<%BANNER%>

Equine Salmonellosis--Molecular Epidemiology of Clinical Isolates and the Effect of Antibiotics on the Cecal Microenviro...


PAGE 1

EQUINE SALMONELLOSIS—MOLECULAR EPIDEMIOLOGY OF CLINICAL ISOLATES AND THE EFFECT OF ANTIBIOTICS ON THE CECAL MICROENVIRONMENT WITH PARTICUL AR REFERENCE TO SHORT-CHAIN FATTY ACIDS AND THE SALMON ELLA PLASMID VIRULENCE ( spv ) GENES By TAMARA SHEA VETRO WIDENHOUSE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 by Tamara Shea Vetro Widenhouse

PAGE 3

I would like to dedicate this work to my family, Mom, Dad, Christopher, Alexis and Carissa—yo u define me. Without you, I am nothing, and cannot imagine my life in your absence. You have made me a better daughter, st udent, teacher, scientist, wife, mother, friend, and human being. You are my universe, and this work is just as much yours as it is mine. I also dedicate this dissertation to every animal that has made the ultimate sacrifice in the name of research. Though it was not by choice, your gifts were never taken for granted, nor will they ever be forgotten.

PAGE 4

iv ACKNOWLEDGMENTS I would like to thank my parents, for ma king sure that I grew up safe, loved, and turned into the kind of person you would not cr oss the street to avoid. They will always be my role models. I would also like to thank my husband Christopher, for standing beside me all these years. Even though a long distance relationship, vet school, a Punky, two doctoral degrees, billions in student loans, a Peach, in-laws, and writing two dissertations might have been enough to va porize any marriage, we have only gotten stronger. May we have many more “adventures” together. I would like to thank my mentor and colleague, Dr. Guy Lester, for th e eternal open door (even though I had to travel across several continen ts to walk through it!), and th e incredible am ount of respect, direction, and free-rein given to me throughout my veterinary school and graduate tenure. A good mentor is difficult to find, and a great one is only dreamt of—he is one of the best. I would also like to thank the few friends who have managed to stick around long after their statute of limitations ran out. I thank Dr. Lori Wendland and Dr. Chris Sanchez, both of whom provided lodgi ng, food, babysitting services, good coffee, libations, a sympathetic ear, and a helping hand whenever it was needed—which was often. I cannot thank them e nough, and am forever grateful. I would especially like to thank Misdee Wrigley-Milligan for her fi nancial support, and the endowment of the Deedie Wrigley-Hancock Fellowship for Equine Colic Research. Her tireless dedication to the University of Florida, College of Ve terinary Medicine, and to the study of equine colic is to be commended. I would like to thank the members of my committee, Dr.

PAGE 5

v Alfred M. Merritt, Dr. Paul A. Gulig, Dr. Saundra TenBroeck, Dr. Steeve Gigure, and Dr. Maureen T. Long. I know at times it s eemed like I had dropped off the face of the earth and that this process might drag on fore ver, but I thank them for their patience and guidance; it is appreciated more th an they will ever realize. I would also like to thank An Nguyen, who went above and beyond the call of duty to provide the isolates and sensitivity data. Most importa ntly, without Hilken V. Kuc k, I would probably still be stuck in the dungeon of a laboratory somewher e or on the Florida Turnpike. He has saved me infinite amounts of time in the lab a nd on the road that was better spent with my family, and have also put up with my “l ess-than-charming” stress-induced personality for many years—I hope he can someday fo rgive me. I sincerely thank him. Last, but certainly not least, I would like to thank those horses and ponies who were intimately involved in my journey, especia lly Tony, Joni, Cody, Oreo, Hide, Seek, Easy, Scott, Bill, Ted, Fly, Rapture, and Willie. Please know that I am forever grateful for their sacrifice and am a better pers on for only having known them.

PAGE 6

vi TABLE OF CONTENTS Page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES...............................................................................................................x LIST OF FIGURES...........................................................................................................xv ABBREVIATIONS...........................................................................................................xx ABSTRACT.....................................................................................................................xx v CHAPTER 1 INTRODUCTION........................................................................................................1 The Genus Salmonella ..................................................................................................1 The Bacteria...........................................................................................................1 Animal Models of Disease....................................................................................2 The Salmonella Virulence Plasmid.......................................................................3 Salmonella Plasmid Virulence ( spv ) Genes..........................................................4 Short-Chain or Volatile Fatty Acids and Salmonella ...................................................6 Antibiotic-Associated Diarrhea (AAD) in the Horse...................................................7 The Gastrointestinal Microenvironment.......................................................................9 The Normal Flora..................................................................................................9 Short-Chain Fatty Acids—Productio n and Intestinal Function...........................10 Effects of Antimicrobial Therapy: Dysbacteriosis..............................................14 Specific Aims..............................................................................................................15 Hypotheses..................................................................................................................16 2 SALMONELLA IN HORSES DISEASE DEFINITION AND GENERAL AND MOLECULAR EPIDEMIOLOGY............................................................................18 Background.................................................................................................................18 Disease Overview................................................................................................18 Prevalence............................................................................................................19 Reported Risk Factors for Salmonella Infection.................................................19 Salmonella Serovars Associated with Equine Infection......................................20 Role of Microbial Virulence Factor s in Equine Salmonella Infection................21

PAGE 7

vii Disease Prevention Diet, Probiotics, Immunity...............................................21 Disease Treatment...............................................................................................24 The Salmonella Virulence Plasmid.....................................................................24 Salmonella Plasmid Virulence ( spv ) Genes........................................................25 Function of the spv genes.............................................................................25 Significance of the spv genes.......................................................................26 Specific Aims..............................................................................................................26 Materials and Methods...............................................................................................27 Case Selection.....................................................................................................27 Microbiological Techniques................................................................................27 Field samples................................................................................................27 Clinical and reference isolates......................................................................28 Salmonella Identification and An tibiotic Resistance Profile...............................30 Salmonella Isolate Storage..................................................................................30 Reference Strains.................................................................................................31 Plasmid Profiling of Salmonella Isolates............................................................31 Polymerase Chain Reaction (PCR) Identification of spv Genes.........................34 Salmonella Plasmid Transformations into Susceptible Bacteria—Effects on Antibiotic Resistance.......................................................................................35 Statistical Methods..............................................................................................37 Results........................................................................................................................ .37 Asymptomatic Population...................................................................................37 Clinical Cases......................................................................................................37 Relationship Between Gender or Age and Outcome...........................................38 Case Seasonality..................................................................................................39 Group and Serovar Distribution..........................................................................40 Outcome by Group or Serovar............................................................................42 Plasmid Profiling.................................................................................................44 spv Gene Analysis...............................................................................................47 Outcome by Presence of the Virulence Plasmid or spv Genes............................54 Effect of Clinical and Labor atory Parameters on Outcome................................56 Relationship Between Proportion of Po sitive Fecal Salmonella Cultures and Outcome...........................................................................................................58 Antibiotic Resistance Profiles.............................................................................58 Antibiotic Resistance Transformation.................................................................59 Site of Salmonella Isolation.................................................................................62 Multi-Serovar Salmonella Infections..................................................................65 Discussion...................................................................................................................67 3 EXPERIMENTS.........................................................................................................75 Background.................................................................................................................75 The Horse: Classic Larg e Intestine Fermenter....................................................75 Equine Cecal Anaerobic Flora and SCFAs in the Normal Animal.....................76 Antimicrobial Effects on Normal Anaerobic Flora.............................................77 Antimicrobial Effects on SCFAs.........................................................................81 Current Theory on the Pathogenesis of Antibiotic-Associated Diarrhea (AAD)82

PAGE 8

viii Effects of SCFA on Anaerobic Growth of Bacteria............................................84 Acid Tolerance Response of Salmonella and Other Enterobacteriaceae............84 SCFA Effects on Salmonella Growth and Invasion............................................86 SCFA Effects on Expression of spv Genes in vitro .............................................88 SCFAs and Salmonella Colonization and Infection of Avian Species................88 SCFAs and Salmonella in Swine.........................................................................89 SCFAs and Salmonella Colonization a nd Infection of Bovine Species..............90 Specific Aims..............................................................................................................90 Materials and Methods...............................................................................................91 IACUC Approval.................................................................................................91 Subject Coding for Experiments and Data Analysis...........................................91 Surgical Placement of Cecal Cannula in the Horse.............................................91 Antibiotic Treatment of Horses...........................................................................94 Equine Cecal Sampling Procedure......................................................................95 Physical Effects on the Horse..............................................................................96 Effects on Fecal Consistency...............................................................................97 Effects on Cecal Content Character....................................................................97 Equine Cecal Anaerobe Quantification...............................................................97 pH Analysis of Equine Cecal Contents...............................................................98 Short-Chain Fatty Acid Analysis of Equine Cecal Contents...............................98 Protozoal Quantification of Cecal C ontents from Horses Treated with Antibiotics........................................................................................................99 In vitro Short-Chain Fatty Acid Growth Comparison.........................................99 In vitro Effects of Cecal Liquor from Antib iotic-treated Horses on Anaerobic Growth of Salmonella ....................................................................................100 Statistical Methods............................................................................................102 Results.......................................................................................................................1 02 Effects on the Horse..........................................................................................102 Effects on Cecal pH...........................................................................................103 Effects on Cecal Protozoal Counts....................................................................104 Effects on Cecal SCFA Quantities and Proportions..........................................106 Effects on Cecal Anaerobic Bacteria.................................................................111 In vitro Effects of SCFAs on Anaerobic Growth of Salmonella .......................113 Effect of Acetate................................................................................................116 Effect of the Plasmid and spv Genes on Acetate Response...............................116 Effect of Butyrate..............................................................................................117 Effect of the Plasmid and spv Genes on Butyrate Response.............................117 Effect of Propionate...........................................................................................118 Effect of the Virulence Plasmid and spv Genes on Propionate Response.........118 In vitro Effects of Cecal Liquor from Antib iotic-treated Horses on Anaerobic Growth of Salmonella ....................................................................................119 Effect of the Virulence Plasmid and spv Genes on Anaerobic Growth of Salmonella Exposed to Cecal Liquor from Antibiotic-Treated Horses.........122 Discussion.................................................................................................................122

PAGE 9

ix 4 SUMMARY, CONCLUSIONS, AND FUTURE DIRECTIONS............................133 APPENDIX A SALMONELLA EPIDEMIOLOGY DATA COLLECTION SHEET....................139 B INDEX OF SUPPLIERS AND CONTACT INFORMATION................................141 C SALMONELLA ISOLATE INDEX........................................................................143 D SALMONELLA DATABASE CASE DESCRIPTIVE INFORMATION..............146 E SALMONELLA ISOLATE ANTIMICRO BIAL SUSCEPTIBILITY DATA........174 F DESCRIPTIVE STATISTICS..................................................................................187 LIST OF REFERENCES.................................................................................................196 BIOGRAPHICAL SKETCH...........................................................................................215

PAGE 10

x LIST OF TABLES Table page 1-1. Short-chain fatty acid chemi cal formulas and common names.................................10 2-1. spv gene characteristics.............................................................................................26 2-2. Composition of bact erial culture media.....................................................................29 2-3. Salmonella serovar Typhimurium reference strains used in this study.....................31 2-4. Composition of buffers and solutions used in plasmid extraction protocols.............33 2-5. Times and temperatures for PCR reactions...............................................................34 2-6. Primers utilized in PCR reactions..............................................................................35 2-7. Breed distribution of 84 equine salmonella cases 1999-2002...................................38 2-8. Effect of gender on mortality in 96 cases of equine salmonellosis...........................38 2-9. Effect of age on mortality in 85 cases of equine salmonellosis.................................39 2-10. Average minimum temperatures in Gainesville, Florida, USA (1961-1990)..........40 2-11. Salmonella serovars isolated from 98 equine cases 1999-2002..............................41 2-12. Salmonella isolates of environmental a nd species other than equids collected 1999-2002.................................................................................................................41 2-13. Effect of salmonella group on mortalit y in 88 cases of equine salmonellosis........44 2-14. Plasmid-positive salmonella isolates by serovar 1999-2002...................................44 2-15. Summary outcome as determined by presence of the virulence plasmid and spv genes in 98 equine salmonella cases........................................................................54 2-16. Effect of spv gene presence on mortality in 86 cases of equine salmonellosis where outcome was known.................................................................................................55 2-17. Logistic regression model wi th variables predictive of outcome............................57

PAGE 11

xi 2-18. Antibiotic susceptibilities for 101 equine salmonella isolates. The reported % susceptible, % intermediate, and % resistan t, are only for those isolates with data for that antibiotic......................................................................................................58 2-19. Antibiotic susceptibility report for Ca se 78, with intermediate resistance to enrofloxacin..............................................................................................................59 2-20. Clinical salmonella isolates from 105 equine cases by location of cultured specimen...................................................................................................................62 2-21. Systemic sites of salmonella infection in horses by serovar...................................63 2-22. Relationship of the virulence plasmid and spv genes to isolate location in 98 cases of equine salmonellosis............................................................................................64 2-23. Details of multiserovar salmonella infections in six horses 1999-2002..................66 3-1. Summary of literature reports quan tifying equine cecal anaerobic bacteria.............76 3-2. Literature reports quantifying norm al equine cecal SCFA concentrations...............77 3-3. Literature summary of antibiotic effects on f ecal bacteria and short-chain fatty acids..................................................................................................................79 3-4. Summary of single dose antibiotic eff ects on the equine cecal microenvironment..82 3-5. Coding legend for experimental animals...................................................................91 3-6. Antibiotic treatments of horses..................................................................................94 3-7. Cecal liquor pH of cannulated horses before and after 4 days of control (no) antibiotic treatment.................................................................................................104 3-8. Cecal liquor pH of cannulated horses trea ted with intramuscular ceftiofur sodium at 2 mg/kg twice daily, before a nd after 4 days of treatment.....................................104 3-9. Cecal liquor pH of cannulated horses tr eated with intravenous oxytetracycline at 10 mg/kg once daily, before and after 4 days of treatment.........................................104 3-10. Cecal liquor pH of cannulated hor ses treated with oral trimethoprimsulfamethoxazole at 30 mg/kg twice daily, be fore and after 4 days of treatment..104 3-11. Total protozoal counts per ml of cecal c ontents from cannulated horses before and after 4 days of control (n o) antibiotic treatment.....................................................105 3-12. Total protozoal counts per ml of cecal cont ents from cannulated horses treated with intramuscular ceftiofur sodium at 2 mg/kg twice daily, before and after 4 days of treatment.................................................................................................................105

PAGE 12

xii 3-13. Total protozoal counts per ml of cecal cont ents from cannulated horses treated with intravenous oxytetracycline at 10 mg/kg on ce daily, before and after 4 days of treatment.................................................................................................................105 3-14. Total protozoal counts per ml of cecal cont ents from cannulated horses treated with oral trimethoprim-sulfamethoxazole at 30 mg/kg twice daily, before and after 4 days of treatment....................................................................................................105 3-15. Cecal liquor concentrations of individual and total SCFA s from cannulated horses, before and after 4 days of control (no) treatment...................................................107 3-16. Cecal liquor concentrations of individual and total SC FAs from cannulated horses treated with intramuscular ceftiofur sodium at 2 mg/kg twice daily, before and after 4 days of treatment.................................................................................................108 3-17. Cecal liquor concentrations of individual and total SC FAs from cannulated horses treated with intravenous oxyt etracycline at 10 mg/kg once daily, before and after 4 days of treatment....................................................................................................109 3-18. Cecal liquor concentrations of individual and total SC FAs from cannulated horses treated with oral trimethoprim-sulfamet hoxazole at 30 mg/kg twice daily, before and after 4 days of treatment..................................................................................110 3-19. Mean counts of culturable anaerobic bacteria from seri al dilutions of raw equine cecal liquor, from 5 cannulated horses, befo re and after 4 days of control (no) treatment. The dilution shaded in green was chosen for comparison...................111 3-20. Mean counts of culturable anaerobic bacteria from seri al dilutions of raw equine cecal liquor, from 5 cannulated horses trea ted with intramuscular ceftiofur sodium at 2 mg/kg twice daily, before and after 4 da ys of treatment. The dilution shaded in green was chosen for comparison..........................................................................112 3-21. Mean counts of culturable anaerobic bacteria from seri al dilutions of raw equine cecal liquor, from 5 cannulated horses trea ted with intravenous oxytetracycline at 10 mg/kg once daily, before and after 4 days of treatment. The dilution shaded in green was chosen for comparison..........................................................................112 3-22. Mean counts of culturable anaerobic bacteria from seri al dilutions of raw equine cecal liquor, from cannulated horses treated with oral trimethoprimsulfamethoxazole at 30 mg/kg twice daily, be fore and after 4 days of treatment. The dilution shaded in green was chosen for comparison.....................................112 C-1. Index of salmonella isolates by group and serovar, with plasmid and spv gene status..............................................................................................................143 D-1. Salmonella case descriptive information: breed, age, sex, presenting complaint, risk factors for salmonellosis, specimen origin, and salmonella group(s) and serovar(s). Blank cells indicate missing or unavailable records..............................................147

PAGE 13

xiii D-2. Salmonella case descriptive information: serovar, date sample taken, presence of diarrhea, total hospitalization cost, case outcome, hospitalization days, number of positive cultures, hematologic indices at time of positive culture, and total protein changes during hospitalizati on. Blank cells indicate missing or unavailable records....................................................................................................................159 D-3. Salmonella case descriptive informati on: serovar, antibiotic therapy prior to admission and types, antibiotic therapy dur ing hospitalization and types. Drugs in boldface type were used specifically to tr eat the salmonella infection. Blank cells indicate missing or unavailable records, a nd three dashes indicates that nothing could be determined from the record.....................................................................165 E-1. Salmonella isolate MIC antibiotic sensi tivity profiles. Blank cells indicate missing data. Legend: AMI = amikacin, AMOX = amoxicillin-clavulanic acid, AMP = ampicillin, CEFA = cefazolin, CEFZ = ceftazidime, NAX = ceftiofur, CHLP = chloramphenicol.....................................................................................................175 E-2. Salmonella isolate MIC antibiotic sensitiv ity profiles. Blank cells indicate missing data. Legend: CLIN = clindamycin, DOX = doxycycline, ENRO = enrofloxacin, ERYT = erythromycin, GENT = gentamicin, IMIP = imipenem..........................179 E-3. Salmonella isolate MIC antibiotic sensi tivity profiles. Blank cells indicate missing data. Legend: NITR = nitrofurantoin, OX = oxacillin, PEN = penicillin, RIF = rifampin, TET = tetracycline, TMP = trimethoprim-sulfamethoxazole.................183 F-1. Descriptive statistics for individual cecal SCFA measurements before and after four days of control (no) antibiotic treatment in 5 horses..............................................187 F-2. Descriptive statistics for individual cec al SCFA measurements in 5 horses treated with intramuscular ceftiofur sodium at 2 mg/kg twice daily, before and after 4 days of treatment............................................................................................................187 F-3. Descriptive statistics for individual cec al SCFA measurements in 5 horses treated with intravenous oxytetracyclin e at 10 mg/kg once daily, before and after 4 days of treatment.................................................................................................................188 F-4. Descriptive statistics for individual cec al SCFA measurements in 5 horses treated with oral trimethoprim-sulfamethoxazole at 30 mg/kg twice daily, before and after 4 days of treatment.................................................................................................188 F-5. Descriptive statistics fo r cecal protozoal counts in 5 hor ses treated with control (no treatment), ceftiofur sodium, oxytetracyclin e, or trimethoprim-sulfamethoxazole, before and after 4 days of treatment.......................................................................189 F-6. Descriptive statistics for cecal pH in 5 horses treated with control (no treatment), ceftiofur sodium, oxytetracycline, or tr imethoprim-sulfamethoxazole, before and after 4 days of treatment.........................................................................................189

PAGE 14

xiv F-7. Descriptive statistics for salmonella gr owth in M9 supplemented with 10% sterile filtered cecal contents from 5 individual horses treated with oxytetracycline (TET), or trimethoprim-sulfamethoxazole (TMPS)...........................................................189 F-8. Descriptive statistics for salmonella gr owth in M9 supplemented with 10% sterile filtered cecal contents from 5 individual horses treated with control (no treatment) or ceftiofur sodium (NAX).....................................................................................190 F-9. Descriptive statistics for all salmonella growth in LB broth supplemented with 10% sterile filtered cecal contents pooled from 5 horses treated with no treatment, ceftiofur (NAX), oxytetracycline (TET), or trimethoprim-sulfamethoxazole (TMPS). Units=CFU/ml, N=number of dilutions counted...................................191 F-10. Descriptive statistics for all salm onella growth in M9 minimal medium supplemented with 10% sterile filtered cecal contents pool ed from 5 horses treated with control (no treatment) or ceftiofu r sodium (NAX), oxytetracycline (TET), or trimethoprim-sulfamethoxazole (TMPS). Units for mean measurement are CFU/ml, and N=number of dilutions counted........................................................192 F-11. Descriptive statistics for salmonella gr owth in M9 minimal medium supplemented with sodium chloride, acetate, butyr ate, or propionate at 30 or 100mM...............193

PAGE 15

xv LIST OF FIGURES Figure page 1-1. spv gene expression regulati on is dependent on growth pha se and cellular location in Salmonella ..............................................................................................................5 1-2. Polysaccharide metabolism and SCFA production pathways in the rumen..............11 1-3. Diagram of major gast rointestinal microbial dige stive and energy functions, nitrogen and carbon recycling, and SCFA production.............................................12 1-4. Summary of potential en terotrophic eff ects of SCFA...............................................14 2-1. API20E rapid identification stri p showing typical reaction results for Salmonella species......................................................................................................................28 2-2. Salmonella group C2 isolate as provided on Hektoen-Enteric agar...........................28 2-3. Age distribution of 98 equine salmonella cases 1999-2002......................................39 2-4. Seasonal distribution of sa lmonella cases from horses 1999-2002...........................40 2-5. Mortality distribution, within serovar, of non-surviving e quine salmonella cases 1999-2002.................................................................................................................43 2-6. Mortality by salmonella group in 88 cases with known outcomes............................43 2-7. Plasmid profiles of 9 clin ical salmonella isolates. Refe r to Appendix C for specific isolate information. Lanes: 1) Prev iously extracted 100-kb plasmid of 3306, 2) Case 8, 3) Bovine isolate of S Typhimurium var. Copenhagen, 4) Case 11, 5) Bovine isolate of S Typhimurium var. Copenhagen, 6) Case 12, 7) Bovine isolate of S Typhimurium var. Copenhagen, 8) Ca se 6, 9) Case 10, 10) Case 7................47 2-8. Plasmid profiles of 4 clin ical salmonella isolates. Refe r to Appendix C for specific isolate information. Lanes: 1) superco iled marker DNA, 2) 100-kb plasmid of 3306, 3) Case 78, 4) Case 71, 5) Case 66, 6) Case 77, 7) 100-kb plasmid of 3306, 8) blank.....................................................................................................................48

PAGE 16

xvi 2-9. Plasmid profiles of 5 clin ical salmonella isolates. Refe r to Appendix C for specific isolate information. Lanes: 1) superco iled marker DNA, 2) 100-kb plasmid of 3306, 3) Case 89, 4) Case 92, 5) Case 85, 6) Case 93, 7) Case 96, 8) 100-kb plasmid of 3306......................................................................................................49 2-10. Plasmid profiles of 4 clini cal salmonella isolates. Refe r to Appendix C for specific isolate information. Lanes: 1) blank, 2) 100-kb plasmid of 3306, 3) Aged (>1month) plasmid extract of 3306, 4) Case 46, 5) Case 44, 6) Case 43, 7) Case 53, 8) blank...............................................................................................................49 2-11. Plasmid profiles of 4 clini cal salmonella isolates. Refe r to Appendix C for specific isolate information. Lanes: 1) superco iled marker DNA, 2) 100-kb plasmid of 3306, 3) Aged (>2month) plasmid extract of 3306, 4) Case 41, 5) Case 63, 6) Case 64, 7) Case 65, 8) s upercoiled marker DNA...................................................50 2-12. Plasmid profiles of 3 clini cal salmonella isolates. Refe r to Appendix C for specific isolate information. Lanes: 1) superco iled marker DNA, 2) 100-kb plasmid of 3306, 3) 100-kb plasmid of 3306, 4) Case 82, 5) Case 83, 6) Case 40, 7) 100-kb plasmid of 3306, 8) blank.......................................................................................50 2-13. Plasmid profiles of 7 clini cal salmonella isolates. Refe r to Appendix C for specific isolate information. Lanes: 1) 100-kb plasmid of 3306, 2) Case 32, 3) Case 36, 4) Case 37, 5) Case 91, 6) Case 90, 7) Case 87, 8) Case 86.........................................51 2-14. PCR product results for spvA and spvC genes in 9 clinical salmonella isolates, with positive and negative controls. Refer to Appendix C for specific isolate information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) 3337 spv negative control, 3) 3306 spv positive control, 4) Case 86, 5) Case 87, 6) Case 88, 7) Case 117, 8) Case 100, 9) Case 101, 10) Case 102, 11) Case 103, 12) Case 104, 13) 1-kb ladder DNA marker (Promega), 14) blank...............................................................51 2-15. PCR product for asd gene in 9 clinical salmonella isolates (same isolates and orientation as Figure 2-14). Refer to Appendix C for specific isolate information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) 3337 spv negative control, 3) 3306 spv positive control, 4) Case 86, 5) Case 87, 6) Case 88, 7) Case 117, 8) Case 100, 9) Case 101, 10) Case 102, 11) Case 103, 12) Case 104, 13) 1-kb ladder DNA marker (Promega), 14) blank..........................................................................52 2-16. PCR product results for spvA genes in 11 clinical salmonella isolates, with positive and negative controls. Refer to Appendix C for specific isolate information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) 3306 spv positive control, 3) 3337 spv negative control, 4) lost isolate, 5) Case 8, 6) Case 7, 7) Case 12, 8) Case 13, 9) Case 10, 10) Case 9, 11) Case 5, 12) Case 3, 13) Case 11, 14) Case 4....................53

PAGE 17

xvii 2-17. PCR product results for spvC genes in 11 clinical salmonella isolates, with positive and negative controls. Refer to Appendix C for specific isolate information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) 3306 spv positive control, 3) 3337 spv negative control, 4) Case 21, 5) Case 19, 6) Case 16, 7) Case 22, 8) Case 27, 9) Case 26, 10) Case 25, 11) Case 24, 12) Ca se 23, 13) Case 15, 14) Case 14............53 2-18. PCR product results for the asd gene in 11 clinical sa lmonella isolates (same isolates and orientation as Figure 2-17). Refer to Appe ndix C for specific isolate information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) 3306 spv positive control, 3) 3337 spv negative control, 4) Case 21, 5) Case 19, 6) Case 16, 7) Case 22, 8) Case 27, 9) Case 26, 10) Case 25, 11) Case 24, 12) Case 23, 13) Case 15, 14) Case 14.....................................................................................................................54 2-19. Outcome in equine salmonella cases as influenced by presence of the spv gene locus.........................................................................................................................5 5 2-20. Outcome in equine salmonella cases, as influenced by absence of the spv gene locus.........................................................................................................................5 6 2-21. Plasmid profiles of 3 clin ical salmonella isolates and E. coli transformed with plasmid DNA from those isolates. Refe r to Appendix C for specific isolate information and Appendix E for antimicrobi al susceptibilities Lanes: 1) 100-kb plasmid of 3306, 2) Untransformed E. coli DH5 3) E. coli DH5 transformed with Case 97, grown in CEF, 4) E. coli DH5 transformed with Case 92, grown in CEF, 5) E. coli DH5 transformed with Case 98, grown in CEF, 6) Transforming plasmid DNA from Case 97, 7) Transf orming plasmid DNA from Case 92, 8) Transforming plasmid DNA from Case 98..............................................................60 2-22. Plasmid profiles of 2 clin ical salmonella isolates and E. coli transformed with plasmid DNA from those isolates. Refe r to Appendix C for specific isolate information and Appendix E for antimicr obial susceptibilities. Lanes: 1) Untransformed E. coli DH5 2) 100-kb plasmid of 3306, 3) E. coli DH5 transformed with Case 98, grown in AMP, 4) E. coli DH5 transformed with Case 98, grown in NAX, 5) E. coli DH5 transformed with Case 98, grown in CEF, 6) E. coli DH5 transformed with Case 92, grown in AMP, 7) E. coli DH5 transformed with Case 92, grown in CEF, 8) blank.....................................................................60 2-23. Plasmid profiles of 2 clin ical salmonella isolates and E. coli transformed with plasmid DNA from those isolates. Refe r to Appendix C for specific isolate information and Appendix E for antimicr obial susceptibilities. Lanes: 1) Untransformed E. coli DH5 2) 100-kb plasmid of 3306, 3) E. coli DH5 transformed with Case 97, grown in AMP, 4) E. coli DH5 transformed with Case 92, grown in NAX, 5) E. coli DH5 transformed with Case 97, grown in NAX, 6) E. coli DH5 transformed with Case 97, grown in CEF, 7) blank, 8) blank...........61

PAGE 18

xviii 2-24. Plasmid profile of Case 97—lane 5. Th e red box delineates 3 large plasmid bands that are visible in the upper part of the lane. This isolate transferred ceftiofur, cefazolin, and ampicillin resistance via two different plasmids (the lower two).....62 2-25. Systemic equine salmonella isolates compared to gastrointestinal isolates by group....................................................................................................................64 3-1. Pathogenesis of anti biotic-associated diarrhea..........................................................83 3-2. Components of indwelling cecal cannul a placed into experimental horses. Clockwise from the top: side view of sliding flange placed on the lateral serosal aspect of cecal wall, cannula with fixed interior flange, silicone filled, thick walled tubing used to plug cannula, front view of sliding flange, hose clamp to secure plug within cannula..........................................................................................................92 3-3. Experimental horse E (2) with cecal cannula 3 years post-implantation..................93 3-4. Close-up view of cannula in situ in experimental horse E (2 ). Note the formation of a firm swelling intimately associated with the cannula insertion. This is internal granulation tissue forming around the interior silicone flanges which will result in the eventual expulsion of the device........................................................................93 3-5. Collection of equine cecal contents fr om indwelling silicone cannula. Note the rapid flow and liquid nature of the contents.............................................................96 3-6. Mean cecal anaerobic culture counts e xpressed as CFU / ml of liquor from five horses before and after treatment with control (no treatment), ceftiofur, oxytetracycline, or trimethoprim-sulfamethoxazole..............................................113 3-7. The effect of LB broth with sodium ch loride (control treatment) added at 30mM or 100mM on anaerobic growth of S. Typhimurium 3306 vs. 3337. All solutions were pH 6.5. Error bars re present 95%CI for two repli cates of the experiment...114 3-8. The effect of LB broth with sodium acetate added at 30mM or 100mM on anaerobic growth of S. Typhimurium 3306 vs. 3337. All solutions were pH 6.5. Error bars represent 95%CI for two repli cates of the experiment...........................................114 3-9. The effect of LB broth with sodi um butyrate added at 30mM or 100mM on anaerobic growth of S. Typhimurium 3306 vs. 3337. All solutions were pH 6.5. Error bars represent 95 %CI for two replicates of the experiment..........................115 3-10. The effect of LB broth with sodi um propionate added at 30mM or 100mM on anaerobic growth of S. Typhimurium 3306 vs. 3337. All solutions were pH 6.5. Error bars represent 95 %CI for two replicates of the experiment..........................115 3-11. The effect of LB broth with sodium acetate at 30mM or 100mM compared to NaCl at 30mM or 100mM on anaerobic growth of S. Typhimurium. All solutions were pH 6.5. Error bars represent 95%CI for four replicates of the experiment...........116

PAGE 19

xix 3-12. The effect of LB broth with sodium butyrate at 30mM or 100mM compared to NaCl at 30mM or 100mM on anaerobic growth of S. Typhimurium. All solutions were pH 6.5. Error bars repr esent 95%CI for four replic ates of the experiment..117 3-13. The effect of LB broth with sodium propionate at 30mM or 100mM compared to NaCl at 30mM or 100mM on anaerobic growth of S. Typhimurium. All solutions were pH 6.5. Error bars repr esent 95%CI for four replic ates of the experiment..118 3-14. The effect of LB broth with 10% added filter-sterilized cecal contents pooled from five horses by treatment on the anaerobic growth of S. Typhimurium. The horses were treated with control (no treatmen t), ceftiofur (NAX), oxytetracycline (TET), or trimethoprim-sulfamethoxazole (TMPS)...........................................................119 3-15. The effect of M9 minimal medium (+ glucose) with 10% added filter-sterilized cecal contents pooled from five horses by treatment on the anaerobic growth of S. Typhimurium. The horses were treated with control (no treatment), ceftiofur (NAX), oxytetracycline (TET), or trim ethoprim-sulfamethoxazole (TMPS)........120 3-16. The effect of M9 minimal medium (+ glucose) with 10% added filter-sterilized cecal contents from antibiotic-treate d horses on the anaerobic growth of S. Typhimurium. Data points are the mean of 5 individual horses tr eated with control (no treatment) or ceftiofur (NAX). Time 6h is a missing data point....................121 3-17. The effect of M9 minimal medium (+ glucose) with 10% added filter-sterilized cecal contents from antibiotic-treate d horses on the anaerobic growth of S Typhimurium. Data points are the mean of 5 individual hor ses treated with oxytetracycline (TET) or trimethopr im-sulfamethoxazole (TMPS)......................122

PAGE 20

xx ABBREVIATIONS % percent(age) positive negative pH pH gradient almost equal to C degrees Centigrade (Celsius) F degrees Fahrenheit 95%CI 95% confidence interval A acetate AAD antibiotic-associated diarrhea ADH test for arginine dihyd rolase, red/orange = + ADP adenosine diphosphate AMI amikacin AMOX amoxicillin-clavulanic acid AMP ampicillin AMY amygdalin fermentation/oxidation test, yellow = + ARA arabinose fermentation/oxidation test, yellow = + asd aspartate semialdehyde dehydrogenase ASP acid shock protein ATP adenosine triphosphate ATR acid tolerance response B butyrate BHI brain heart infusion bp base pairs C. Clostridium CaCl2 calcium chloride CEC competent Escherichia coli CEF or CEFA cefazolin CEFZ ceftazidime cfu colony forming units CH4 methane CHLP chloramphenicol CIT test for citrate utiliz ation, blue-green/blue = + CLIN clindamycin cm centimeter CO2 carbon dioxide COD cause of death

PAGE 21

xxi CON control df degrees of freedom DMC direct microscopic count DNA deoxyribonucleic acid dNTP DNA nucleotides (A,C,G,T) DOA dead on arrival DOX doxycycline E. Escherichia e.g. for example EB ethyl butyrate EDTA ethylene diamine tetra acetic acid ENRO enrofloxacin ERYT erythromycin et al. and others euth. euthanatized ex vivo outside the living body FOS fructo-oligosaccharides FUO fever of unknown origin g grams g gravity, 10-11 N.m/s2 GDUD gastro duodenal ulcer disease GEL gelatinase production te st, diffusion of black = + GENT gentamicin GLU glucose fermentation/oxidation test, yellow = + gyr gyrase h hour(s) H2 hydrogen H2O2 hydrogen peroxide H2S test for hydrogen sulfide production, black = + HCl hydrochloric acid HE Hektoen-Enteric agar i.e. that is IACUC Institutional Animal Care and Use Committee IB isobutyrate ICH iodochlorhydroxyquin IM intramuscular IMIP imipenem in vivo inside the living body IND test for indole production, red = + INO inositol fermentation/oxidation test, yellow = + IV intravenous IVA isovalerate kb kilobase(s) kg kilogram kV kilovolts

PAGE 22

xxii l or L liter lb pound (weight) LB Luria-Bertani LBN Luria-Bertani (sodium) LDC test for lysine decarboxylase, red/orange = + LI large intestine log10 logarithm base 10 M molar m meters M molar M9 minimal media MAN mannitol fermentation/oxidation test, yellow = + MDa megadaltons MEL melibiose fermentation/oxidation test, yellow = + mg milligram mg/kg milligrams per kilogram bodyweight MgCl2 magnesium chloride MgSO4 magnesium sulfate MIC minimum inhibitory concentration min minutes ml milliliter mM millimolar mm millimeters mmol millimoles MOPS morpholinepropanesulphoni c acid (buffer solution) MQMFK modified Qiagen Midi F ilter Kit for plasmid analysis mRNA messenger ribonucleic acid N normal N/A not applicable NaCl sodium chloride NAHMS National Animal Health Monitoring System NAL nalidixic acid Nalr nalidixic acid resistant NaOH sodium hydroxide NAX ceftiofur sodium ND none determined NG nasogastric NITR nitrofurantoin No. number NVFA non-volatile fatty acids ODC test for ornithine decarboxylase, red/orange = + ONPG test for beta galactosidase, yellow = + OX oxidase test, violet = + P propionate PBS phosphate buffered saline

PAGE 23

xxiii PCR polymerase chain reaction PEN penicillin PF pelvic flexure PFGE pulsed field gel electrophoresis pg picograms PGMAA pH-gradient mediated anion accumulation pH negative logarithm of hydrogen ion concentration pKa negative logarithm of the acid dissociation constant Ka PMN polymorphonuclear leukocyte PO per os (orally) ppm parts per million PRAS pre-reduced anaerobically sterilized q every QBT Equilibration Buffer (Qiagen Midi Filter Kit) QC Wash Buffer (Qiagen Midi Filter Kit) QF Elution Buffer (Qiagen Midi Filter Kit) R plasmid or factor resistance plasmid or factor RFLP restriction fragment length polymorphism RHA rhamnose fermentation/oxidation test, yellow = + RIF rifampin rpm revolutions per minute rpoS alternative sigma factor (referring to the gene) rpoS alternative sigma factor (referring to the protein) RT room temperature s second(s) S. Salmonella SAAAD Salmonella-attributed antibiotic-associated diarrhea SAC sucrose fermentation/oxidation test, yellow = + SC small colon SCFA short-chain fatty acid SD standard deviation SDS sodium dodecyl sulfate SEM standard error of the mean SI small intestine SOR sorbitol fermentation/oxidation test, yellow = + SPF specific pathogen free spp bacterial species spv Salmonella plasmid virulen ce (referring to the gene) spv Salmonella plasmid virulen ce (referring to the protein) subsp. sub-species TBE tris-borate EDTA TDA test for deaminase, brown/red = + TE tris-EDTA TET (oxy)tetracycline TMP or TMPS trimethoprim sulfamethoxazole

PAGE 24

xxiv TNTC too numerous to count TSP or TP total serum protein or total protein U units URE test for urea hydrol ysis, red/orange = + V valerate v/v volume per volume var. variant (serovariant) VFA volatile fatty acid VMTH Veterinary Medi cal Teaching Hospital VP Voges-Proskauer test for acetoin, pink/red = + w/v weight per volume wt. weight bacterial strain

PAGE 25

xxv Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EQUINE SALMONELLOSIS—MOLECULAR EPIDEMIOLOGY OF CLINICAL ISOLATES AND THE EFFECT OF ANTIBIOTICS ON THE CECAL MICROENVIRONMENT WITH PARTICUL AR REFERENCE TO SHORT-CHAIN FATTY ACIDS AND THE SALMON ELLA PLASMID VIRULENCE ( spv ) GENES By Tamara Shea Vetro Widenhouse May 2004 Chair: Guy D. Lester Major Department: Veterinary Medicine Antibiotic-associated diarrhea (AAD) is a common and potentially fatal disorder in horses and is often attributable to Salmonella spp. Disturbances in anaerobic microflora are thought to cause altered inte stinal levels of bacteriost atic short-chain fatty acids (SCFA). Salmonella virulence plasmid ( spv ) genes are reported to increase ability of Salmonella to grow in organs outside the gastroin testinal tract. Horses treated with intravenous oxytetracycline (TET), oral tr imethoprim-sulfamethoxazole (TMPS), and intramuscular ceftiofur (NAX) had significant differences in concentrations of seven individual cecal SCFA with TET having the mo st significant effects, followed by TMPS and NAX. No differences were detected in cecal protozoal counts, total culturable cecal anaerobes, or cecal pH compared to untreated horses. Epidemiological techniques were used to investigate 106 cases of salmonella infection in horses at a veterinary teaching hospital over 2 years. Total mortality was

PAGE 26

xxvi 36.5%. Plasmid profiles, spv gene analysis, serovar, and antibiotic sensitivity data were recorded for all isolates. O dds ratios predicted an increase d risk of a fatal outcome in horses younger than 4 years of age (3.3 times), horses infected with group B salmonellae (15.7 times relative to group D), and horse s whose salmonella isolate possessed the spv genes (12.3 times). Extra-intestinal salmone lla isolates were 12.2 times more likely to contain the spv genes. The majority of large plas mids in salmonella serovars isolated from horses were not virulence plasmids, but likely antibiotic resistance plasmids (3/3 tested transferred multiple resistances) This information suggests that the spv genes may play a similar role in horses as they do in humans, mice, and calves: to potentiate systemic infection after gastrointestinal infection. Sterile-filtered cecal liquor from horses treated with ceftiofur or trimethoprimsulfamethoxazole increased the in vitro anaerobic growth rates of Salmonella relative to plain media, and slightly more than untreated control horses cecal liquor. Salmonella grew equally as well (but much slower th an NAX or TMPS) in TET treated horses cecal liquor and plain M9 medium. The SCFAs acetat e, butyrate, and propionate, added to M9 minimal medium at 30mM or 100mM, e xhibited a dose-dependent inhibition of anaerobic salmonella growth that was not attributable to the spv genes, with propionate 100mM > butyrate 100mM > acetate 100mM propionate 30mM > butyrate 30mM > acetate 30mM.

PAGE 27

1 CHAPTER 1 INTRODUCTION The Genus Salmonella The Bacteria The first mention of the yet-to-be-named genus Salmonella was a report in 1880 on a “typhoid bacillus” observed in the spleen and mesenteric lymph nodes of a fever patient.1 A second organism discovered at approxi mately the same time, which failed to agglutinate in serum from typhoid patients, was designated “bacille paratyphique.” The first documented cases of salmonellosis in animals were described by Salman and Smith in 1886 of swine affected with hog cholera. This bacterium was later designated S. Choleraesuis, and the genus eventually named after the former.1 Salmonellae are gram-negative members of the family Enterobacteriaceae. As of August 2002, the genus is represented by 2,523 distinctive serova riants (serovars)2 of flagellated, facultativ ely anaerobic bacilli.3 Salmonellae are speciated and subcharacterized by their O (LPS), H (flagellar), and Vi (capsular) antigens. O antigens are located on the surface of the outer memb rane and are determined by specific polysaccharide sequences. H antigens are expressed on flagella, and they are composed of the proteins called flagellin. H antigens ar e biphasic and occur in either or both of two forms, phase 1 and phase 2. The bacteria ar e capable of switching from one phase to the other depending on environmental pressures.1 Vi is a unique antig en in that it overlays the O antigen and is present in a limited num ber of serovars, the most important being Salmonella Typhi, a host-adapte d serovar of humans.4;5 These bacteria are stable and

PAGE 28

2 ubiquitous in the environment, and they are ca pable of colonizing and infecting nearly all higher species, although some serovars are know n to have host pref erences as well as syndrome phenotypes. Animal Models of Disease Non-typhoidal salmonellae are global ente ric pathogens of humans and other vertebrates, and decades of research have been devoted to the epidemiology, pathogenesis, diagnosis, control, and effectiv e treatment of salmonellosis. To date, the most economic and thoroughly characterized an imal model of salmonellosis has been the mouse. Salmonella infection in the mouse typically produces a syndrome of fever and bacteremia. Until recently, it was thought that the mouse species did not develop enteritis secondary to orogastri c inoculation with Salmonella ,6 which is typical of the pathogenesis in most mammals, including humans and horses.7 In humans, cows, and horses, the inflammatory reaction of the gastrointestinal tract is predominantly neutrophilic, while in the mouse the mononuclear cell is the principal inflammatory cell.8 The serovarand route-dependant clinical response to experiment al infection in the mouse is different to that of most other vertebrates. Watson et al showed that the cellular route of intestinal invasion is different between mice and calves, with M cells and Peyer’s patches being the preferred targets in the murine host in contrast to enterocytes in calves.9 This creates difficulty in the extrapolation of experiment al data from mice to larger mammals. Several alternative models to human gastroin testinal salmonellosis have been developed, using the calf10 or pig,11;12 but these are expensive, logist ically difficult to maintain, and carry significant animal welfare concerns.4;6 In 2003, a newly proposed mouse model of enteric salmonellosis was described and successfu lly tested by Barthel et al. in Belgium.13 This model more closely approximates the neut rophilic inflammatory infiltrate seen in

PAGE 29

3 response to the bacteria, yet the mice still do not become diarrheic. Despite this, the model is still a step forw ard in terms of salmonella investigation on genetic, immunologic, and environmental levels, as th ere are readily availa ble genetic knockout strains and immunohistologic media and protoc ols developed for the mouse species. The Salmonella Virulence Plasmid The term plasmid is used to describe autonomously replicating extrachromosomal DNA. This DNA is not critical to cell survival in vitro but can confer specific characteristics that allow the host cell to su rvive during adverse conditions or to cause disease.14 Pathogenic salmonellae possess a coll ection of these attributes, called virulence factors.15;16 These include factors that c onvey acid resistance, enhance the ability to invade non-phagocytic cells, elicit inflammation, support resistance to destruction by phagocytic cells, suppress th e immune system of the host organism, enhance intracellular replication, and encode an timicrobial resistance. Several of these factors can be attributed to the presence of a large 50-100 kb plasmid, originally termed the “cryptic plasmid” as its purpose was unclear, but now described as a virulence plasmid.17;18 The virulence plasmid of Salmonella has been characterized extensively in the mouse typhoid model and appears to be most important in the ability of the organism to multiply in systemic tissues after dissemination from the gastrointestinal tract.19-21 Clinical significance of this virulence plasmid has been exam ined in several studies, and there remains disagreement regarding contributi on of the virulence plasmid to bacteremia and replication in extra-intestinal tissues. Clinical isolate data from several human studies also agree with the murine model: virulence plasmids are more likely to be present in those isolates obtained from syst emic sources such as blood, liver and spleen, compared to unrelated isolates obtained from feces.22;23 Conflicting re ports utilizing

PAGE 30

4 comparable experimental methods have show n no causal relationship between bacteremia and presence of the virulence plasmid in humans.24;25 Discordant results are also seen within the model using calv es orally infected with S. Dublin. The virulence plasmid containing the spv genes was shown to be important in determining the severity of diarrhea in calves,10 while other investigators demonstrat ed no influence of the virulence plasmid (as compared to naturally occurring plasmid-free, or plasmid-cured isolates) on enteropathogenesis either in vivo or in ex vivo ligated ileal loop experiments.26 This serovar-host-syndrome interrelationship is most certainly a confounding factor in determining the pathophysiologic importanc e of the salmonella virulence plasmid. Salmonella Plasmid Virulence ( spv ) Genes The plasmids of several serovars cont ain a 7.8-kb salmonella plasmid virulence ( spv) region, which contains five genes ( spvRABCD ) that are highly conserved across the serovars that possess them.10;20;27 Those serovars tend to be natural host-adapted salmonellae, including S. Dublin, S. Choleraesuis, S. Abortusovis, and S. GallinarumPullorum, but have also been found in broad host range serovars such as S. Typhimurium, and S. Enteriditis.10 The genes contained within that small region are sufficient to replace the virulence phenotype of the entire plasmid in animal systemic infection models.27 spvR encodes a transcriptional activ ator of the LysR/Met R fam ily of regulatory proteins and is transcribed independently from the four effector genes (spvABCD) SpvR binds to the spvR and spvA promoters and directs tr anscription of itself and spvABCD during stationary phase growth.28 The full significance of spv genes on bacterial pathogenicity is becoming more clear, and they have been asso ciated with enhanced virulence in mouse systemic infection models,20;29-30;31 as well as showing enhanced expression after invasion of both phagocytic and epithelial cells. 32;33 The spv genes are not necessary for

PAGE 31

5 the bacteria to colonize the mouse gastrointestinal tract or invade mucosal cells to initiate a systemic infection.20 They are also not required to survive in mouse secondary organs such as liver and spleen.34 They have been shown, however to accelerate proliferation of the organism in the re ticuloendothelial system,21 are essential to cause cytopathology in mononuclear cells,35 and are associated with increased mo rtality in the calf model of oral infection.10 A simplistic diagram showing the current opinion of how the spv genes are regulated in salmonella serova rs is shown in Figure 1-1. Figure 1-1. spv gene expression regulation is depe ndent on growth phase and cellular location in Salmonella A significant proportion of salmonella serova rs isolated from clinical cases of human22 and bovine23 diarrhea do not contain virulenc e plasmids and, therefore, the spv genes. It appears that the ability of the spv genes in Salmonella to cause or enhance pathology depends on other bacterial factors (e.g. chromosomal) as well as host factors. R A B C D R A B C D rpoS rpoS Log Phase Stationary Phase or Intracellular mRNA transcripts

PAGE 32

6 The S Enteriditis virulence plasmid containing the spv genes was placed into a plasmidcured S Dublin strain, and virulence was restored in a mouse model,36 while the virulence plasmid from S Dublin pSDL2 only variably tr ansferred a virulent phenotype to serovars that did not commonly carry a virulence plasmid.37 Although the spv genes are present and conserved across several serovars, many different syndromes and outcomes of infection have been clinically or experimentally observed within those subgroups. The work described in this disserta tion attempts to analyze the role of the virulence plasmid and the spv genes in the pathogenesis and epidemiology of equine salmonellosis.10;38 Short-Chain or Volatile Fatty Acids and Salmonella Salmonella spp. are enteroenvironmentally transmitted pathogens of humans and animals. It is therefore exp ected that during their life cycle they are exposed to extremes in temperature, oxygen availability, pH, osmo larity, nutrient availability, organic acid concentration, and presence of other bact ericidal compounds such as reactive oxygen species. Salmonella have shown remarkable ability to sample their environmental conditions and use this information as a si gnal for growth, stasis, or expression of virulence factors. This abil ity is known as “quorum sensing” and is present in several opportunistic and/or pathoge nic species of bacteria.39 It has also been shown in Salmonella that the induction of a stress resist ance response to one condition, (e.g., low pH) confers resistance to multiple stress conditions.40 Ironically, the end-result of carbohydrate feeding to horses (ins tead of a complete forage-based diet) may actually be priming resident or transient salmonella or ganisms and extending their spectrum of resistance to organic acids a nd other stressors. This may indirectly predispose horses to development of salmonella-induced diarrheal disease by seeding their environment with

PAGE 33

7 bacteria that are more virulent than thei r acid-susceptible or otherwise stress-nave cohorts. Short-chain fatty acids are normally found in relatively high concentrations in the forestomachs of ruminants,41 the cecum and large inte stine of all warm-blooded vertebrates,42 and the crop, cecum, and large intestine of birds.43 Acetate, butyrate, and propionate are typically found in the highest percentages, with smaller amounts of isomeric and variable sized carbon-chain compounds.41 In a general se nse, these acidic end products of anaerobic fermentation reacti ons help to keep th e endogenous population of bacteria within th e intestines at a stab le level and discourage transient pathogens from becoming established. They can also be ab sorbed and function as an energy source for the host animal, or they can be di rectly utilized by colonocytes. Antibiotic-Associated Dia rrhea (AAD) in the Horse Diarrhea is one of the most common a nd recognized side e ffects of antibiotic therapy in all species, especi ally the horse. Symptomatically, it can range from mild loss of fecal consistency to projectile liquid feces and/or intestinal pseudomembrane formation. A long-standing hypothesis suggests that disruption of the normal chemical and biological balance within the intestine is responsible for the development of colitis, either during or after the cessation of antibiotic therapy. This relationship may or may not be true in horses. In one case-control study, horses which had rece ived parenteral or oral antibiotics were 40 times more likely to develop diarrhea than horses which had received no therapy.44 Also, in a documented outbreak of salmonella diarrhea in a large hospital, horses that had received parenter al antimicrobial therapy were at 10.9 times greater risk of having Salmonella isolated from their feces th an were matched cohorts not receiving antibiotics.45 However, three longitudinal st udies have demonstrated no clear

PAGE 34

8 association between antibiotic administra tion and salmonella infection in horses.46-48 Any antibiotic, given by any rout e, to any horse, for any length of time, has the potential to cause diarrhea,49 though orally administered agents and those drugs having a biliary component to their metabolic-cycle pose a greater risk.50 Oxytetracycline,51-55 penicillin V and G,56 aminopenicillins,57 metronidazole,56 lincosamides,58-60 trimethoprimpotentiated sulphonamides,61;62 third generation cephalosporins,56 and macrolides63 all have diarrhea as a reported side effect in the horse, though there are conflicting data for specific antibiotics (e.g., trimet hoprim-potentiated sulphonamides).49;51 The situation becomes pivotal in the equine species due to several factors, most importantly 1) the large capacity of the di gestive tract, therefore the potenti al of enormous amplification and dissemination of the infectious agent into the environment, and 2) the intensive management of horse operations—with overly susceptible animals such as neonatal, geriatric, pregnant, and immunocompromised indi viduals often kept in direct contact with asymptomatic animals shedding Salmonella From a therapeutic standpoint the horse also presents more unique challenges. Fi rst, the potentially large volume of fluid excreted per day is difficult to replace—oral and/or parenteral fluid therapy is the cornerstone of therapy in treatment of horses with large colon diseas e. Second, the horse is uniquely susceptible to many secondary comp lications of enterocoli tis that in and of themselves could be as life-threatening as the diarrhea itself. The large bio-burden of gram-negative bacterial cell wall (endotoxi n) contained within the adult equine gastrointestinal tract is more than adequate to cause severe disease or mortality should it gain access to the circulatory system. Thir d, it has been shown clinically as well as experimentally that horses can asymptomatica lly harbor and shed virulent organisms for

PAGE 35

9 unpredictable amounts of time, either followi ng acute infections or without previous illness64;65 and the ability to positively identify a carrier animal based on appearance alone is impossible. The Gastrointestinal Microenvironment The Normal Flora The terms “resistance to colonization” or “competitive exclusion” have been used to describe the passive ability of the gastro intestinal tract to k eep pathogenic organisms from becoming established.66;67 In humans, the anaerobic component of the commensal microflora has been determined to be primarily responsible for maintaining the colonization resistance toward pathogens.68 Despite the multitude of potentially virulent organisms ingested on a continual basis, the innate functions of the intestinal microenvironment restrict a pathogens abil ity to attach, multiply, invade and cause disease. Intestinal anatomy and motility, muco sal epithelial and immune cells, the enteric nervous system, residential bacteria, pr otozoa and their by-products, and mucosal immunoglobulin all combine with digesta to comprise this effective barrier to pathogens.69 The predominant species and demographics of the bacterial popul ation change with respect to the section of th e intestine being colonized. Host diet, oxygen tension, pH, redox potential, and intestinal motility all determine the constitution of the normal intestinal flora, and even this may change on an individual or daily basis. Generally speaking, anaerobic bacteria significantly increa se as a percentage of the total bacteria progressing aborally through th e gastrointestinal tract.70 These anaerobic bacteria are responsible for the breakdow n of otherwise indigestib le saccharide bonds and the production of SCFA and gases such as meth ane and carbon dioxide. Short-chain fatty

PAGE 36

10 acids are also important food sources for th e colonic mucosal cells and are used by the host organism as an energy source.42 Short-Chain Fatty Acids—Production and Intestinal Function Short-chain fatty acids are bacterial byproducts of fermentation reactions that occur in an anaerobic environment. N on-spore forming anaerobes are the principal facilitators of this process thr ough the Embden-Meyerhof-Parnas pathway.71 They have been studied extensively with respect to production sites, rates of appearance, and biological fate in many species.41 SCFAs are important for development and proper function of the rumen, intestine, and mucosal epithelium. The SCFAs, methane, carbon dioxide and hydrogen are the main end-products of anaerobic bacter ial fermentation of carbohydrates, while the branched -chain SCFAs are breakdown products of proteins and are produced independently of the others.72;73 Table 1-1. Short-chain fatty acid ch emical formulas and common names Chemical Formula Common Name CH3-COOH Acetate CH3-CH2-COOH Propionate CH3-(CH2)2-COOH Butyrate CH3-CH-COOH 3CH Isobutyrate CH3-(CH2)3-COOH Valerate CH3-CH-CH2-COOH 3CH Isovalerate CH3-(CH2)2-CO2CH2-CH3 Ethyl butyrate (ethyl butanoate) Herbivores (especially the ruminants) obt ain significant amounts of energy (up to 70-80% of daily maintenance) from the absorption and metabolism of SCFAs, which are produced via bacterial breakdown of dietary li gnin, pectin, cellulose, and hemicellulose.

PAGE 37

11 SCFA production and anaerobic re spiration pathways in the ru minant with substrates and intermediate compounds ar e shown in Figure 1-2, modified from Van Soest.74 Figure 1-2. Polysaccharide metabolism and SCFA production pathways in the rumen. Modified from Van Soest.74 Tan boxes indicate subs trate, red boxes indicate SCFAs, green boxes indicate NVFAs, blue boxes indicate important intermediate compounds, and purpl e boxes indicate accumulated endproducts.

PAGE 38

12 Humans and other monogastric species such as the dog obtain much less energy (69%) from the utilization of endogenously produced SCFAs.42 Additional sources of substrate include sloughed intestinal epithe lial cells, blood, mucins, digestive enzymes, and miscellaneous resistant starches.73 Figure 1-3 depicts the interrelationship between anaerobic microbial function and the products of fermentation. Figure 1-3. Diagram of major gastrointestinal microbial di gestive and en ergy functions, nitrogen and carbon recycling, and SCFA production In addition to local consumption, SCFAs are shuttled directly into the portal circulation for peripheral and hepatic metabolism. Short-chai n fatty acid contributions to maintenance energy requirements of the host range from less than 10% in humans and Amino Acids Microbial Growth Carbon Skeleton Microbial Maintenance

PAGE 39

13 dogs up to more than 80% in the ruminant a nd large-intestine fermenters such as the horse.75 Those SCFAs are utilized by the rumen or ceco-colonic mucosal epithelial cells as an energy source and also influence inte stinal blood flow and water and electrolyte secretion and absorption.73 Short-chain fatty acids are in timately involved in the proper function and regulation of the terminal di gestive processes as shown in Figure 1-4.73;75;76 “Colonic starvation” or “nutritional colitis” are phrases used to describe a diarrhea seen in patients fed either total parenteral (intra venous) nutrition or ente ral tube formulas low in fiber.77 The hypothesis involves decreased SC FA production in the colon, with the colonocytes becoming malnourished, leading to abnormal water and sodium absorption. It was also shown that deranged fermentati on in the large intes tine in response to antibiotic administration did not necessarily predict the deve lopment of diarrhea, though all patients that developed antibioticassociated diarrhea had fermentation abnormalities.78 This suggests that purely the abse nce or impairment of SCFA synthesis is not enough to cause diarrhea but may be an essential predisposing condition.

PAGE 40

14 Figure 1-4. Summary of potential enterotrophi c effects of SCFA Effects of Antimicrobial Therapy: Dysbacteriosis61 The incidence of AAD is estimated to be between 5-25% of all humans receiving antibiotics, though patient risk group, type of antibiotic, a nd route of administration will affect true prevalence.79 Current hypotheses suggest that the gastrointestinal side-effects of antimicrobials are manifested through di sruption of autochthonous anaerobic flora, particularly Bacteroides, Bifidobacte rium, Lactobacillus and Streptococcal spp Anaerobic bacteria are critical for ferm entation of carbohydrates and production of SCFAs, and it is these acids that are believed to have natural and regulatory bactericidal and bacteriostatic properties against en teric commensals as well as pathogens.80 Several investigators have reported significant di sruptions in anaerobic flora and SCFA concentrations in animals, humans, xeno-transplanted flora models, and in vitro colon simulation systems treated with antimicrobials.81-88 Intestinal colonization and increased multiplication rates of S. Typhimurium in response to st reptomycin treatment in mice Energy Source Growth factors Regulate gastrointestinal hormones e.g. insulin and glucagon Autonomic nervous system Vasodilatory, mediate intestinal blood flow Anti-carcinogenic ( in vitro ) Regulate mucosal sodium and water absorption

PAGE 41

15 were associated with decreased concentra tions of fecal SCFAs and increased luminal pH.89 In vitro supplementation of SCFA to the cecal contents of treated animals inhibited salmonella growth in this model.80 Further studies have show n that this protection may be conferred by specific SCFAs as elevated concentrations of propionic or formic acid added to feed conferred si gnificant protection against S. Typhimurium cecal colonization in chicken hatchlings.90 Another theory links the etiology of AAD to the reduction or disappearance of SCFAs. These acids are regulators of sodi um and water uptake in the colon, and their absence causes an indirect accumulation of sodium and water in the intestinal lumen.91 Sodium is a potent cellular osmolyte which draws more water across membranes and into the lumen, causing increases in fecal water co ntent. This theory does not adequately account for the magnitude of diarrhea seen in some AAD patients, but it could easily be an initiator or contributor to pathogenesis. Alternative popular assumpti ons of the pathogenesis of AAD include unchecked overgrowth of Clostridium difficile (especially in human ne onates) with production of potent enteroand cyto-toxins or the vacati ng of attachment sites or toxin receptors normally occupied by host commensal bacteria.92 C. difficile has been identified as a pathogen in equine AAD.93 Specific Aims The specific aims of the reported studies were to: Collect Salmonella spp. isolates from clinical ca ses of equine salmonellosis and from normal horses. Examine case history and collect relevant host data (age, breed, gender, presenting disease, risk factors, biochemical profiles, antimicrobial susceptibilities, treatments) for all salmonella isolates.

PAGE 42

16 Determine if the salmonella isolates carried large plasmids. Determine if the salmonella isolates carried spv genes. Determine the cecal SCFA concentrations luminal pH, total culturable anaerobic bacterial counts, and protozoal counts of horses before and after treatment with selected antibiotics. To examine spv + and spv salmonella isolates in te rms of growth rate during anaerobic culture in nutrient broth supplemented with sterile-filtered cecal contents from antibiotic-treated ve rsus non-treated horses. To examine spv + and spv salmonella isolates in te rms of growth rate during anaerobic culture in nutrient broth adjust ed to the mean luminal cecal pH and supplemented with individual SCFAs nor mally found in horse cecal liquor. To examine plasmid containing spv isolates for antibiotic resistance determinants located on the plasmids. Hypotheses Large plasmids in salmonella isolates are directly correlated with presence and type of disease. o Isolates from normal horses will not have plasmids. o Isolates from cases of diarrhea will variably contain plasmids. o Isolates from systemic cases will always contain plasmids. Salmonella isolates with large plasmids will also contain spv genes on those plasmids. The administration of repeat ed doses of commonly used antimicrobial agents to healthy horses will reduce the total cultu rable anaerobic bacterial population of the cecum, resulting in a reduction in the c oncentration or disruption of relative proportion of SCFAs and an increase in luminal pH. The administration of antimicrobial ag ents to healthy horses will reduce the numbers of cecal protozoa. Sterile-filtered cecal contents from horses that were not treated with antibiotics will inhibit the growth of Salmonella compared to sterile-filt ered cecal contents from animals that received antibiotics in a spv -dependant manner. Nutrient broth containing individu al SCFA will inhibit growth of Salmonella under anaerobic conditions in a dose-dependant and spv -dependant manner.

PAGE 43

17 Large plasmids in salmonella isolates that do not contain spv genes are likely antibiotic resistance plasmids.

PAGE 44

18 CHAPTER 2 SALMONELLA IN HORSES DISEASE DEFINITION AND GENERAL AND MOLECULAR EPIDEMIOLOGY Background Disease Overview In spite of pharmacological and therapeutic advances, diarrhea in the adult horse continues to be one of the most challenging and frustrating medical syndromes facing the equine veterinarian. Salmonella spp. are one of the primary etiological agents of equine diarrhea, although a large number of diarrhea cases will progress or resolve without a definitive diagnosis. Salmonella infection of hor ses is not limited to the intestinal tract. There is potential for bacteremia, particul arly in foals, with seeding of synovial structures, bone, lung, umbilical remnants, brain and meninges, liver, and kidneys. Salmonellosis can quickly become a financial disaster for the intensively managed horse farm or equine hospital given the potentia l copious nature of contaminating feces produced by one diarrheic adult horse, along with the environmental persistence of the organism. There are also serious human h ealth issues regardin g the zoonotic potential from treating and handling these animals. There are four recognized clinical syndrom es of salmonella infection in horses: 1) an asymptomatic carrier or latent state;64 2) a severe and sometimes fatal fibrinonecrotic enterotyphlocolitis; 3) bacteremia—with or without seconda ry foci of infection; and 4) pyrexia, depression, and leukopeni a without diarrhea—similar to the “enteric fever” syndrome seen in humans infected with S. typhi.94

PAGE 45

19 Prevalence Excretion of Salmonella into the environment commonl y occurs in horses without signs of enteric disease. This may be an animal that has recently recovered from infection, an animal that has acquired bacter ial organisms via ingestion of contaminated feed, water, or bedding material and is si mply a transient portal, or a chronically colonized host that has adapted a traditiona lly pathogenic relationship into a commensal one. It is the apparently h ealthy, but chronically colonized animal that represents the greatest danger to the popul ation. Estimates vary widely depending on the population sampled and the diagnostic methodology used of the percentage of the horse population that is shedding Salmonella The recent National Animal Health Monitoring System (NAHMS) survey reported that 0.8% of resi dent horses sampled in the US excreted Salmonella in their feces.95 The majority of horses in this survey had normal fecal appearance at the time of sampling, although 2.1% had loose or watery feces. The prevalence of salmonella shedding was not higher in animals that had received antibiotics within the past 30 days. Reported Risk Factors for Salmonella Infection It has been noted that horses have an increased risk of developing salmonellainduced diarrhea after certain “stressors” ha ve been placed on them, including but not limited to transportation,96 hospitalization,47 nutritional excess or deficiency,54 dietary change,97 colic—especially la rge colon impaction,47;98 nasogastric intubation,44;45 debilitating injury or illn ess, antibiotic therapy,44;45;48 parturition, weaning, surgery, anesthesia, or anthelmintic therapy.99;100 The challenge inoculum for these “at risk” individuals can be up to 100fold smaller than for non-st ressed and immunocompetent cohorts.94;101;102 It is for these reasons that horses admitted to veterinary hospitals, even

PAGE 46

20 on an outpatient basis, are highly susceptible to infection. The populations at greatest risk are those horses with gastrointestinal diseases admitted to referral hospitals for medical or surgical therapy.44;45;101 Salmonella Serovars Associat ed with Equine Infection Approximately 60% of known salm onella serovars belong to the S. enterica subsp. Enterica group and within this group the O-antigen designations A, B, C1, C2, D, & E account for 99% of all warm-blooded animal in fections. All O-antigen groups have been isolated from horses, but groups B, D, and E are the most common.101 Commonly, phenotypic and molecular analyses are married to form the most accurate picture of an isolate as possible. Analysis of antimicr obial susceptibility, se rogroup, serovar, phage type, plasmid profile, ribotype, or restric tion endonuclease examination allows more specific identification of salmonella organisms. Newer and more precise methods of distinguishing salmonellae include polymera se chain reaction (PCR) fingerprinting, multiplex PCR, pulsed-field gel electrophoresis (PFGE),103 restriction fragment length polymorphism (RFLP), IS200 typing,104 and real-time PCR.105 PCR has demonstrated itself to be one of the most sensitive and expedient methods of detecting Salmonella spp. in equine fecal samples, though culture is still the most cost-e ffective and widely available.106 This is most helpful from an epid emiologic and control standpoint or for biologic surveillance programs. Though the tr eatment does not vary between serovars, specific identification could help in cases of outbreak, treatment failure, or when more than one strain of Salmonella is suspected. Salmonella serovars frequently reported is olated from horses over the last 40 years include Agona, Anatum, Arizonae, Enterica, Enteriditis, Heidelber g, Infantis, Krefeld, London, Miami, Muenchen, Muenster, Newport, Oraneienburg, Rubislaw, Saintpaul,

PAGE 47

21 Senftenberg, Thompson, Typhimurium, and Typhimurium var. Copenhagen.46;47;95;97;99;107-111 Almost all serovars of Salmonella infecting horses are non-host adapted strains,101 with the exception of Abortu sequi, which does not cause gastrointestinal disease, but ra ther early abortion in mares and systemic sepsis in newborn foals.112 Horses are also susceptible to some of the normally host-adapted serovars of other species such as S. Dublin (bovine) and S Choleraesuis (porcine).110 The herbivorous and gregarious nature of horses makes them efficient dispersal agents as well as susceptible recipients for the entero-environmental cycling of Salmonella Compounding this issue, salmonellae are ubiquitous and environmentally resistant and can remain infectious in f ecal material for years under the appropriate conditions.113 Role of Microbial Virulence Factor s in Equine Salmonella Infection Specific virulence factors that mediate sy stemic or gastrointestinal salmonella infections in horses have not b een extensively studied. Likely this is due to reluctance or difficulty in using the horse as a model of disease. Retrospective studies examining isolates obtained from clinical cases of salmonellosis have been published, but investigators focused on more epidemiological than molecular techniques of comparison. Disease Prevention Diet, Probiotics, Immunity Methods utilized by veterinarians to d ecrease the morbidity and mortality of salmonella infection in horses have either lim ited scientific basis or are applied based on results obtained from other species. Very little information is available on specific preventative strategies or therapies once clinical signs become evident. Fructo-oligosaccharides (FOS) have been uti lized extensively as feed additives in the poultry and companion animal industries fo r many years. They exert their effects by

PAGE 48

22 increasing the amount of fermentable carbohydrate that reaches the la rge intestine, which can be acted upon by the bacterial population. Th is in turn raises th e concentrations of organic acids and drops the pH, which pr esents an inhospitable environment to pathogenic species.114 “Direct-fed microbial” and “competitive exclus ion” are terms used frequently in the poultry industry to describe a practice and physiol ogic phenomenon of directly feeding or facilitating the establishment of a desirable microbial population in order to discourage colonization by an undesirable one, typically Salmonella Transfaunation via fecal slurry or cecal or colonic contents from a recen tly euthanatized or cannulated horse are techniques used in a hospital situation to re-establish comm ensal protozoa and bacterial flora in horses with diarrhea. Enemas of sl urried fecal material from normal individuals, have been shown quite effective at treating or preventing antibi otic associated diarrhea in humans, but are unlikely to be beneficial in horses due to anatomical differences.115 A commercial probiotic preparation is available for use in horses (Probios Equine One Gel, Chr. Hansen BioSystems), however clinic al efficacy data of this type of product in horses is limited. In a prospective study of hos pitalized horses neither of two commercial probiotic formulations had any effect on salm onella shedding, incidence of diarrhea, or length of hospitalization fo llowing abdominal surgery.98 A recent prospective study examining the probiotic potential of Lactobacillus rhamnosus strain GG in horses failed to show efficient colonization of the adult gastrointestinal tract unless extremely large doses were administered, though foals were more consistently and efficiently colonized.116 These conflicting results should be fu rther investigated, as human evidence is strongly in favor of the us e of direct-fed microbials in the prevention and management

PAGE 49

23 of antibiotic-associated diarrhea or other diarrheas attributed to dysbacteriosis. Significant benefit could be obtained from a small daily dose of orally administered bacteria during periods of increased suscep tibility to salmonellosis, such as during extended travel or preceding and concu rrent with antibiotic administration. Immunity to Salmonella is dependent on a combination of cell-mediated recognition and destruction by activated granulocytes, as well as an antibody driven humoral response. Salmonella antibody-containing equine plasma products are commercially available. These products are al most exclusively used for the treatment of systemic salmonellosis in foals, or as preventa tive therapy in foals with failure of passive transfer in areas with a histor y or high prevalence of diseas e. These products are usually cost prohibitive for use in adult horses, and more importantly, are serovar specific, thus providing no cross protection to the significant number of othe r serovars able to infect horses. Mucosal immunization of horses with mutant strains of Salmonella rendered non-pathogenic has also been examined. Sheo ran et al. demonstrat ed strong production of S. Typhimurium specific mucosal IgA in je junal, nasal, and vaginal compartments after intra-nasal vaccina tion of ponies with a cya crp-pabA mutant of S Typhimurium.117 This strain is attenuated fo r virulence by deletion of the genes necessary for adenylate cyclase production ( cya ) and the cyclic AMP receptor protein ( crp ). This live vaccine did not cause any signs of disease, was not shed in the feces, nor was it transferred to cohabitated non-vaccin ates. Mucosal specific antibody is an attractive first line of defense against en teric pathogens, and exploitation of the gastrointestinal mucosal immune system in th e horse is attractive in terms of prevention and protection.

PAGE 50

24 Disease Treatment The treatment of salmonella infection is controversial and dependent on several factors, including severity of disease, immune status, metabolic state, age, concurrent malignancy, drug cost, drug availability, side-eff ects, and the presence of colonizable foci (e.g., implanted materials, catheters). Conven tional antibiotic therapy of uncomplicated salmonella gastroenteritis in human beings is often not efficacious and may actually prolong the convalescent phase and/or extend the length of time that Salmonella is shed from the feces.118-120 Even antibiotics preferred for the directed therapy of Salmonella in horses and humans (e.g., fluoroqui nolones) have not had any scientifically reproducible or predictive effects on fecal carriage postinfection. Post-convalescent shedding is an important salmonella-related morbidity i ssue facing the equine practitioner. Contamination of the environment with pers istent, virulent, and potentially antibiotic resistant bacteria is a cause for concern in a horse facility, espe cially a veterinary hospital. Outbreaks of nosocomial salmonello sis have resulted in institutional shutdowns world-wide.109;121-128 In these circumstances antibiotic therapy of clinically silent or uncomplicated cases would be useful if the period of environmental contamination could possibly be shortened, thereby limiting expo sure of other animals while the facility is depopulated and disinfected.129 The Salmonella Virulence Plasmid Clinical isolate data from several human studies agrees with the murine model of salmonellosis: virulence plasmids are more likel y to be present in t hose isolates obtained from systemic sources such as blood, liver, a nd spleen, compared to unrelated isolates obtained from feces.22;23 Conflicting reports utilizing co mparable experimental methods have shown no causal relationship between b acteremia and presence of the virulence

PAGE 51

25 plasmid in humans.24;25 Discordant results are also seen within the model using calves orally infected with S. Dublin. The virulence plasmid containing the spv genes was shown to be important in determinin g the severity of diarrhea in calves,10 while other investigators demonstrated no influence of the virulence plasmid (as compared to naturally occurring plasmid-free or plasmi d-cured isolates) on pathogenesis either in vivo or in ex vivo ligated ileal loop experiments.26 This serovar-host-syndrome interrelationship is most certainly a confounding factor in determining the pathophysiologic importance of th e salmonella virulence plasmid. Salmonella Plasmid Virulence ( spv ) Genes Function of the spv genes The function of the spv genes in Salmonella has been a focus of investigation for many years. Highly conserved genomic elements should theoretically be important to the survival and host-to-host transmission of pa thogenic bacterial species. Of the entire spv locus, it has been shown that only spvB and spvC are essential for full virulence in the mouse model of subcutaneous infection,31 and more recently, that spvB was required for cytotoxic pathology (progressive detachment of adherent cells, vacuolization) and apoptosis after phagocytosis by hum an monocyte-derived macrophages.35 The apparent accelerated growth of spv positive strains (as compared to spv negative) and their ability to cause systemic disease may actually be an extension of their ability to survive and travel within macrophages to th ese sites. A summary of the current understanding of the molecular and functional in formation regarding the spv genes can be found in Table 2-1.

PAGE 52

26 Table 2-1. spv gene characteristics Gene Activity Protein Localization Significance spvR Transcriptional activator of spvR and spvABCD Cytoplasm Positive regulator (promoter) of itself and the other spv genes spvA unknown Outer membrane130 Unknown, mutations do not affect virulence in mouse intraperitoneal infection model. spvB ADPribosyltransferase Cytoplasmic and transported out of cytoplasm, small amounts in inner membrane130 Essential for full virulence (mouse model).31 Effector protein causing depolymerization of actin cytoskeleton within macrophages. Inhibition of phagolysosome fusion. spvC unknown Cytoplasm Essential for full virulence (mouse model)31 spvD unknown Exported outside of cell Mutations attenuate virulence (mouse model)131 Significance of the spv genes The role of spv genes in equine salmonella infect ion has not been investigated. There is conflicting evidence, as demonstr ated in the mouse and calf models, that spv genes play a primary role in the establishmen t and persistence of sy stemic infections and do not contribute significantly to the enteric phase of the disease. Anaerobiasis was shown to significantly retard the growth rate of Salmonella with a significantly reduced cell density at stationary phase, and the spv genes were not expressed.132 This lends further support to th e hypothesis that the spv genes are not involved in the enteric phase of infection, but this has not been examined in species other than the calf and mouse. Specific Aims The overall aim of this section of th e study was to describe the general and molecular characteristics of Salmonella spp. isolated from ho spitalized symptomatic

PAGE 53

27 animals in North Central Florida and contrast these isolates from those collected from asymptomatic animals in the same geographic region. The specific aims were: To collect, describe, and store Salmonella spp. isolates from hospitalized horses. To collect, describe, and store Salmonella spp. isolates from asymptomatic horses at pasture. To determine if the salmonella isolates carried plasmids and classify them based on size. To determine if the salmonella isolates carried spv genes. To examine plasmid containing spv negative isolates for antibiotic resistance determinants located on those plasmids. Materials and Methods Case Selection Bacterial cultures were obtained from hospitalized foals and adult horses with clinical signs consistent with salmonella in fection. Material submitted to the clinical microbiology laboratory included feces, gast ric secretions, blood, synovial fluid, and tissue samples from post-mortem examinations. Sequential fecal samples were also collected from asymptomatic horses at severa l farms in the North Central Florida area, over a period of 24 months. Individual re cords were kept for each animal and horses were sampled at least three times on separate occasions. Microbiological Techniques Field samples Freshly voided or rectal fecal samples we re collected and placed into sterile, labeled containers. Two to five grams of fecal material was placed into selenite broth and incubated at 37C in a 5% CO2 environment for 12-18 h to maximize isolation of Salmonella spp. A Hektoen-Enteric plate (Remel Inc., Lenexa, KS) was streaked for isolation from the overnight culture broth. Th e plates were incubated 18-24 h at 37C in a 5% CO2 environment. Non-lactose-fermenting and H2S-producing colonies were

PAGE 54

28 selected and streaked onto urease slants (Rem el Inc., Lenexa, KS) which were incubated 18-24 h at 37C in a 5% CO2 environment. Urease-negative organisms were further characterized utilizing API 20E enteric test strips (bioMrieux USA, Durham, NC) for positive identification of Salmonella spp. An incubated strip with reactions typical of Salmonella is shown in Figure 2-1. Tests and inte rpretations from left to right include ONPG -, ADH +, LDC +, ODC +, CIT +, H2S +, URE -, TDA -, IND -, VP -, GEL -, GLU +, MAN +, INO -, SOR +, RHA +, SAC -, MEL +, AMY -, ARA +. Figure 2-1. API20E rapid identification st rip showing typical reaction results for Salmonella spp. Clinical and reference isolates All clinical isolates were provided as pur e cultures on Hektoen-Enteric (HE) agar plates (Figure 2-2) by the clinical microbi ology service at the Un iversity of Florida College of Veterinary Medici ne, Gainesville, Florida. Figure 2-2. Salmonella group C2 isolate as provided on Hektoen-Enteric agar

PAGE 55

29 Bacterial cultures of all salmonella refere nce strains and clinical isolates were subsequently grown in Luria-Bertani broth (L B) or on LB agar w ithout antibiotics at 35 C, in a 5% CO2 atmosphere unless otherwise indicat ed. One and a half percent (w/v) agar was added to LB broth for plates. Co mposition of culture media is in Table 2-2. Table 2-2. Composition of bacterial culture media Media Ingredients per Liter and/or Supplier with Catalog Number Sterilization Storage LB Broth 10 g tryptone 5 g yeast extract 5 g NaCl Sigma-Aldrich L3152 Autoclave 15 min at 121C RT LB Agar 10 g tryptone 5 g yeast extract 5 g NaCl 15 g agar Autoclave 15 min at 121C 2-6C LB-N Broth 10 g tryptone 5 g yeast extract 8.5 ml 5 M NaCl (0.85% w/v) Autoclave 15 min at 121C RT Hektoen-Enteric Agar Sigma-Aldrich H7532 Autoclave 15 min at 121C 2-6C Minimal Medium (M9) 200 ml 5x M9 Salts 20 ml 1 M glucose 2 ml 1 M MgSO4 0.1 ml 1 M CaCl2 Autoclave 15 min at 121C *before* addition of filter sterilized glucose RT or 26C Brain-Heart Infusion (BHI) Broth Sigma-Aldrich B7403 Autoclave 15 min at 121C RT Selenite Broth w/ Cystine Remel 064506 Pre-sterilized commercial product 2-6C Urea Agar Remel 065210 Pre-sterilized commercial product 2-6C PRAS Brucella bloodagar with 75 micrograms per ml gentamicin Anaerobe Systems AS-141G Pre-sterilized, custom manufactured, commercial product RT

PAGE 56

30 Antibiotic supplementation to LB agar plat es, when referenced, was made in the following concentrations: ampi cillin (AMP) 100 micrograms/ml; nalidixic acid (NAL) 50 micrograms/ml; ceftiofur sodium (NAX) 8 micrograms/ml; cefazolin sodium salt (CEF) 8-32 micrograms/ml; and tetracycline hydr ochloride (TET) 25 micrograms/ml. Salmonella Identification and Anti biotic Resistance Profile Isolates identified as Salmonella were grouped using group specific antisera (Fisher Scientific International, Hampton, NH) and serotyped through National Veterinary Service Laboratories (NVSL), Ames, IA. Those isolates positively identified as Salmonella were sub-cultured and frozen. Antibiotic resistance profiles were determined for each isolate via an automated minimum inhibitory concentra tion (MIC) system (Sensititre Microbiology Systems, software version—SAMS V2.3 Release 1, Tr ek Diagnostics, Cleveland, OH, USA). Salmonella Isolate Storage Subcultures of salmonella strains were stored as pure cultures at -80C in LB or Brain-Heart Infusion (BHI) and 35% (v/v) gl ycerol. A standing overnight culture was prepared by selecting approximately ten to fi fteen colonies from th e plate provided by the microbiology laboratory. Multiple colonies we re sampled to avoid the selection of any single genotype or abnormal colony. The isolate was inoculated into LB and incubated at 37 C without agitation. The following morning, 1 ml of this culture was added to 30 ml of either BHI or LB and incubated at 37 C with agitation for 1 to 1.5 hours (approximate OD600 = 0.5-0.6). The sample was centrifuged at 10,000 x g for 10 min to pellet the cells. The supernatant was removed, and the cells were re-suspende d into 2 ml BHI or LB. Two ml of 70% glycerol was added to the cell suspensi on and mixed gently. The

PAGE 57

31 isolates were transferred immediately to pr e-labeled standard cryoge nic storage vials and flash frozen in a dry ice and ethanol bath. Reference Strains Reference Salmonella serovar Typhimurium strains 3306 and 3337 were kindly provided by Dr. Paul Gulig in the Department of Molecular Genetics and Microbiology at the University of Florida College of Medi cine. Specific information regarding these strains is detailed in Table 2-3. The sequenced size of the S Typhimurium strain LT2 virulence plasmid pSLT is 93,939 bp,133 but the virulence plasmid sizes of similar strains such as SR-11 may vary. The reported size fo r the virulence plasmid of the strain used herein is approximately 100,000 bp. Table 2-3. Salmonella serovar Typhimurium reference strains used in this study Strain Genotype Source Phenotype SR-11 3306 gyrA1816 pStSR100+ Dr. Paul Gulig Nalr, virulent, spv+ SR-11 3337 gyrA1816 pStSR100Dr. Paul Gulig Nalr, spv-, avirulent, plasmid cured derivative of 3306 Plasmid Profiling of Salmonella Isolates Plasmid extraction was achieved using a m odification of a commercial kit for large construct and very low copy number plasmi d purification (Qiagen Filter Midi Kit, Qiagen, Inc., Valencia, CA). The S. Typhimurium plasmid copy number has been estimated to be between 2-3 per cell.133 Bacteria were grown in 50 ml LB for 12-16 h (approximately A600 = 1 – 1.5). Cells were divided into two sterile 50 ml polypropylene centrifuge tubes and pelleted by centrifugation at 7,000 rpm in JA-20 rotor for 15 min at 4 C. The cells were re-suspended th oroughly by vortexing in 10 ml buffer P1Resuspension Buffer per tube. Ten ml of buf fer P2-Lysis Buffer was added to each tube, the cells were mixed by gentle rolling and inversion and incubated at RT for 5 min

PAGE 58

32 exactly. Ten ml of chilled buffer P3-Neutra lization Buffer was added per tube and the samples were mixed immediately but gently by inversion. The tubes were incubated for 15 min at RT. Columns (Qiagen-tip 100, Qiagen Inc., Valencia, CA) were equilibrated with 4 ml of buffer QBT and columns were allowed to empty by gravity flow during this incubation to be ready when needed. Sample s were poured into pre-labeled high-speed centrifuge tubes (Oak Ridge Cent rifuge Tubes, Fisher Scientific International, Hampton, NH) and centrifuged at 15,000 rpm in a B eckman JA-20 rotor, for 10 min at 4 C. The supernatant was removed and applied to a vertically supported filtration syringe (QIAfilter cartridge, Qiagen Inc., Valencia, CA), and the plunger was inserted. The filtrate was dispensed onto the columns gently and slowly, over a period of approximately 10-20 min, keeping visible sample in the reservoir of the column at all times. The column was then washed with 2 volumes of 10 ml buffer QC-Wash Buffer at RT and allowed to empty by gravity flow. Th e wash solutions were discarded and clean 40 ml high-speed centrifuge tubes were placed under the columns to collect the eluted DNA. Plasmid DNA was el uted with 5 ml 56 C buffer QF-Elution Buffer per column. The plasmid DNA was precipitated with 0.7 volumes (4 ml per isolate) of RT isopropanol. The samples were centrifuged at 16,000 rpm in a JA-20 rotor for 30 min at 4 C and the supernatant was removed. The DN A pellet was washed with 70% ethanol, dried, and re-suspended in 150 microliters of TE for agarose gel analysis and transformation experiments. A summary of the solution ingredients for the modified plasmid extraction procedures and storage is found in Table 2-4. Agarose gel analysis was performed via common method. Equal volumes of the plasmid extract and 10x sample loading buffe r were combined as a droplet on paraffin

PAGE 59

33 paper, loaded, and run on a 0.5% agarose in Tris-borate-EDTA (TBE ) gel at 125 volts for 1.5 h. Size was estimated by comparison to the approx. 100-kb plasmid from S Typhimurium 3306 run on the same gel in addition to a super coiled DNA ladder with a range from 16.2 to 2-kb pairs (Gibco BRL, Carlsbad, CA). Plasmid bands were visualized by staining the gel with ethidi um bromide (1 microgram/ml) and photographed using a digital gel imaging and documenting system (Chemi System, UVP BioImaging Systems, Upland, CA). If an isolate had at least one large plas mid it was considered plasmid-positive. A plasmid was only considered to be a virulence plasmid if the spv gene primer sets hybridized to the isolate. Otherwise, it was simply a large plasmid of unknown type. Table 2-4. Composition of buffers and soluti ons used in plasmid extraction protocols Reagent Composition Storage P1 (Resuspension Buffer) 50mM Tris-HCl pH 8.0 10mM EDTA 100 micrograms/ml RNase A 2-8C P2 (Lysis Buffer) 200mM NaOH 1% SDS (w/v) RT P3 (Neutralization Buffer) 3.0M potas sium acetate pH 5.5 RT or 2-8C QBT (Equilibration Buffer) 750mM NaCl 50mM MOPS pH 7.0 15% isopropanol (v/v) 0.15% Triton X-100 (v/v) RT QC (Wash Buffer) 1.0M NaCl 50mM MOPS pH 7.0 15% isopropanol (v/v) RT QF (Elution Buffer) 1.25M NaCl 50mM Tris-HCl pH 8.5 15% isopropanol (v/v) RT Tris-EDTA (TE) 10mM Tris-HCl pH 8.0 1mM EDTA RT 10x Sample Loading Buffer 40% Sucrose 0.17% Xylene Cyanol 0.17% Bromophenol Blue RT

PAGE 60

34 Polymerase Chain Reaction (PCR) Identification of spv Genes PCR was performed to evaluate all clinical isolates for the presence of the spv genes. Positive ( 3306) and negative ( 3337) isolates, as well as a series for a chromosomal gene, aspartate semialdehyde dehydrogenase ( asd ), were run as controls in each experiment. These controls were vita l for two reasons: 1) to verify that the reactions, reagents, and conditions were appropriate, and 2) to ensure that the isolates were truly Salmonella spp. (which was most important in validating negativ e reactions). A loop of pure culture was added to 200 mi croliters of sterile water in a 1.5 ml microcentrifuge tube, and boiled for 10 min to be used as template DNA. The master mix and all reactions were prepared and maintained on ice until the run was started. Master mix was made fresh for each experime ntal run, and consisted of 24.75 microliters deionized H20, 5 microliters 10x PCR buffer, 8 microliters 1.25mM dNTP mix, 0.25 microliters 5U/microliter Taq DNA polymerase, and 4 microliters 50mM MgCl2 (GibcoBRL, Carlsbad, CA). Five microliters of template DNA was added to each tube for a total reaction volume of 50 microliters per tube. Th e samples were placed in a thermocycler (Programmable Thermal C ontroller PTC-100, MJ Research, Inc. Reno, NV) for the cycle described in Table 2-5. Table 2-5. Times and temper atures for PCR reactions STEP 1 Melt 94C 180 seconds STEP 2 Melt 94C 60 seconds STEP 3 Anneal 45C or 50C 60 seconds STEP 4 Extend 72C 120 seconds REPEAT STEPS 2, 3, & 4 FOR 30 CYCLES STEP 5 End 72C 180 seconds STEP 6 Hold 4C indefinitely Primer sets for the spv gene PCR were provided by Dr. Paul Gulig. These consisted of 3’ and 5’ primers for asd spvA spvC and spvR All clinical isolates were

PAGE 61

35 examined for presence of the asd gene of Salmonella and probed with at least two different spv gene primer sets. An isolate wa s determined to be positive for the spv gene locus if two conditions were met: 1) if the asd product was present, and 2) if two or more of the spv gene products were present. These primer sets are situated to bracket the entire open reading frame of the gene, so th e PCR products are in essence whole spv genes. All isolates were probed with no less than two spv gene primer sets each (usually spvA and spvC occasionally including spvR ). These primers were ex tremely effective in their ability to identify the genus Salmonella and the presence of spv genes. Sequence and other important primer informa tion are contained in Table 2-6. Table 2-6. Primers utilized in PCR reactions Primer Sequence Product Size spvA 5’ 5’-CCCCCGGGATGAATAT GAATCAGACCACCA-3’ --spvA 3’ 5’-GGGAATTCTGGTAGCGCGGGAAGC-3’ 784 bp asd 5’ 5’-CAGCACATCTCTT AGCAGGAAAAAAACGC-3’ --asd 3’ 5’-GGGAAGCTTCTACGCCAACTGGCGCA-3’ 1,100 bp spvR 5’ 5’-CCCCGGGATCCATGGATTT CTTGATTAATAAA-3’ --spvR 3’ 5’-CCCCGGGAATTCGC TGCATAAGGTCAGAAGG-3’ 905 bp spvC 5’ 5’-CCCCCGGGATGCCCA TAAATAGGCCTAATC-3’ --spvC 3’ 5’-GCCGGAATTCGTCAGTAAGGG-3’ 875 bp Salmonella Plasmid Transformations into Susceptible Bacteria—Effects on Antibiotic Resistance Based on a discovery that the minority of la rge plasmids in the clinical salmonella isolates were virulence plasmids, transformatio ns of extracted plasmid DNA into a select antibiotic-sensitive strain of Escherichia coli were performed to investigate the possibility that the large plasmids may be car rying antibiotic resistance (R) determinants. Three clinical salmonella isolates were selected based on their antimicrobial sensitivity profiles and the pr esence of a single large plasmi d that did not contain the spv genes. Successful transference was c onfirmed through plasmid extraction of the

PAGE 62

36 transformed E. coli isolates and gel electropho resis with untransformed E. coli as well as the original plasmid extracts used for the transformation. Briefly, the procedure was performed as follows: competent E. coli DH5 cells (Invitrogen Corporation, Carlsbad, CA) were thawed on ice. A 40 microliter aliquot of E. coli was added to an ice-cold electroporation cuvette. All solutions were maintained on ice throughout the procedure unless otherwise specified. Two microliters of plasmid extract in TE was added to the E. coli and mixed gently with a pipette. Th e mixture was electroporated at 1.25 kV, 25 microfarad capacitance 200 ohms resistan ce, on a Bio-Rad Gene-Pulser (Bio-Rad, Hercules, CA). The time constant was as cl ose to 4.9 as possible, and if below 4.5, the procedure was repeated with less plasmid ex tract. Nine-hundred microliters LBN broth was added to the cuvette and it was incubate d in a water bath at 37C for 30 minutes. The cell suspension was transferred to a 1.5 ml microcentrifuge tube and 100 microliters was spread onto several different LB agar plates, each containing a relevant selective antibiotic based on the antimicrobial resistance profile of the original salmonella isolate. The plates were incubated overnight at 37 C and observed for growth the next day. Transformants (as evidenced by growth on selec tive plates) were grown in LB broth with continued selective pressure and subjected to the same extraction procedure described previously to verify the presence an d size of newly acquired plasmid DNA. Chemical transformations were performe d by mixing 5 microliters of extracted donor plasmid DNA in water with 40 micr oliters of recipient strain (same DH5 E. coli as for electroporation transformations). Th e mixture was incubated on ice for 15 min and then transferred to 42 C water bath for 2 min to heat-shock the cells. Five-hundred microliters of LBN broth was added, and the mixture was incubated at 37 C without

PAGE 63

37 shaking for 30 min. One-hundred microliter s of the mixture was then plated on appropriate selective plates as descri bed for the electroporation transformation. Statistical Methods Clinical isolates, patient information, and resistance data were collected weekly as they became available over a period of three years (December 1999 through September 2002). Survival was based on discharge from the hospital. Animals were initially grouped with respect to age into the followi ng categories: <0.5y; 0.5-4y; 5-8y; 9-12y; 1315y; and >15y. They were then further groupe d as follows for sta tistical comparison: 05y; 6-15y; and >15y. Cases with missing da ta were excluded from calculation of descriptive percentages (e.g., su rvival, gender, age). The e ffects of gender, age, and breed on outcome were investigated independe ntly using the Pears on Chi-square test. Stepwise logistic regression was used to form a model in which multiple clinical variables could be used to predict outco me. A Kruskal-Wallis test was used to investigate the percentage of submitted sa mples that were positive with respect to outcome. Results Asymptomatic Population Isolates were sought from asymptomatic animals in order to contrast bacterial genotype with isolates from clinical cases. Salmonella spp. were not identified from any of 381 cultures performed on 105 different as ymptomatic horses over a period of two years. Clinical Cases There were 106 hospitalized animals during th e period of interest that had at least one positive culture of Salmonella Within this population there were more males (61%)

PAGE 64

38 than females (39%), although this difference was not significant. The mean age of the population was 4.9 years, with a range of 2 da ys to 38 years. The age distribution by categorical group is shown in Figure 2-3. Fifty seven perc ent of the clinical cases survived. Of those non-survivors 26% died spontaneously. Breed distribution for the affected horses is shown in Table 2-7. Table 2-7. Breed distribution of 84 equine salmonella cases 1999-2002 Breed No. of Cases % of Cases Thoroughbred 32 38.09 Quarter Horse 15 17.86 Paso Fino 10 11.90 Miniature Horse 5 5.95 Arabian 4 4.76 Paint Horse 4 4.76 Warmblood 4 4.76 Standardbred 3 3.57 Pony 3 3.57 Mixed Breed 2 2.38 Draft 1 1.19 Appaloosa 1 1.19 Relationship Between Gender or Age and Outcome There was no gender bias with respect to short-term survival (Table 2-8). There was a statistically significant association be tween age and case outcome (Table 2-9). Horses less than 5 years of age were 3.3 times more likely to die when infected with Salmonella than older animals. Table 2-8. Effect of gender on mortality in 96 cases of equine salmonellosis* Gender Died Survived % Dead Odds RatioLower 95%CI Upper 95%CI Female 13 19 40.6 1.25 0.54 2.88 Male 22 30 42.3 1.00 0.47 2.13 *Pearson Chi-square value of 0.277 with 1 degree of freedom, p=0.599

PAGE 65

39 0 5 10 15 20 25 30 35Number of Cases <6mo6m-4y5y-8y9y-12y13y-15y>15yAge Figure 2-3. Age distribution of 98 equine salmonella cases 1999-2002 Table 2-9. Effect of age on mortality in 85 cases of equine salmonellosis* Age Group Died Lived % Dead Odds Ratio Lower 95%CI Upper 95%CI 0 – 4 yrs 28 28 50.00% 3.33 1.18 9.42 5 – 15 yrs 6 17 26.09% 1.00 0.28 3.63 > 15 yrs 2 4 33.33% 1.67 0.24 11.45 Pearson Chi-square value of 6.002 with 2 degrees of freedom, p = 0.05. Case Seasonality Cases were examined for the month of o ccurrence and the data are shown in Figure 2-4. The majority of cases in the present st udy occurred during the warmer months of the year, with 68% between the months of Ap ril and September. Thirty-year average minimum temperatures in Gainesville, Florida remained above 60.3 F (15.7 C) during the months of May through October (Table 2-10).

PAGE 66

40 Table 2-10. Average minimum temperatures in Gainesville, Florida, USA (1961-1990)* Jan Feb Mar Apr MayJunJulAugSepOctNovDec Year Avg C 6.1 6.8 10.2 13.1 17.1 20.6 21.7 21.8 20.7 15.710.97.4 14.3 F 43.0 44.2 50.4 55.6 62.8 69.1 71.1 71.2 69.3 60.351.645.3 57.7 *Obtained from www.worldclimate.com 0 2 4 6 8 10 12 14 16Number Cases J an uary Febru a ry Marc h Ap ril M a y J u n e July Au g u st Septe mb er Octob e r N ovemb er Dec e mber Figure 2-4. Seasonal dist ribution of salmonella ca ses from horses 1999-2002 Group and Serovar Distribution All serovars in the present study have been previously identifie d in horses with the exception of a group F Salmonella serotyped as Rubislaw. There were three cases serotyped as Salmonella 4,5,12:i—monophasic, a type closely related to S Typhimurium, whose antigenic formula is 1,4,5,12:i—1,2 biphasic.134 Eight isolates (7.5%) were not serotyped as the samples were either cont aminated with other bacterial genera or contained more than one group or serovar of Salmonella The breakdown of salmonella group, serovar, and prevalence in this study is summarized in Table 2-11.

PAGE 67

41 Forty-eight additional isolat es from environmental sampling and animal species other than horses were collected and archiv ed during the reporting pe riod. These isolates were serotyped, and antibiotic sensitivity pr ofiles determined, but no other analyses were performed. Serovar and species of isolation information regarding these isolates is detailed in Table 2-12. Table 2-11. Salmonella serovars isol ated from 98 equine cases 1999-2002 Serovar Group Number of Cases % of Isolates Java B 23 23.45 Newport C2 13 13.27 Typhimurium B 8 8.16 Typhimurium var. CopenhagenB 7 7.14 Javiana D 7 7.14 Miami D 7 7.14 Saintpaul B 6 6.12 Muenchen C2 5 5.10 Anatum E 4 4.08 4,5,12:i-monophasic B 3 3.06 Newington E 2 2.04 London E 2 2.04 Mbandaka C1 2 2.04 Hartford C1 1 1.02 Agona B 1 1.02 Braenderup C1 1 1.02 Infantis C1 1 1.02 Meleagridis E 1 1.02 Reading B 1 1.02 Rubislaw F 1 1.02 Tallahassee C2 1 1.02 Thompson C1 1 1.02 Table 2-12. Salmonella isolates of environmen tal and species other than equids collected 1999-2002 Isolate Origin Serovar Group No. of Isolates Avian Manila E 1 Avian Infantis C1 1 Bovine Typhimurium B 4 Bovine Typhimurium var. Copenhagen B 4 Bovine Anatum E 2 Bovine Newport C2 1

PAGE 68

42 Table 2-12. Continued Isolate Origin Serovar Group No. of Isolates Bovine Mbandaka C1 1 Canine Typhimurium var. Copenhagen B 2 Canine Adelaide Not A-E 2 Canine Miami D 1 Environmental Java B 15 Environmental Newport C2 3 Environmental Typhimurium var. Copenhagen B 2 Environmental Javiana D 1 Environmental Anatum E 1 Environmental Typhimurium B 1 Environmental Tallahassee C2 1 Non-human Primate Typhimurium B 1 Other Mammal Hartford C1 1 Reptile Sub group 3 1 Rodent Typhimurium var. Copenhagen B 1 Outcome by Group or Serovar The relationship between group or serova r and outcome was investigated. The results are presented graphically in Figur e 2-5. There was a significant difference between isolates according to antigenic grouping in terms of mortality (p=0.033; Figure 2-6); survival was decrease d with isolation of group B salmonella serovars (43% survival). Other groups included C1 (60% survival), C2 (59% survival), D (92% survival), E (67% survival), and F (100% survival). Odds ratio data were determined by salmonella group; if the horse was infected with a group B Salmonella it was 15.7 times more likely to die (1.9 to 129.25, 95%CI) than if it were infected with a group D. Statistical summary of data is shown in Ta ble 2-13. Additional odds ratios, relative to infection with group D, were: C1—6 times more likely to die; C2—7 times more likely to die; and E—6 times more likely to die. Alt hough these ratios appeared large they were not significant as the respectiv e confidence inte rvals included 1.0.

PAGE 69

43 0102030405060708090100% Mortality Within Serotype 4 5 1 2 : i m o n o p h a s i c A g o n a A n a t u m H a r t f o r d J a v a M e l e a g r i d i s M i a m i M u e n c h e n N e w p o r t R e a d i n g S a i n t P a u l T a l l a h a s s e eT h o m p s o n T y p h v a r C o p e n T y p h i m u r i u m Figure 2-5. Mortality distribu tion, within serovar, of non-surviving equine salmonella cases 1999-2002 0 5 10 15 20 25 30Number Cases BC1C2DEF Salmonella Group Alive Dead Figure 2-6. Mortality by salmonella group in 88 cases with known outcomes

PAGE 70

44 Table 2-13. Effect of salmone lla group on mortality in 88 cas es of equine salmonellosis* Group Died Lived % Dead Odds Ratio Lower 95%CI Upper 95%CI B 26 20 56.5 15.7 1.90 129.25 C1 2 3 40.0 6.0 0.42 85.25 C2 7 10 41.2 7.0 0.74 65.95 E 2 4 33.3 6.0 0.42 85.25 D 1 12 7.7 1.0 0.06 17.90 F 0 1 0.00 ---------* Pearson Chi-square value of 12.129 with 5 degrees of freedom, p = 0.033. Within group B organisms, S. Typhimurium was associated with the highest mortality rate (75.0% mortality), followed by 4,5,12:i-monophasic (66.7% mortality), S. Saintpaul (60.0% mortality), S. Typhimurium var. Copenhagen (50.0% mortality), and S. Java (45.0% mortality). Four serovars (Tallahassee, Read ing, Meleagridis, and Agona) may appear more virulent due to low numbers of cases. Plasmid Profiling Plasmid profiles were completed for 104 c linical salmonella isolates. Several isolates in the main database were not anal yzed due to equivocal identificati on, inability to culture the provided isolate sample, or loss during storage. The majority of examined isolates, 64.4% (67/104), contai ned at least one large (> 20-k b) plasmid. Several isolates had additional smaller plasmids and some had more than one large plasmid. The breakdown of plasmid carriage by serovar is listed in Table 2-14. Table 2-14. Plasmid-positive salmone lla isolates by serovar 1999-2002 Serovar Number Plasmid Positive / Number Serovar % Positive of Total Isolates % Positive Within Serovar Java 23 / 25 22.12 92.00 Newport 5 / 13 4.81 38.46 Typhimurium 8 / 8 7.69 100.00 Typhimurium var. Copenhagen 7 / 8 6.73 87.50 Javiana 3 / 5 2.88 60.00 Miami 1 / 7 0.96 14.29

PAGE 71

45 Table 2-14. Continued Serovar Number Plasmid Positive / Number Serovar % Positive of Total Isolates % Positive Within Serovar Muenchen 6 / 6 5.77 100.00 Saintpaul 2 / 6 1.92 33.33 Unidentified 2 / 4 1.92 50.00 Anatum 2 / 4 1.92 50.00 4,5,12:i-monophasic 3 / 3 2.88 100.00 Newington 0 / 2 0.00 0.00 Hartford 2 / 2 1.92 100.00 London 1 / 2 0.96 50.00 Mbandaka 1 / 2 0.96 50.00 Agona 0 / 1 0.00 0.00 Braenderup 0 / 1 0.00 0.00 Infantis 0 / 1 0.00 0.00 Meleagridis 0 / 1 0.00 0.00 Reading 0 / 1 0.00 0.00 Rubislaw 1 / 1 0.96 100.00 Tallahassee 0 / 1 0.00 0.00 TOTAL67 / 104 64.42% Selected examples of results from agar ose gel electrophoresis plasmid profiles are shown in Figures 2-7 through 2-13. Figure 27 shows the plasmid profiles of several isolates extracted using the Birnboim and Doly method.135 All isolates were considered plasmid-positive; however, lanes 3, 5, 6, and 7 s how plasmids that were slightly larger than the 100-kb plasmid of 3306. Three of those four were S Typhimurium var. Copenhagen isolates of bovine orig in, and the fourth was an equine S Newport. Figure 2-8 shows the plasmid profiles of Case 71 in lane 4, Case 66 in lane 5, and Case 77 in lane 6. All of these isolates were determined to be plasmid-negative. These isolates were identified as S Miami, S Newport, and S Miami respectively. Figure 2-9 shows the plasmid profile of five cases which all de monstrated large (> 16-kb) plasmids, but varying in size relative to the 100-kb plas mid of the control strain. Only Case 89 (identified as S Typhimurium var. Copenhagen) possessed the spv genes—all others

PAGE 72

46 were negative. Interestingly, that plasmi d appears very close in size to the 100-kb virulence plasmid of the control S Typhimurium strain, while the other four isolate plasmids are large, but not necessarily equivalent in size. Figure 2-10 shows the plasmid profile of Case 44 in lane 5 and Case 43 in la ne 6. Both isolates appeared to possess a single large ( 100-kb) and a single small plasmid. Both of these isolates were identified as S Typhimurium var. Copenhagen from horses having antibiotic administration prior to admission, and neither of thes e isolates possessed the spv genes. Figure 2-11 shows the plasmid profile of Case 63 in lane 5. Ca se 63 appeared to possess multiple plasmids ranging in size from approximately 2kb to > 100-kb but did not possess the spv genes. This isolate was identified as S London from a horse with pr ior antibiotic administration (penicillin G). Interestingly, this isolate was not considered to be one of the more multidrug resistant strains to the 19 antimicrobials tested (resistance to more than 13/19 typical in multi-drug resistant strains). The isolate from Case 63 was resistant to clindamycin, doxycycline, erythromycin, oxacillin, penicilli n, rifampin, and tetracycline. Figure 2-12 shows the plasmid profile of Case 83 in lane 5 and Case 40 in lane 6. Case 83 did not appear to have any plasmid of any size visibl e on the gel, while Case 40 appeared to possess both a small and large plasmid (betw een 4-kb and > 100-kb respectively). Case 83 was identified as S. Reading from a necropsy larg e intestine specimen with unknown cause of death. Case 40 was identified as S. Typhimurium var. Copenhagen from a horse with prior antibiotic administration (trimethoprim-sulfa methoxazole, metronidazole and penicillin), and this was one of the few is olates of this serovar to not possess the spv genes. Figure 2-13 shows the plasmid profile of Case 36 in lane 3. This isolate was serotyped as the only group F Salmonella identified in horses ( S Rubislaw). Group F

PAGE 73

47 salmonellae have not been reported as common equid isolates in the literature. This isolate appeared to possess two large and two small plasmids but did not carry the spv genes and was obtained from 3/5 fecal samples submitted. Figure 2-7. Plasmid profiles of 9 clinical salmonella isolat es. Refer to Appendix C for specific isolate information. Lanes: 1) Previously extracted 100-kb plasmid of 3306, 2) Case 8, 3) Bovine isolate of S Typhimurium var. Copenhagen, 4) Case 11, 5) Bovine isolate of S Typhimurium var. Copenhagen, 6) Case 12, 7) Bovine isolate of S Typhimurium var. Copenhagen, 8) Case 6, 9) Case 10, 10) Case 7. spv Gene Analysis Of the 67 isolates found to be plasmid-pos itive 19.4% (12.5% of all isolates) were also PCR-positive for the spv genes examined. All positive isolates generated expected PCR product for all genes examined, they were an “all or none” result. Also, all isolates that were spv positive were also plasmid-positive, and it was assumed that the genes were located on a plasmid. Figures 2-14 through 2-18 show examples of the results obtained for the spv gene analyses of clinical isolat es. All isolates positive for the spv genes were exclusively group B salmonellae. Within this group they were also limited to only three serovars, Typhimurium, Typhimurium va r. Copenhagen, and 4,5,12:i-monophasic, an antigenically close relation to S. Typhimurium. 100 kb

PAGE 74

48 Figure 2-8. Plasmid profiles of 4 clinical salmonella isolat es. Refer to Appendix C for specific isolate information. Lanes: 1) supercoiled marker DNA, 2) 100-kb plasmid of 3306, 3) Case 78, 4) Case 71, 5) Case 66, 6) Case 77, 7) 100-kb plasmid of 3306, 8) blank. In Figure 2-14, there are multiple background bands in the spvA gel as well as one band in the negative control lane of the spvC gel. Since none of these bands were of the same intensity as the control, nor were they an appropriate size, they were considered artifacts. No clinical isolate tested in Figure 2-14 was considered positive for the spv genes. There appears to be a faint ba nd of appropriate size in lane 9 of the spvC gel; however, since the product band was of low intensity (compared to the positive control) and there was no correspondi ng positive result in the spvA or spvR (not shown) gels, the isolate was determined to be negative. 100 kb 16.21 kb 2.06 kb

PAGE 75

49 Figure 2-9. Plasmid profiles of 5 clinical salmonella isolat es. Refer to Appendix C for specific isolate information. Lanes: 1) supercoiled marker DNA, 2) 100-kb plasmid of 3306, 3) Case 89, 4) Case 92, 5) Ca se 85, 6) Case 93, 7) Case 96, 8) 100-kb plasmid of 3306. Figure 2-10. Plasmid profiles of 4 clinical sa lmonella isolates. Refer to Appendix C for specific isolate information. Lanes: 1) blank, 2) 100-kb plasmid of 3306, 3) Aged (>1month) plasmid extract of 3306, 4) Case 46, 5) Case 44, 6) Case 43, 7) Case 53, 8) blank. 100 kb 100 kb 16.21 kb 2.06 kb

PAGE 76

50 Figure 2-11. Plasmid profiles of 4 clinical sa lmonella isolates. Refer to Appendix C for specific isolate information. Lanes: 1) supercoiled marker DNA, 2) 100-kb plasmid of 3306, 3) Aged (>2month) plasmid extract of 3306, 4) Case 41, 5) Case 63, 6) Case 64, 7) Case 65, 8) supercoiled marker DNA. Figure 2-12. Plasmid profiles of 3 clinical sa lmonella isolates. Refer to Appendix C for specific isolate information. Lanes: 1) supercoiled marker DNA, 2) 100-kb plasmid of 3306, 3) 100-kb plasmid of 3306, 4) Case 82, 5) Case 83, 6) Case 40, 7) 100-kb plasmid of 3306, 8) blank. 100 kb 16.21 kb 2.06 kb 100 kb 16.21 kb 2.06 kb

PAGE 77

51 Figure 2-13. Plasmid profiles of 7 clinical sa lmonella isolates. Refer to Appendix C for specific isolate information. Lanes: 1) 100-kb plasmid of 3306, 2) Case 32, 3) Case 36, 4) Case 37, 5) Case 91, 6) Case 90, 7) Case 87, 8) Case 86. Figure 2-14. PCR product results for spvA and spvC genes in 9 clinical salmonella isolates, with positive and negative controls. Refer to Appendix C for specific isolate information. Lanes: 1) 1kb ladder DNA marker (Promega), 2) 3337 spv negative control, 3) 3306 spv positive control, 4) Case 86, 5) Case 87, 6) Case 88, 7) Case 117, 8) Case 100, 9) Case 101, 10) Case 102, 11) Case 103, 12) Case 104, 13) 1-kb ladder DNA marker (Promega), 14) blank. 100 kb 1000 bp 1000 bp

PAGE 78

52 Figure 2-15. PCR product for asd gene in 9 clinical salmone lla isolates (same isolates and orientation as Figure 2-14). Refer to Appendix C for specific isolate information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) 3337 spv negative control, 3) 3306 spv positive control, 4) Case 86, 5) Case 87, 6) Case 88, 7) Case 117, 8) Case 100, 9) Case 101, 10) Case 102, 11) Case 103, 12) Case 104, 13) 1-kb ladder DNA marker (Promega), 14) blank. Figure 2-15 shows the PCR products of the salmonella asd gene for the same isolates (and same orientation in the gel) as Figure 2-14. Note that there is no product visible for the isolate in lane #7 of either Figure 2-14 or 2-15. This isolate was positively identified as S. Newport previously; how ever, the sample taken for template DNA in the PCR mixture on this day was ta ken from a HE plate—where Salmonella spp. are identified based on their ability to produce hydrogen sulfide. Normally these samples were taken from isolates growing on LB plat es. Apparently the hydr ogen sulfide or some other compound in the culture medium interf ered with the PCR reaction, which would have caused a false-negative result to be generated had this control not been run simultaneously. An isolate was only evaluated for spv genes pending positive determination of the asd gene, which essentially valid ated that the isolate was a Salmonella spp. Figure 2-16 demonstrates five clinical salm onella isolates that were positive for the spvA gene. Serovars represented by these five isolates include S Typhimurium and 4,5,12:i-monophasic. 1000 bp

PAGE 79

53 Figure 2-16. PCR product results for spvA genes in 11 clinical salmonella isolates, with positive and negative controls. Refer to Appendix C for specific isolate information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) 3306 spv positive control, 3) 3337 spv negative control, 4) lost isolate, 5) Case 8, 6) Case 7, 7) Case 12, 8) Case 13, 9) Case 10, 10) Case 9, 11) Case 5, 12) Case 3, 13) Case 11, 14) Case 4. Figure 2-17. PCR product results for spvC genes in 11 clinical salmonella isolates, with positive and negative controls. Refer to Appendix C for specific isolate information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) 3306 spv positive control, 3) 3337 spv negative control, 4) Case 21, 5) Case 19, 6) Case 16, 7) Case 22, 8) Case 27, 9) Case 26, 10) Case 25, 11) Case 24, 12) Case 23, 13) Case 15, 14) Case 14. Figure 2-17 demonstrates four clinical salmonella isolates that were positive for the spvC gene. Serovars represented by these four isolates include S. Typhimurium var. Copenhagen, S. Typhimurium, and 4,5,12:i-monophasic. 1000 bp 1000 bp

PAGE 80

54 Figure 2-18. PCR product results for the asd gene in 11 clinical salmonella isolates (same isolates and orientation as Fi gure 2-17). Refer to Appendix C for specific isolate information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) 3306 spv positive control, 3) 3337 spv negative control, 4) Case 21, 5) Case 19, 6) Case 16, 7) Case 22, 8) Case 27, 9) Case 26, 10) Case 25, 11) Case 24, 12) Case 23, 13) Case 15, 14) Case 14. Figure 2-18 shows the PCR products of the salmonella asd gene for the same isolates (and same orientation in the gel) as Figure 2-17. This is the typical appearance of Salmonella spp. probed with the asd primer set. All of these isolates could subsequently be examined for the spv genes since they were positively determined to be salmonellae. Outcome by Presence of the Virulence Plasmid or spv Genes Short-term outcome was examined with resp ect to the presence or absence of the virulence plasmid or spv genes. Results are detailed in Table 2-15. Table 2-15. Summary outcome as determined by presence of the virulence plasmid and spv genes in 98 equine salmonella cases spv+ spvTOTAL Lived 2 47 49 Died 9 28 37 Unknown Outcome 2 10 12 TOTAL 13 85 There was a significant correlation between the presence of spv genes and mortality in this study population (p=0.001). Table 2-16 and Figures 2-19 and 2-20 depict the differences in short-term outcome between cas es with respect to the virulence plasmid and spv genes. Nine out of 11 cases (81.8%) with spv -positive salmonella strains had a 1000 bp

PAGE 81

55 fatal outcome as opposed to 28/75 (37.3%) of the cases with spv -deficient strains. The spv genes were restricted to group B salm onellae including serovars Typhimurium, Typhimurium var. Copenhagen, and 4,5,12:imonophasic. Horses infected with spv gene-positive salmonella serovars were 12.3 time s more likely to die than if they were infected with a spv negative strain. Also, if the or ganism was detected outside the gastrointestinal tract it was significantly more likely to be spv positive. Table 2-16. Effect of spv gene presence on mortality in 86 cases of equine salmonellosis where outcome was known* Exposure Died Survived% Dead Odds RatioLower 95%CI Upper 95%CI spv Positive 9 2 81.8 12.30 2.59 58.41 spv Negative 28 47 37.3 1.00 0.52 1.91 Pearson Chi-square value of 14.070 with 1 degree of freedom, p=0.001. spv +18% 82% Lived Died Figure 2-19. Outcome in equine salmonella cases, as influenced by presence of the spv gene locus.

PAGE 82

56 spv -63% 37% Lived Died Figure 2-20. Outcome in equine salmonella cases, as influenced by absence of the spv gene locus. Effect of Clinical and Labora tory Parameters on Outcome A large number of independent clinical and laboratory variables were investigated with respect to predicting out come in horses infected with Salmonella Individual variables with a p-value less th an 0.2 were included in the original model. There was no significant difference between survivors and non-survivors with resp ect to total serum protein (TSP) at admission (p=0.197) or TSP at death or discharge (p=0.198), however both factors significantly impacted outcome wh en investigated using forward stepwise logistic regression. The median TSP at ad mission and discharge for the survivors was 6.25 (mean = 6.45, with 6.1 to 6.6 95%CI) and 6.15 (mean = 6.1, with 5.9 to 6.4 95%CI) respectively. The median TSP at admission and death for the non-survivors was 6.8 (mean = 6.7, with 6.4 to 7.0 95%CI) and 5.5 (mean = 5.7, with 5.3 to 6.2 95%CI),

PAGE 83

57 respectively. The total white blood cell and neutrophil counts at presentation did not predict outcome. Regardless of clinical presentation or syndrome, the presen ce or absence of diarrhea was recorded where available for each case. True to the predominantly enteric nature of this disease, 74 out of 85 (87.1%) cases exhibited diarrhea at some point during hospitalization. Eleven cases (12.9%) did not develop diarrhea at any point during hospitalization, and the information could not be determined for 21 cases. The presence of absence of diarrhea did not predict survival. Forward stepwise logistic regression analysis indicated that four categorical predictor variables had a significant impact on outcome: spv gene status, TSP at admission, TSP at death or discharge, and da ys of hospitalization were all related to outcome. The average number of days sp ent hospitalized was 10.2, with a minimum of one day (6 cases) and a high of 48 days. Of those cases that had a 3-day or less hospitalization period, there was a 91.66% mo rtality rate—these cases likely were admitted with severe disease, and were euthanatized due to expense, prognosis or complications, with the Salmonella not being confirmed until after death. The significant variables with test statistics are included in Table 2-17. Table 2-17. Logistic regression model w ith variables predictive of outcome Variable B SE Wald StatisticdfSignificance Exp (B) spv genes -2.7101.0776.332 10.012 0.067 TSP at admission 1.0500.3986.967 10.008 2.859 TSP at death or discharge -1.0990.3569.516 10.002 0.333 Days of hospitalization -0.1690.0617.633 10.006 0.845

PAGE 84

58 Relationship Between Proportion of Po sitive Fecal Salmonella Cultures and Outcome A Kruskal-Wallis test was used to investig ate the percentage of submitted samples that were positive with respect to outcome. There was a significant difference (p=0.042) between those that lived and t hose that died. Horses that survived had a median of 60% of their fecal cultures that were positive (mean = 62.71%, with 53.8% to 71.6% 95%CI), as compared to horses that died spontaneous ly or that were euthanatized, where the median percentage of positive cultures was 100% (mean = 77%, with 65.8% to 88.2% 95%CI). Antibiotic Resistance Profiles Antibiograms were obtained for 101 isolates Complete MIC and resistance data for all isolates can be found in Appendix E. A summary of the antibiotic susceptibilities of 101 cases is displayed in Table 2-18 a nd the complete antibiogram of Case 78, demonstrating the only isolate with re duced sensitivity to the fluoroquinolone enrofloxacin, is shown in Table 2-19. Table 2-18. Antibiotic susceptibilities for 101 equine salmonella isolates. The reported % susceptible, % intermediate, and % re sistant, are only for those isolates with data for that antibiotic. Antibiotic % Susceptible % Intermediate % Resistant Clindamycin 1.1 0 98.5 Erythromycin 1.15 0 98.5 Penicillin 0 0 100.0 Oxacillin 1.1 0 98.9 Rifampin 1.1 0 98.9 Doxycycline 58.9 0 41.0 Tetracycline 61.0 0 39.0 TrimethoprimSulfamethoxazole 61.4 0 38.6 Amoxicillin-Clavulanic Acid 67.4 0 32.6 Ampicillin 68.3 0 31.7 Ceftiofur 69.0 0 31.0 Cefazolin 70.0 2.0 28.0

PAGE 85

59 Table 2-18. Continued Antibiotic % Susceptible % Intermediate % Resistant Ceftazidime 70.0 10.0 20.0 Gentamicin 75.2 9.9 14.9 Chloramphenicol 85.0 1.0 14.0 Amikacin 92.9 0 7.1 Enrofloxacin 99.0 1.0 0 Imipenem 100.0 0 0 Nitrofurantoin 100.0 0 0 Table 2-19. Antibiotic susceptibility report fo r Case 78, with intermediate resistance to enrofloxacin Drug MIC RS Drug MIC RS Drug MIC RS Amikacin <=2 S Amox/Clav >16 R Ampicillin >16 R Cefazolin >16 R Ceftazidime 32 R Ceftiofur >4 R Chloramp 32 R Clindamy. >2 R Doxycyc. >4 R Enroflox. 1 I Erythrom. >4 R Gentamicin >8 R Imipenem <=1 S Nitrofur. <=32 S Oxacillin >4 R Penicillin >16 R Rifampin >4 R Tetracyc. >16 R TMP-Sulfa >4 R Antibiotic Resistance Transformation Electroporation transformation was successful in transferring cefazolin resistance from the three salmonella isolates; but arci ng due to the presence of buffer salts could have potentially damaged the plasmid DNA to the extent of generating false-negative results. Chemical transformations were performed with succe ssful transference of cefazolin, ceftiofur and ampicillin resistance fr om all three isolates Figures 2-21 through 2-24 show the gel electrophoresis results fo r these analyses. Cefazolin-resistant E. coli was shown to contain a new plas mid equivalent in size to th e original cefazolin-resistant transforming salmonella isolate (F igure 2-21). The untransformed E. coli did not possess a plasmid, and is included for comparison.

PAGE 86

60 Figure 2-21. Plasmid profiles of 3 clinical salmonella isolates and E. coli transformed with plasmid DNA from those isolates. Refer to Appendix C for specific isolate information and Appendix E for an timicrobial susceptibilities. Lanes: 1) 100-kb plasmid of 3306, 2) Untransformed E. coli DH5 3) E. coli DH5 transformed with Case 97, grown in CEF, 4) E. coli DH5 transformed with Case 92, grown in CEF, 5) E. coli DH5 transformed with Case 98, grown in CEF, 6) Transforming plasmid DNA fr om Case 97, 7) Transforming plasmid DNA from Case 92, 8) Transformi ng plasmid DNA from Case 98. Figure 2-22. Plasmid profiles of 2 clinical salmonella isolates and E. coli transformed with plasmid DNA from those isolates. Refer to Appendix C for specific isolate information and Appendix E for an timicrobial susceptibilities. Lanes: 1) Untransformed E. coli DH5 2) 100-kb plasmid of 3306, 3) E. coli DH5 transformed with Case 98, grown in AMP, 4) E. coli DH5 transformed with Case 98, grown in NAX, 5) E. coli DH5 transformed with Case 98, grown in CEF, 6) E. coli DH5 transformed with Case 92, grown in AMP, 7) E. coli DH5 transformed with Case 92, grown in CEF, 8) blank. 100 kb 100 kb

PAGE 87

61 Figure 2-23. Plasmid profiles of 2 clinical salmonella isolates and E. coli transformed with plasmid DNA from those isolates. Refer to Appendix C for specific isolate information and Appendix E for an timicrobial susceptibilities. Lanes: 1) Untransformed E. coli DH5 2) 100-kb plasmid of 3306, 3) E. coli DH5 transformed with Case 97, grown in AMP, 4) E. coli DH5 transformed with Case 92, grown in NAX, 5) E. coli DH5 transformed with Case 97, grown in NAX, 6) E. coli DH5 transformed with Case 97, grown in CEF, 7) blank, 8) blank. In Figure 2-23, lane 5 shows an isolate (C ase 97) that transfer red resistance to NAX, CEF, and AMP. The plasmid transferring resistance to ceftiofur is larger than the other two transforming plasmids (which appear to be the same size). Looking at the plasmid profile of Case 97 in Figure 2-24, lane 5—there are three large plasmid bands visible (2 smaller than 100-kb and 1 larger). This isolate most likel y is carrying the AMP and CEF resistance genes on the same plasmi d and the NAX resistance gene on another larger plasmid. On closer examination of Figure 2-21—red box, the second larger plasmid is visible in Case 97 (lane 6), al ong with the other transforming plasmids of homogenous size. 100 kb

PAGE 88

62 Figure 2-24. Plasmid profile of Case 97—lane 5. The red box delineates 3 large plasmid bands that are visible in the upper part of the lane. This isolate transferred ceftiofur, cefazolin, and ampicillin resist ance via two different plasmids (the lower two). These experiments demonstrate that cefazo lin, ampicillin, and ceftiofur resistance in Cases 92, 97, and 98 were carried on plasmids which are smaller than the 100-kb virulence plasmid and do not contain the spv genes. Site of Salmonella Isolation The majority of isolates in this study were obtained from fecal samples (80.0%). Isolates from various segments of the gastroin testinal tract were examined separately and as part of the group of gastro intestinal isolates. If isolat es from all enteric sites are considered together, the proportion of gastrointestinal isolates in the study rises to 93.3%. Isolate distribution by site of inf ection is summarized in Table 2-20. Table 2-20. Clinical salmonella isolates fr om 105 equine cases by location of cultured specimen Anatomic Site Number of Cases % of Total Cases Feces 84 80.0 Small Intestine (necropsy or surgery) 7 6.7 Large Intestine (necropsy) 4 3.8 Synovial (joint) Fluid 2 1.9 Lung (necropsy) 1 1.0 100 kb

PAGE 89

63 Table 2-20. Continued Anatomic Site Number of Cases % of Total Cases Duodenum (necropsy) 1 1.0 Gastric Reflux 1 1.0 Abscess 1 1.0 Rectal Biopsy 1 1.0 Blood 1 1.0 Liver (necropsy) 1 1.0 Physis (necropsy) 1 1.0 The relationship between serova r and site of infection is summarized in Table 2-21. Figure 2-25 illustrates the systemic isolates comp ared to the gastrointestinal isolates by group. Table 2-21. Systemic sites of salmon ella infection in horses by serovar Serovar Number and % of Systemic Isolates Hartford 1/1 100.0 Typhimurium 3/8 37.5 Muenchen 1/5 20.0 Typhimurium var. Copenhagen 1/7 14.3 Newport 0/13 0.0 Java 0/23 0.0 Javiana 0/7 0.0 Miami 0/7 0.0 Saintpaul 0/6 0.0 Anatum 0/4 0.0 4,5,12:i-monophasic 0/3 0.0 Newington 0/2 0.0 London 0/2 0.0 Mbandaka 0/2 0.0 Agona 0/1 0.0 Braenderup 0/1 0.0 Infantis 0/1 0.0 Meleagridis 0/1 0.0 Reading 0/1 0.0 Rubislaw 0/1 0.0 Tallahassee 0/1 0.0

PAGE 90

64 0 10 20 30 40 50 60 70 80 90 100 BC1C2DEF Salmonella Group % Systemic % Gastrointestinal Figure 2-25. Systemic equine salmonella isolat es compared to gastrointestinal isolates by group Extra-intestinal isolates were 16.18 times more likely to carry the spv genes on a virulence plasmid than enteric isolates (p =0.001). All salmonellae that contained the spv genes also carried a large plasmid; no isolates were plasmid-negative and spv -positive. This was significant with a 95% confid ence interval of 3.54 to 74.06. Data are summarized in Table 2-22. Table 2-22. Relationship of the virulence plasmid and spv genes to isolate location in 98 cases of equine salmonellosis* Site of Isolation spv Positive spv Negative% Positive Odds Ratio Lower 95%CI Upper 95%CI Extra-intestinal 4 3 57.1% 12.15 2.34 63.10 Intestinal 9 82 9.9% 1.00 0.38 2.65 Pearson Chi-square value of 18.994 with 1 degree of freedom, p=0.001.

PAGE 91

65 Multi-Serovar Salmonella Infections Six horses had more than one serovar of Salmonella isolated from them during hospitalization. All of the hor ses with multi-serovar salmonella infections developed or were admitted with diarrhea, all survived, and all of the isolates were obtained from fecal specimens. Table 2-23 details the groups, ser ovars, and relevant case information from these 6 horses.

PAGE 92

66 Table 2-23. Details of multi-serovar salm onella infections in six horses 1999-2002 Case ID Sex Age Specimen Origin Salmonella Group Serovar Clinical Syndrome OutcomeDiarrhea 80 Feces C2 muenchen 36 M 7y Feces F rubislaw Diarrhea, Fever Lived YES 70 Feces B java 69 M 6y Feces D javiana Post-Op Colic Diarrhea Lived YES 52 Feces C2 newport 58 F 3m Feces D miami Diarrhea, Fever Lived YES 75 Feces B java 71 M 3m Feces D miami Diarrhea Lived YES 73 Feces C1 hartford 39 F 6y Feces C2 newport Diarrhea, Fever Lived YES 81 Feces B Multiple serovars (NVSL sample) 37 M 3m Feces C2 muenchen Colic Diarrhea, Chronic Diarrhea Lived YES

PAGE 93

67 Discussion Risk factors for the development of salmone lla infection in horses have been well described and were not investigated in th e present study. The original aim of the proposed study was to contrast Salmonella spp. shed from diseased animals in a hospital setting with those recovered fr om a population of asymptomatic animals at pasture. This comparison was to focus on isolate serovar, grouping, plasmid, and spv gene status. Unfortunately, despite extensive and repeated culturing of animals at pasture we were unable to isolate Salmonella from any asymptomatic animal. This finding was a surprise, even in the face of a low pr evalence (0.8%) reported in the recent NAHMS survey of North American horses.95 With prevalence estimations ranging between 0% to 70% of the horse population, depending on the risk group being sampled and the type of diagnostic test used, it was e xpected to find at least one hor se asymptomatically shedding Salmonella in their feces. The sampling was done ove r several seasons to ensure that the influence of temperature, weather patterns and time of year was minimized, and several samples were taken from each animal over a pe riod of time to maximi ze the possibility of identifying periodic shedding episodes. There are several explanations for this negative result: too few samples examined per horse, cu lture techniques too insensitive to identify the low levels of bacteria shed by healthy hors es, or more likely, that the true prevalence was so low that insufficient numbers of an imals were sampled. The sample collection procedure and culture techniques were vali dated with samples from hospital patients known to be shedding Salmonella Enrichment (with sodium selenite cystine, or tetrathionate broth) and culture is cu rrently the gold sta ndard for diagnosing Salmonella from fecal samples in horses, neverthe less the technique is not 100% sensitive.50

PAGE 94

68 The study focus shifted towards a clos er examination of the hospitalized population, including descriptive da ta, risks factors within this population associated with outcome, including organism group, serovar and spv gene status. With respect to the descriptive data we were restricted by an inability to obtain accu rate hospital population demographics for the period in question. The breed distribution likely reflected the regional and hospital population. No breed predilection has been reported for non-host adapted salmonella infection in simila r populations of hospitalized horses.44;136 The mean age of affected animals in the present study was low in comparison to published values, but likely reflects th e referral horse population in North Central Florida. This teaching hospital has a large cas eload of young horses and foals due to close proximity to breeding farms, and this factor likely contributed significantly to the low mean age. Twenty-eight cases (28.57%) were in horses le ss than 6 months old, consistent with the opportunistic nature of Salmonella in the very young, immunosuppressed, or geriatric animals.107 Olsen et al. showed a very similar distribution regarding isolation rates by age in humans, with over 48% of 441,863 isolat es coming from individuals less than 19 years of age.137 The unbalanced distribution of the case population may also be a reflection of compounded risk factors asso ciated with age (e.g., younger horses may undergo surgery more often than older ones, or younger horses are kept in larger groups and may have an increased exposure to pa thogens relative to solitary individuals). As expected and reported in the literature, the largest num ber of cases in the present study occurred during the warmer months of the year, 68% from April through September. The seasonal predominance of salmonellosis in horses is typically highest during the warmer summer months95;99 and this seasonality wa s likely extended due to

PAGE 95

69 the warm Florida climate. Thirty-year aver age minimum temperatures in Gainesville, Florida remained above 60.3 F (15.7 C) during the months of May through October (Table 2-10).The wide spectrum of Salmonella recovered in this popul ation is consistent with previous studies in horses. There were significant associations between salmonella grouping and spv gene presence and mortality. Animals infected with group B Salmonella were nearly 16 times more likely to die than infected with the common Group D bacteria. It was not surpri sing that the highest percentage of non-survivors occurred in the group B organism S Typhimurium and related serovar groups S Typhimurium var. Copenhagen and 4,5,12:i-monophasic. S Typhimurium is a serious pathogen worldwide, with higher mortality rates than many othe r serovars, even within the group B. S Typhimurium and S Typhimurium var. Copenhagen we re shown to cause significantly higher fatality rates than all other serova rs in two studies of hospitalized horses.99;138 This effect could likely be attributed to the presence of viru lence plasmids, other antimicrobial resistance factors, or undetermin ed virulence factors significant in horses. Plasmid-bearing, spv gene positive organisms we re restricted to group B Salmonella Eighty seven and a half percent of S Typhimurium isolates were spv gene positive; 29% of S Typhimurium var. Copenhagen isolates were spv gene positive; and all 3 isolates of 4,5,12:i-monophasic contained spv virulence genes. It is im portant to point out however that many group B Salmonella do not carry spv genes. This includes S Java (none of 23 isolates), S Saint Paul (0 of 6), S Agona and S Reading (0 of 1, respectively). Extra-intestinal isolates were limited to groups B, C1, and C2. The serovars recovered from those isolates included S. Hartford, S. Typhimurium, S. Muenchen, and S. Typhimurium var. Copenhagen. S. Typhimurium was the only serovar with more than

PAGE 96

70 one systemic isolate, and more than 37% of all S. Typhimurium isolates were from systemic sites. Systemic isolates had a significantly higher potential of carrying the spv genes. This finding may indicate a similar ro le for the salmonella virulence plasmid and these genes in horses, as demonstrated in calves,26 humans,22;139 and mice.21 Montenegro et al. showed that virulence plasmids were detected in nearly 100% of extra-intestinal isolates from human blood as well as cattle or swine internal organs.23 In summary, spv gene-containing isolates in horses are likely restricted to certain group B salmonellae, are more likely to be reco vered outside the intes tinal tract, and are more commonly associated with a negative outcome than nonspv gene-containing isolates. The fact that all spv positive isolates were Gr oup B salmonellae is also in agreement with published reports. Eleven diffe rent serovars have been reported to carry virulence plasmids (including S. Typhimurium); however, not all isolates within those serovars will necessarily contain a virulence plasmid.140 In the present study, one S. Typhimurium and six S. Typhimurium var. Copenhagen isolates did not possess the spv genes. Verification that the spv genes were located on the plas mid (and not integrated into the chromosomal DNA) was not performed, but could be determined by transferring the gel electrophoresis products to solid membranes, and then DNA-DNA hybridization to the plasmid band (Southern blot). Chromosomal integration of the spv genes has only been reported in subspecies II, IIIa, IV, and VII which typically infect cold-blooded vertebrates.141;142 These subspecies do not inf ect warm-blooded vertebrates—only subspecies I isolates have dem onstrated mammalian pathogenicity.134 It was also shown in a mouse-avirulent subspecies IV isol ate that the chromosomally integrated spv genes

PAGE 97

71 were not normally expressed and complement ation with the entire virulence plasmid from S. Typhimurium did not cause the is olate to become mouse virulent. The ability to recover through bacterial culture, Salmonella spp. from fecal samples, correlated with outcome. In genera l, animals with significant enteric disease and higher mortality were more likely to retu rn a larger proporti on of positive cultures than those with milder disease. This may be related to the number of organisms being shed and/or to the immune status of the animal, with immunocompromised individuals unable to significantly respond to the organism. A recent retrospective study determined that low serum total protein concentrations were associated with failure to survive in horses admitted for acute diarrhea.143 Using limited clinical and laboratory data we perfor med a stepwise logistic regression analysis in order to unmask factors that may be impor tant in predicting outcome in horses with salmonella infection. We also concluded th at total plasma protein was an important determinant of outcome, in addition to spv gene status, and dura tion of hospitalization. Unfortunately none of these factors, with the exception of total plasma protein at admission, could be used reliably to predict out come in the clinical setting. The protein concentration at admission was higher in th e non-surviving group, lik ely reflecting more severe hemoconcentration in those horses as sociated with acute fl uid losses. Although this finding is of clinical interest it is unlikel y by itself to influence the decision to pursue treatment. Overall, spv gene-containing isolates in horse s are likely restricted to group B organisms, more likely to be recovered outsi de the intestinal tract, and more commonly associated with a negative outcome than spv gene-negative isolates.

PAGE 98

72 Examination of in vitro salmonella sensitivity data is an important facet of clinical practice. Not only does sensitivity data guide therapy but also is important in terms of monitoring for drug resistance. The recogniti on of fluoroquinolone resi stance in resident strains of Salmonella is particularly important. A recen t report detailed an outbreak and general increase in the number of multidrug-resistant S Newport being isolated from humans.144 These isolates were resistant to amox icillin/clavulanate, ampicillin, cefoxitin, ceftiofur, cephalothin, chloramphenicol, streptomycin, sulfamethoxazole, and tetracycline, and the resistance can be attrib uted to the presence of plasmids carrying a blacmy gene, which produces AmpC-type enzymes th at confer resistance and are termed Newport MDR-AmpC strains. One isolate (cas e 14) was from a horse previously treated with antibiotics (penicillin and trimethoprim-sulfamethoxazole) that developed diarrhea attributed to S Newport. This isolate had a resist ance pattern strikingly similar to the multidrug-resistant S. Newport described in th e report (resistant to amoxicillin/clavulanate, ampicillin, cefazolin, ceftazidime, ceftiofur, chloramphenicol, clindamycin, doxycycline, erythromycin, oxacilli n, penicillin, rifampin, tetracycline, and trimethoprim-sulfamethoxazole) and also carrie d a large plasmid that did not contain the spv genes. These strains of Salmonella are commonly associated with dairy farms, sick cows, and unpasteurized milk or cheese.144 Only four serovars of Salmonella accounted for the 22 multidrug-resistant isolates identified in this study (resistant to 8 drugs out of the 12 clinically relevant drugs tested). S. Java accounted for ten, S. Typhimurium var. Copenhagen accounted for five, S. Javiana accounted for two, and S. Newport accounted for one. Interestingly, S. Typhimurium var. Copenhagen was also isolat ed from four hospitalized cows during the

PAGE 99

73 same time period, and similar to what was reported for S. Newport, dairy cattle could be reservoirs as well as modulators of resistance pressure in this serovar. It should be noted that in vitro sensitivity data does not directly correlate with in vivo susceptibility due to the normally intracellular location of this organi sm. This is particularly true for non-lipid soluble antibiotics such as gentamicin. Ba sed on this population of organisms and their susceptibility data, clindamycin, erythromyci n, penicillin, oxacilli n, and rifampin cannot be recommended for therapeutic treatment of salmonella infections, due to more than 95% of all isolates being resist ant to each of those drugs. This is expected based on the mode of action and gram-positiv e bacterial spectrum of these drugs. Drugs typically used for peri-operative prophylaxis such as the firs t generation cephalosporin cefazolin or the aminoglycoside gentamicin, had approximately 28% and 15% resistance respectively. Amikacin, enrofloxacin, imipenem, and nitrofuran toin each had greate r than 92% of all isolates susceptible. A popular antibiotic selec tion for treatment of sa lmonella infections in adult horses is the fluor oquinolone enrofloxacin. Enrofl oxacin has an excellent gramnegative spectrum, is accumulated within macrophages, and is effective against intracellular organisms such as Salmonella In this study, a singl e isolate demonstrated intermediate resistance to this drug (Case 78) ; all others were susceptible. Case 78 was from an 18y old Welsh Pony that presented for a gastric impaction. Post-operatively, this horse developed diarrhea a ssociated with a group B— S. Java, and was later euthanatized. This patient was treated with penicilli n, gentamicin, and metronidazole during hospitalization and also was trea ted with enrofloxacin for the Salmonella Surveillance for fluoroquinolone resistance in Salmonella (especially veterinary isolates) is extremely important, as these drugs are reserved fo r life-threatening infections in humans.145-147

PAGE 100

74 Antibiotic resistances shown to be carried on plasmids include beta-lactamases and extended-spectrum beta-lactamases,148 ampicillin,149 tetracycline,150 quinolones,151 trimethoprim and sulfonamides.149;152 The 3 cases demonstrating successful transference of antibiotic resistances we re all the same serovar ( S. Java), a prevalent serovar in the hospital during that time period. Prescott reported that multiple-antibiotic resistance is a problem only in S. Typhimurium and not in other salmonella serovars.153 This was found not to be the case, as most of the multiple-an tibiotic resistant serovars in this study were not S. Typhimurium, but S. Java. Preliminary data gene rated by this study supports the contention that the majority of large plasmi ds associated with clinical isolates of Salmonella from horses are likely antimicrobial resistance plasmids or R plasmids.

PAGE 101

75 CHAPTER 3 EXPERIMENTS Background The Horse: Classic Large Intestine Fermenter The horse is exquisitely dependent on the protozoal and microbial population within its cecum and large colons to fermen t otherwise indigestible cellulose foodstuffs, similar to the forestomach ecosystem in ruminants.154 The cecum in the horse and pony is analogous to the rumen, and SCFA pr oduction and metabolism in the cecum alone have been shown to supply approximately 30% of a horse’s daily digestible energy intake.155 This biological fermentation vat wo rks best under the ideal conditions of constant influx of substrate and consumption or efflux of by-products; i.e., the horse is best suited to eat on a continuous basis, as opposed to the meal feeding activities of omnivores and carnivores.156 Domestication of the horse and modernization of the horse industry have significantly ch anged management strategies, specifically diet composition and practices associated with feeding thos e diets. An eventual migration toward confinement, structured exercise, and concen trate meal feeding has exaggerated problems uncommonly seen in wild horse populations, su ch as colic, gastric ulcer disease, and infectious diarrhea.157 Significant differences in dietary composition would theoretically have great impact on the amount of SCFA in the cecum and colon, but surprisingly, this is not the case. It has been shown that horses fed concentrated grain and carbohydraterich diets do not produce excessive amounts of SCFA as compared to horses fed control diets,155;158 but they can have significantly diffe rent SCFA ratios than horses consuming

PAGE 102

76 exclusively forage ba sed diets or grazing.159 These ponies developed lower acetate and higher propionate molar proportions on incr eased carbohydrate diets. The results in equine research agree with human studies where no appreciable effects of dietary carbohydrate content on fecal SCFA concentrati ons were identified, but differences in the relative proportions of SCFA were reported.160 Abrupt dietary decreases in the proportions of resistant star ches to fermentable carbohydr ates, however, can alter the microbiota and local climate of the cecum, such as decreasing the pH and total protozoal counts and increasing the numbers of lactate-producing anaerobes.161;162 The effect of dietary manipulation of the in testinal microenvironment, particularly with respect to growth, attachment, and invasion of pathoge ns, is worthy of further investigation. Equine Cecal Anaerobic Flora a nd SCFAs in the Normal Animal A summary of the reported values for to tal culturable anaerobes in the equine cecum are detailed in Table 3-1 and a summa tion of several reports in the literature measuring equine cecal SCFA concentrations in normal animals is shown in Table 3-2. Table 3-1. Summary of liter ature reports quantifying equi ne cecal anaerobic bacteria Author Total Bacteria and Units of Measurement Fistulated or Whole Animal Maczulak et al.163 DMC = 2.37 – 4.72 x 109 per ml Culture = 1.86 – 3.65 x 108 per ml Fistulated Kern et al.164 DMC = 458 – 702 x 107 per gram Culture = 35 181 x 107 per gram Whole McCreery et al.165 DMC = 1010 1011 per gram Mackie and Wilkins166 Culture = 25.85 x 108 per gram Whole Kern et al.154 DMC = 642 x 107 per gram Culture = 492 x 107 per gram Whole Julliand et al.167 Culture = 4.2 x 108 per ml Fistulated Goodson et al.162 Culture = 104 per gram Fistulated Medina et al.168 Culture = 2.4 x 108 per ml Fistulated

PAGE 103

77 Table 3-2. Literature reports quantifying nor mal equine cecal SCFA concentrations Acetate Butyrate Propionate TOTAL Glinsky155 73.7 % 51.00 mM 9.5 % 6.57 mM 16.7 % 11.56 mM 69.2 mM Eldsden41 73.2 % 7.2 % 19.65 % Kern164 43.08 M/ml 4.03 M/ml 13.48 M/ml 70.00 M/ml ( mmol/L) Kern154 74.7% 72.76 M/g 5.6% 5.45 M/g 18.4% 17.92 M/g 97.4 M/g Mackie and Wilkins166 99.9 mM 3.8 mM 12.5 mM 118 mM de Fombelle169 71.99 % 57.75 mmol/L 6.96 % 5.60 mmol/L 19.39 % 15.55 mmol/L 80.22 mmol/L Horspool170 34.4 mmol/L 53.1 % 12.2 mmol/L 21.3 % 10.9 mmol/L 18.7 % 65.0 mmol/L Medina et al.168 47.15 mM 70.2 % 3.86 mM 5.68 % 15.25 mM 22.15 % 67.3 mM RANGE 34.4 – 99.9 mM 3.8 – 12.2 mM 10.9 – 15.5 mM 65 – 118 mM Antimicrobial Effects on Normal Anaerobic Flora Antimicrobial therapy at best is an artistic treatment modality, because its therapeutic effects extend to al l susceptible bacteria living in or on the host, not just the pathogenic strains. Disruption of the norma l commensal microenvironment can often be equally as detrimental to the host as the infection being treated. In vitro and in vivo methods have historically been used to exam ine the effects of antib iotic administration on the autochthonous flora of th e gastrointestinal tract, an environment exquisitely dependent on the presence and activity of nu merous species of commensal bacteria. Strictly anaerobic bacteria ar e responsible for many of the me tabolic functions attributed to this ecosystem. Table 3-3 shows a current summary of multiple studies examining the effects of various an tibiotics on fecal SCFA producti on (some in vitro studies are included) as well as effects on the levels of culturable aerobi c and anaerobic bacteria in the feces. Overall, it is appa rent that disruptions in the nor mal flora of the skin, mucous

PAGE 104

78 membranes, gastrointestinal and urogenital tracts are comm on and direct effects of antibiotic administration, and that route, dose rate, mode of action, duration of therapy, and patient metabolic status are all important in mediati ng the particular sequence and severity of adverse effects attribut ed to a particular antimicrobial.

PAGE 105

79Table 3-3. Literature summary of antibiotic eff ects on fecal bacteria and short-chain fatty acids Antibiotic Species Route of Administration Duration of Treatment Effect on Fecal Aerobe Culture* Effect on Fecal Anaerobe Culture* Effect on Fecal SCFA* Amikacin Equine Equine Intravenous Oral One dose One dose ---171 ---171 ---171 --to 171 ---171 ---171 Amoxicillin Human Oral 7 days to ---172 to ---172 Ampicillin Human Oral 5 days 172 172 Bacitracin Human Equine Oral Oral 6 days 4 days 86 (examined microflora associated functions – no cultures) 86 (examined microflora associated functions – no cultures) 82 Ceftriaxone Human Intramuscularly 5 days to sl. 85 85 85 Clarithromycin Human Oral 7 days (E. coli)88 88 Clinafloxacin Human Oral 7 days 84 84 Clindamycin Swine In vitro Colon Simulation Technique (COSITEC) 5 days ( in vitro )92 Co-triomoxazol (TrimethoprimSulfamethoxazole) Human Oral 6 days to ---82 Doxycycline Human Oral 6 days to ---82 Erythromycin Human Oral 6 days 82

PAGE 106

80Table 3-3. Continued Antibiotic Species Route of Administration Duration of Treatment Effect on Fecal Aerobe Culture* Effect on Fecal Anaerobe Culture* Effect on Fecal SCFA* Metronidazole Swine In vitro Colon Simulation Technique (COSITEC) 6 days ( in vitro )83 Moxifloxacin Human Oral 7 days 88 to ---88 Nalidixic Acid Human Oral 6 days to ---82 Ofloxacin Human Oral 6 days to ---82 Penicillin Human Oral >3 days ---73 Pivampicillin Human Oral >3 days ---73 Streptomycin Murine Oral 7 days 80 Tetracycline Human flora murine model Oral 6 weeks to 87 ---87 to ---87 TrimethoprimSulfadiazine Equine Equine Intravenous Oral 5 days to 49 to 49 Vancomycin Human Swine Oral In vitro Colon Simulation Technique (COSITEC) 6 days 6 days 82 ( in vitro )83 = strong suppression, = moderate suppression, = minimal suppression, --= no suppression, = increase

PAGE 107

81 Antimicrobial Effects on SCFAs The effects of orally administered antibio tics on gastrointestinal SCFAs are likely not a direct effect on the SCFAs themselves but rather an effect on the organisms producing them. Several studies have s hown that antibiotic administration has a significant effect on the distribution and con centration of SCFA in the feces of normal humans,82;84;85 but not horses, although data are limited.171 The effect of a single oral or intravenous dose of amikacin, an aminoglycosid e with minimal predicted activity in an anaerobic environment such as the distal gast rointestinal tract, was examined in normal horses.173 Another publication reported the eff ects of a single intravenous dose of oxytetracycline on cecal SCFAs in one pony, th ough the nature of the article was to validate methodology, and no in terpretation was offered.174 In one of the only comprehensive prospective studies involving Equidae the investigator examined the effects of several antibiotics given by various routes, on cecal levels of SCFAs in horses, ponies, and donkeys.170 This study found that antibiotic administration typically altered cecal and fecal SCFA levels by increasing lactic acid concentrations. Lactic acid is a SCFA but is considered to be a non-volatile fatty acid (NVFA) in co ntrast to acetate, propionate, and butyrate. A summary of that work is detailed in Table 3-4. The primary shortcoming of this research is that it onl y examined the effects of a single dose of antibiotic on the variables of interest. The effects of repeated dosing regimens are in need of investigation. Based on the paucity and impracticality of studies performed in the horse, as well as the contradictory nature of results in comparison to humans, the effects of antibiotic administra tion on equine gastrointestinal flora and SCFA profiles are yet to be examined.

PAGE 108

82 Table 3-4. Summary of single dose antibiotic effects on the equine cecal microenvironment170 Drug Route Effect on Microflora* Effect on pH* Effect on SCFA* Penicillin G IV ----(lactic) Penicillin G Oral coliforms, streptococci, Clostridium spp. (lactic), (butyric, propionic) Ampicillin IV to --coliforms --(lactic) Ampicillin Oral coliforms, streptococci, lactobacilli, Clostridium spp. --(lactic), (propionic) Amikacin IV ------Amikacin Oral --(lactic) Oxytetracycline IV ----(lactic) Oxytetracycline Oral coliforms, streptococci, lactobacilli, Clostridium spp. --(lactic) = strong suppression / decrease, = moderate suppression / decrease, = minimal suppression / decrease, --= no suppression / decrease, = increase Current Theory on the Pathogenesis of Antibiotic-Associated Diarrhea (AAD) Over the last 15 years, significant progress has been made in determining the risk factors, pathogenic mechanisms, therapies, and predictive outcomes regarding AAD. Focus has been primarily on the action or interaction of the antibiotic with the gastrointestinal flora and the repercussions of disturbing that ecosystem. Figure 3-1 shows a current summary assumption of the path ogenesis and progression of AAD. It is interesting to note that many of the proce sses involved in this pathogenesis are selfperpetuating, with the generation of cascad ing and cyclic effects on the entire host organism.

PAGE 109

83 Figure 3-1. Pathogenesis of antibiotic-asso ciated diarrhea. Adapted from BergogneBrzin, 2000.79 DIARRHEA colonic absorption of water and sodium osmotic draw of water into lumen Chloride secretion Toxin elaboration Host neuroinflammatory response (e.g. neutrophil influx, mucin production, hypermotility) pathogen or opportunist growth (e.g. C. difficile Salmonella spp., Yeast facultative aerobic Gram+ organisms Impaired or abnormal carbohydrate fermentation and disproportionate SCFA synthesis ANTIBIOTIC THERAPY Disruption of microflora associated function Disruption of autochthonous flora Loss of “Colonization Resistance” obligate anaerobes pH Pharmacologic effects of drug on intestinal motility (e.g. macrolides)

PAGE 110

84 Effects of SCFA on Anaerobic Growth of Bacteria It has been thought that only the undisso ciated form of an organic acid is responsible for initiating an antimicrobial effect, by virtue of its ability to cross lipid membranes. Recent studies however, have show n that using the sodium salts of the acids have also demonstrated antimicrobial activit y, and it is now known that both dissociated and undissociated lipophilic aci ds and bases can cross cell membranes according to concentration gradients, though the undissoci ated form is still the preferred route.175 The sodium salts of SCFAs are nearly completely dissociated in solution. This dissociation behavior allows for an easily predictable am ount of SCFA availabl e to cross into the cytoplasm regardless of differences in local pH. Physiologically, as pH decreases, the amount of undissociated acid in a solution in creases, and it is this undissociated (nonionized) form (HA H+ + A-) of weak acids that is thought best able to penetrate the cell wall.176 Once inside in the more alkaline cytoplasm of a cell, the acid again dissociates and there is a gr adual accumulation of protons (H+) and anions (A-) within the cell, above what can be re-associated to intracellular cations such as potassium (K+) and shuttled back out across the membrane.175 Acid Tolerance Response of Salmonella and Other Enterobacteriaceae “What does not kill you makes you str onger” describes the ability of Salmonella to augment their resistance to organic acids or acidic pH after exposure to low pH alone177 or in combination with individual SCFA.178 This is termed th e acid tolerance response (ATR) of Salmonella which is evident in two distinct circumstances, while the cells are growing exponentially or while they are at sta tionary phase. It was also shown that this adapted acid resistance was further enhanced by increasing the length of exposure time to the SCFA, anaerobiasis, and acidic pH in conjunction with SCFA exposure.178 Increased

PAGE 111

85 protein synthesis was shown to be the basi c mechanism responsible for the ATR, and over 50 distinct proteins have been termed acid shock proteins ba sed on their synthesis response to low pH shock.179 This ATR could have implications for the survival of Salmonella in the stomach, intestinal lumen, tr eated foodstuffs, silage, or in the phagosome of the macrophage. It was al so shown that induction of the ATR in Salmonella indirectly confers protection to severa l other unrelated stressors, such as high osmolarity (2.5M NaCl) and reactive oxygen species (20mM H2O2).180 It has been demonstrated that concentrate feeding in cattle decreased large intestine pH and increased the relative concentrations of SCFAs, wh ich induced an acid-resistant phenotype in the native E. coli (rather than selection for an acid-resistant subpopulation).181 This phenomenon could easily apply to the intestinal microflora and/or Salmonella in the case of the domestic horse. Route of infection appears to be an impor tant determinant of pathogen survival and ability to invade through the ga strointestinal barrier in salmonella infections, and may be associated with exposure to low pH. When S Dublin is given orally26;182 to calves, or S. Typhimurium is given orally to mice89 or horses,94;96 typical courses of disease result. Nicpon et al. conducted experiments utilizing chronically fistulated horses, where they were given large (100 ml of 108 cells per ml) challenge doses of virulent S. Typhimurium or S Enteriditis orally or directly into the cecum.183 Only the horses given Salmonella via the oral route developed a ny clinical signs such as feve r, diarrhea, and colic. The horses given Salmonella via cecal fistula (two doses pe r horse) did not develop any signs of disease. This effect coul d be attributed to the ability of the host to overwhelm and

PAGE 112

86 dispatch the minimally prepared and defended organism, or could be due to completely unrelated virulence factors. SCFA Effects on Salmonella Growth and Invasion “SCFA toxicity” is a term used to descri be the bacteriostatic and bacteriocidal effects of SCFA on bacteria. This mechan ism has historically been explained by an uncoupling theory, where an abnormally low luminal pH allows undissociated acid to pass through the cell membrane, become dissociat ed in the more alkaline interior of the cell, and abolish the proton-motive force acro ss the cell membrane, thereby acidifying the cell and halting intracellular metabolism and pr otein synthesis. This theory does not adequately explain why many pathogenic a nd commensal cells are able to survive, continue to function or even flourish in the slightly acidic and SCFA rich environment of the mammalian large intestine. A newer theory has been proposed that can better explain the fermentation acid-resistant phenotype that some bacteria innate ly possess or develop with exposure to SCFA. This theory al so explains the phenomenon of decreased epithelial cell association and decreased growth rates of Salmonella in low pH or SCFA containing culture media. pH Gradient-M ediated Anion Accumulation was defined in 1998 and suggests that SCFA resistant microbes have a reduced pH gradient across their cytoplasmic membrane ( pH) and higher intracellular con centrations of potassium ions to combine with the accumulating acidic anions.175 Some more notable gram-negative bacteria such as E. coli O157:H7 have also refine d their ability to decrease pH, and are notorious foodborne pathogens.175 Acetate and formate tend to promote the invasive phenotype in different salmonella serovars, whereas butyrate and pr opionate tend to inhibit invasion.184;185 Acetate is able

PAGE 113

87 to cross bacterial cell wall membranes in th e undissociated (non-ionized) form, whereas butyrate and propionate cannot.186 It has therefore been sugges ted that acetate can initiate different cellular signals compar ed to butyrate or propionate.185 In vitro growth of Salmonella supplemented with biologically relevant SCFAs in physiologic and aphysiologic con centrations has been studied by several investigators. Unfortunately, most of this work has been done under aerobic conditions, and minimal data are available examining the effects of SCFAs under anaerobic growth conditions. The distal intestinal environment is comp letely anaerobic, and bacteria undergoing anaerobic respiration may respond very different ly than in the presence of oxygen. pH is a critical factor in determining whether or not a SCFA at a particular concentration has an effect on salmonella growth. Durant et al. showed that under aerobic conditions at pH 6, growth rates of S. Typhimurium were decreased in acetate, butyrate, and propionate containing media (25, 50 and 100mM), while no differences were observed when the pH was raised to 7.185 SCFAs are weak acids with pKa’s averagi ng 4.8 and according to the local pH will be present in either ionized or non-ioni zed forms, which should determine their absorption kinetics across biological membra nes. Recently, the pH of the intestinal lumen has been shown not to influence the absorption of SCFAs in the guinea pig and human colon, nor in guinea pig and rat cecum models. This segregation of what was thought to be a dependent interaction is likel y due to the neutral pH environment present at the intestinal mucosal surface.75 The mucosal pH rarely strays from neutral and is independent of changes in luminal pH, which c ould explain why bacteria that are able to quickly attach and invade epithelial cells (e.g., Salmonella and Shigella ) are highly

PAGE 114

88 pathogenic. These bacteria do not solely rely on complex and energy-taxing mechanisms of environmental and host re sistance, but simply bypass th e bacterial gauntlet in the lumen to hide within the cells. SCFA Effects on Expression of spv Genes in vitro It has been demonstrated that stati onary phase growth, heat-shock, nutrient starvation, and energy source depl etion are potent inductors of spv expression in Salmonella .28;132;187;188 It was recently shown that th e specific SCFA s acetate [C2], propionate [C3], butyrate [C4], valerate [C5], and caproate [C6] induce increased in vitro expression of spvR and spvB in S. Dublin during log-phase growth at neutral pH.189 This is important in determining role of the spv genes in gastrointestinal pathogenesis, as they are not currently universally accepted to have an ente ric-related function. The spv genes have been shown to influen ce the severity of enteritis in a host-adapted model of S. Dublin infection in the calf, but this is the only supporting evidence. Examining the response at a lower physiologic pH may be more relevant. SCFAs and Salmonella Colonization and Infection of Avian Species A significant amount of research has been performed with re gard to preventing salmonella colonization and infection in avia n species such as chickens, turkeys, and ducks. Death from infection or salmonella contamination of the eggs or marketable carcass can result in significant lost annual revenue. SCFAs have been studied and used extensively as food additives and therapeuti c modalities to minimize the presence of Salmonella in layer flocks and hatcheries.90;190;191 Challenge experiments using broiler chicks colonized with anaerobic cecal flora fr om adults on day of hatch showed that significantly fewer Salmonella were isolated from the digest ive tracts of those chicks that were fed 10% lactose as a feed additive. T hose chicks also had significantly higher cecal

PAGE 115

89 concentrations of acetic and pr opionic acids, with lower pHs than controls fed no lactose or a lower concentration.190 Forced molting induced by feed deprivation in chickens was found to decrease the crop concentrations of acetate, lactate, butyrate and propionate, and significantly sensitize those animals to cr op and cecal colonization, with subsequent systemic spread after oral challenge with S. Enteriditis.43 This situation is similar to what is thought to occur in the gast rointestinal tract of the horse after acute dietary changes. Direct feeding of either carbohydrate or specific SCFAs in avian species has been shown to be protective agai nst salmonella attachment in vivo90 and survival in vitro .192 Diez-Gonzalez et al. demonstrated that increasing amounts of grain fed to cattle significantly increased the SCFA concentratio ns in the colon approximately fourfold, while rumen concentrations remained unchanged.181 Annison et al. re ported that feeding acetylated, butyrylated, or propionylated starches to rats, preferentially raised the colonic concentrations of those SCFA.193 Altogether, these results predict tremendous potential for more natural methods of dietary modifica tion to protect against salmonella infection in susceptible livestock and companion animal species. SCFAs and Salmonella in Swine Salmonella colonization of swine and contamination of meat is an important human health consideration worldwide. The a ssociation between SCFA s and resistance to colonization has been studied extensively in th e porcine as well as the avian food species. Dietary influence on intestinal pH was show n to be the most important factor in mediating the SCFA-attributed anti-bacterial effects of colon contents in swine at slaughter.194 This effect was thought due to prot onation of the SCFAs into non-ionized forms, as there were no differences in the to tal levels or relative proportions of SCFAs from two herds with otherwise signifi cantly different shedding levels of Salmonella The

PAGE 116

90 diets were found to contribute significantly to acid-base excretion in the urine as well as colon contents. SCFAs and Salmonella Colonization and Infection of Bovine Species The calf model of enteric and systemic sa lmonellosis has surfac ed as a reasonable alternative to in vitro or murine models currently availa ble. One advantage to this model is the fact that the bovine species is a ffected by both host-adapted (Dublin) and broadhost range (e.g., Typhimurium, Anatum, Newport) salmonella serovars. Oral infection of calves with serovar Typhimurium shows the most clinical similarity to salmonella colitis in horses and humans. The spect rum of serovars that affect cattle clinically is very similar to horses (Typhimurium, Typhi murium var. Copenhagen, Anatum, and Newport).195 Rumen SCFA mixtures were shown ma ny years ago to be inhibitory to S Typhimurium growth in the presence of a low pH.196;197 However, a survey of the literature from the dairy and beef industries shows they have minimally advanced their understanding of the effects of SCFA on pathogen growth and su rvival. Rather they have examined them in an attempt to refi ne dietary manipulation of lactation198 or feed efficiency.199 Specific Aims The specific aims of this se gment of the study were to: Determine the cecal SCFA concentrations luminal pH, total culturable anaerobic bacterial counts, and protozoal counts of horses before and after treatment with selected antibiotics. To examine spv + and spv salmonella isolates in te rms of growth rate during anaerobic culture in nutrient broth supplemented with sterile-filtered cecal contents from antibiotic-treated ve rsus non-treated horses.

PAGE 117

91 To examine spv + and spv salmonella isolates in te rms of growth rate during anaerobic culture in nutrient broth adjust ed to the mean luminal cecal pH and supplemented with individual SCFAs nor mally found in horse cecal liquor. Materials and Methods IACUC Approval All protocols, procedures and experiments utilizing experimental animals described in this work were approved by the University of Florida Institutional Animal Care and Use Committee (IACUC) prior to co mmencement of any experiments. Subject Coding for Experiments and Data Analysis Coded reference to individual horses and an tibiotic treatments in this document are according to the following legends (Tables 3-5 and 3-6). Table 3-5. Coding legend fo r experimental animals Horse Letter Code Number Code Bill B 1 Easy E 2 Scott S 3 Ted T 4 Willie W 5 Surgical Placement of Cecal Cannula in the Horse A summary of the model and surgical t echnique using large-bore reinforcedsilicone cecal cannulas in horses is included in this document as it has not previously been described in publication. Surgical impl antation of a cecal ca nnula via laparotomy was performed at least one year prior to inclusion in this study. Five thoroughbred geldings, age range 4-15 years were in cluded in the study. The cecum contained a custom, double-flanged, reinforced -silicone cannula (Figure 3-2) surgically placed in the lateral cecal body 30 cm dorsal to the apex and exteriorized through the lower right abdominal wall with 8-12 inches of the cannula visible outsi de of the horse.

PAGE 118

92 Figure 3-2. Components of indw elling cecal cannula placed into experimental horses. Clockwise from the top: side view of sliding silicone flange placed on the lateral serosal aspect of cecal wall, cannula with fixed silicone interior flange, silicone filled, thick walled PVC tubing used to plug cannula, front view of sliding silicone flange, hose clamp to secure plug within cannula. The cannula tubing has an outside diameter of 7/8” to 1” and an internal diameter of 5/8” to 1”, respectively. The cannula was sealed with a tight-fitting silicone plug and secured with a hose clamp to prevent exposure to air, excep t during collection procedures which were kept to the absolute minimum ti me necessary. This basic preparation was extremely well tolerated by the horses and ha s been used by our laboratory with great success for many years as a humane and eff ective method of repeatedly sampling the cecal lumen with no discomfort to the animal Photos of an experimental horse and cannula preparation are shown in Figures 33 and 3-4. These preparations are not permanent, as time has shown the cannula is slow ly expelled via the formation of internal granulation tissue, and the creation of a te mporary ceco-cutaneous fistula, which soon closes by second intention.

PAGE 119

93 Figure 3-3. Experimental horse E (2) w ith cecal cannula 3 years post-implantation. Figure 3-4. Close-up view of cannula in situ in experimental horse E (2). Note the formation of a firm swelling intimately associated with the cannula insertion. This is internal granulation tissue form ing around the interior silicone flanges which will result in the eventu al expulsion of the device.

PAGE 120

94 Antibiotic Treatment of Horses Each horse received one course of a ll 4 experimental treatments with no replication, using a randomized block design to minimize effects of treatment order or between horse differences. The horses were weighed once before each treatment to accurately calculate dose rates. Three antibi otics were chosen based on frequency of use in equine medicine, reported relationship re garding antibiotic-asso ciated diarrhea, and typical route of administration. The antibio tic treatments are detailed in Table 3-6. Table 3-6. Antibiotic treatments of horses Treatment Treatment Code Dose Route of Administration Dosing Interval Control (no treatment) 1 CON ------Ceftiofur sodium (Naxcel) 2 NAX 2 mg/kg IM q 12 h Oxytetracycline (LA200) 3 TET 10 mg/kg IV (Diluted into 1 liter NaCl) q 24 h TrimethoprimSulfamethoxazole 4 TMPS 30 mg/kg PO q 12 h Ceftiofur sodium is a third-generation ce phalosporin with an intermediate spectrum of activity against gram-positive and gram-negative aerobes and some activity against anaerobes. Oxytetracycline is a broad-spec trum agent effective against gram-positive and gram-negative aerobes and anaerobes. Trimethoprim-sulfamethoxazole is a reasonably narrow spectrum potentiated sulfonamide effective against gram-positive and gram-negative aerobes only.200 Each antibiotic course was administered fo r a total of four da ys twice daily for treatments 2 and 4, and once daily for treatment 3. Four day therapy regimens were decided based on the average amount of time it took for previous investigators to notice derangements in fecal SCFA49;82-85 and culturable anaerobic fecal flora.49;86 There was a

PAGE 121

95 minimum of 30 days washout between each trea tment for an individual animal, typically between 30 to 45 days. Equine Cecal Sampling Procedure Horses were loosely restrained with a halter and lead rope in a stall or in the pasture for sample collection. For each antibiotic tr eatment, cecal contents were collected and processed within 24 hours prior to commenci ng therapy and again after three consecutive days of treatment. The time of concentr ate feeding and cecal sampling was consistent each day, in order to minimize normal temporal and dietary influences on the cecal SCFA profile. Several aliquots were quic kly collected directly in to polypropylene specimen containers via gravity flow (F igure 3-5). The first 100 – 200 ml of contents was allowed to drain before the collection was started to avoid colleting any liquor that had been stagnant in the lumen of the cannula. If the cecal contents would not flow easily or quickly, a 36” stainless steel Chambers cath eter was inserted to facilitate sample collection. The sample collecti on containers were f illed to overflow and capped airtight in order to minimize oxygen introduction into the samples. These were immediately transported to the laboratory in a warm wa ter bath. Sample collection via this method attempts to preserve the warm and anaer obic nature of the horse’s cecum and the collected sample as much as possible. I ndividual portions were immediately separated and placed inside an anaerobic chamber for se rial dilutions and quantitative culture, kept on the bench for pH measurement, or processe d for SCFA analysis. Processing for SCFA analysis involved centrifugation of an 80 ml aliquot of raw cecal contents at 10,000 x g and 4C for 20 min. The supernatant was sterile-filtered thro ugh a 0.2 micrometer syringe filter into sterile containers, and the f iltered samples were frozen at -80C for use

PAGE 122

96 as experimental additives to salmonella cu ltures and SCFA quantification later in the course of the study. Figure 3-5. Collection of equine cecal contents from indwe lling silicone cannula. Note the rapid flow and liquid nature of the contents. Physical Effects on the Horse Horses were maintained on grass pasture with twice daily concentrate feeding and fresh water was available at all times. Subj ective observations of the experimental horses were made during the treatment periods as we ll as during washout peri ods to ensure their well being as well as to monitor for any po ssible long-term or latent effects of the treatments. Appetite, fecal character a nd consistency, demeanor, and body condition were observed on at least a weekly basis by investigators, and daily by caregivers.

PAGE 123

97 Effects on Fecal Consistency Throughout the entire study and specifically during the antibioti c treatment periods, feces were observed for each individual on a daily basis (when avai lable) as the horses were being handled. Effects on Cecal Content Character The effect of antibiotic treatment on cecal digesta composition and consistency was evaluated in a subjective manner by one inves tigator (non-blinded) and was described on the treatment sheets at the time of collection of the preor post-treatment sample. Odor, water and fiber content, and color were all described. Equine Cecal Anaerobe Quantification Enumeration of anaerobic bacteria in equine cecal conten ts was carried out according to the method described by Mackie and Wilkins (1988).166 All sample manipulation was performed inside an an aerobic, climate controlled, combination chamber and incubator with a 10% CO2 85% N2 5% H2 atmosphere (Bactron™ 1.5, Sheldon Manufacturing, Cornelius, OR). Sealed aliquots of the fres hly collected cecal contents were placed into the chamber as soon as possible after collection (within 15 min). Each sample was serially diluted from 10-1 to 10-10 in pre-reduced 0.9% sterile phosphate-buffered saline (PBS) and plated on to commercially manufactured pre-reduced anaerobically sterilized (PRAS) media, Br ucella blood agar cont aining 75 micrograms/ml of gentamicin sulfate (AS-141G, Anaerobe Syst ems, San Jose, CA). CFU/ml of cecal contents was estimated according to the Mi les-Misra technique for quantification of viable bacteria. A 20 microliter drop of each dilution was applied to the plate surface, allowed to dry, and incubated within the anae robic chamber with plates inverted at 37 C. Samples were replicate plated four times for average estimation of CFU/ml. Colony

PAGE 124

98 counts of anaerobic flora were performed af ter one and four days of incubation and reported in CFU/ml of cecal contents. No a ttempt was made to identify or classify the bacteria, only quantify. pH Analysis of Equine Cecal Contents Aliquots of the freshly collected cecal cont ents were measured for pH as soon as possible after collection (within 15 min). The mean of three different measurements was taken for the sample value. pH was measur ed using a digital desk top pH meter (Corning Inc. Life Sciences, Acton, MA) which was cal ibrated using two standard buffers (pH 7.0 and pH 10.0) before each sample measurement. The lowest mean postprandial cecal pH values in horses were obtained approximately 4-7 h after a meal168 and sampling time in these experimental horses was adjust ed to occur with in this window. Short-Chain Fatty Acid Analysis of Equine Cecal Contents Samples of the ultra-centrifuged and filtere d cecal supernatant were frozen at -80C until analysis. Thawed supernatant SCFA composition was measured using capillary glass chromatography (Autosystem II, Perk in Elmer, Boston, MA) with splitless automatic injection onto a Nukol Fused S ilica Capillary Colum n, 30 mm x 0.25 mm ID (Supelco Chromatography, Bellefonte, PA) and us ing helium as the carrier gas. SCFAs were detected using a flame ionization detect or and peaks were integrated and compared with external standards using the Turbochrom e 3 integration comput er software (Perkin Elmer, Boston, MA). Previous experiments w ith cecal contents from the same horses had shown that optimum conditions for separation we re obtained with the temperatures set at 250C for the injector, 160C for the oven, and 180C for the detector. Acetate, propionate, butyrate, isobutyrat e, valerate, isovalerate, and ethyl butyrate were identified.

PAGE 125

99 Protozoal Quantification of Cecal Contents from Horses Treated with Antibiotics Protozoal counts were performed accord ing to the method described by Adam.201 A 1 ml sample of freshly collected cecal co ntents was added to a 2 ml microcentrifuge tube containing 1 ml of buffered 10% formalin. The samples were stored at 2-8C until analysis. After thawing, 500 mi croliters of the formalin diluted sample was added to 150 microliters of 0.5% methyl green in 7% acetic acid solution. The sample was vortexed and added to the counting well of a McMaster slide. A 0.15 ml aliquot of diluted sample was examined for the presence of protozoa. Counts were normalized to number of protozoa per ml of cecal contents, and stat istically compared across horses, times, and treatments. In vitro Short-Chain Fatty Acid Growth Comparison Three different SCFAs were compared for their in vitro ability to affect growth rates of Salmonella in an anaerobic environment. A cetate, butyrate, and propionate were chosen based on their natural predominance in the mammalian cecum and large intestine. Inhibitory as well as stimulatory growth was investigated using a nutritionally robust (LB) or minimal media (M9) respectively. Th e sodium salts of acetic acid, butyric acid or propionic acid (Sigma-Aldrich, St. Louis, MO) were added to pre-reduced broth media in an anaerobic chamber at both 30mM and 100mM concentrations. The pH was adjusted to 6.5 in all experimental solutions equivalent to the m ean cecal pH of the untreated experimental horses. NaCl at 30mM and 100mM was added to control tubes as an isosmolar equivalent. 3306 was compared with 3337 for all experiments to examine the effects of virulence plasmid and spv gene presence on growth-rate. Standing overnight cultures of each isolate we re diluted to approximately 1 x 10-3 to 1 x 10-5 in the experimental tubes, and the cells were grown for 10-12 h. Aliquots were taken at time 0,

PAGE 126

100 2 h, 4 h, 6 h, 8 h, 10 h, and 12 h and serially diluted and plated on LB agar to enumerate CFU/ml. Experiments were repeated to valida te results. Growth curves were produced from these data and compared statistically with SCFA and concentration as factors. Since the two S. Typhimurium strains exhibited n early identical shape in their respective growth curve respons es (regardless of initial b acterial concentrations, which may have been different) to the control and s hort-chain fatty acid so lutions, the data were combined and they were treated as replicat es instead of individual experiments. All experimental solutions were prepared to pH 6.5, which approximates the measured luminal pH of the cecum in horses.1 The concentrations chosen for the experime ntal SCFA solutions in this study were above the normal physiologic range for mammalia n large intestine. They were chosen based on similar studies in other species a nd genera, as well as the food protection industry.158;175;184;185;202 This was decided in order to observe (in the smallest number of experiments possible) whethe r an effect on anaerobic gr owth rates existed for the compounds. Continued examination with ti tration down towards the physiologic range would yield the in vitro breakpoint inhibitory concen tration, and this could be supplemented with information obtained from growth in a variety of other media— including raw intestinal contents spiked with SCFA. Newer technologies allow quantification of specific bacteria with in a heterogeneous bacterial suspension.203 In vitro Effects of Cecal Liquor from Antibiotic-treated Horses on Anaerobic Growth of Salmonella The effect of adding filter-sterilized cecal contents, from horses treated with selected antibiotics as compared to un treated controls on anaerobic growth of Salmonella 1 Data from this work estimated the mean lu minal pH in the cecum of horses to be 6.6

PAGE 127

101 in vitro, was performed according to a modifica tion of the method described in mice by Que.80 Cecal contents were collected fr om horses and processed as described previously—24 h prior to initiation of antibio tic treatment and at the conclusion. The frozen aliquots of cecal liquor were thawed at 4C overnight and either pooled per treatment group or added individually at 10% (v/v) concentration to both pre-reduced complete M9 with glucose or pre-reduced LB broth in an anaerobic chamber (Bactron™ 1.5, Sheldon Manufacturing, Cornelius, OR). St atic overnight cultures of salmonella strains 3306 and 3337 were added to the experiment al tubes at starting dilutions ranging from 1 x 10-3 to1 x 10-5. Aliquots were taken at time 0, 2 h, 4 h, 6 h, 8 h, 10 h, and 12 h, serially diluted in PBS, and plated to enumerate CFU/ml. The filter-sterilized cecal contents from the five experimental horses were pooled by treatment for the preliminary experiments and then examined by individual horse at the conclusion. The growth of S. Typhimurium strain 3306 was also compared to 3337 to examine if any effect of the virulence plasmid with spv genes could be determined. Growth was examined in both a nutritionally rich medium and M9 minimal medium with glucose to evaluate the possibility of inhi bitory as well as trophic effects. The filter-sterilized cecal contents from antibiotic-treated horses we re added in 10% concentrations for all experiments. Growth curves were produced from these data and co mpared statistically, with horse and treatment as factors. Ba sed on preliminary results from both pooled additive experiments, it was decided to do th e remaining individual horse experiments in nutritionally limited M9 minimal medium to examine for potential trophic effects of the additives.

PAGE 128

102 pH was measured of the stored liquor sa mples before the addition experiments and was found to correlate with original measured pH values. It was decided that these values would not be nor malized prior to the in vitro experiments, but remain in their original state to more closely approximat e physiologic conditions. If an effect was determined from the unadulterated samples, th e experiment would be repeated with the pH values neutralized to determine if the effect was due to pH alone. Statistical Methods Data from the total culturable anaerobe counts, pH of cecal contents, and protozoal counts were analyzed non-parametrically using Wilcoxons Sign Rank Test for two related values in a commercial statistical analysis program (SPSS for Windows version 11.0, SPSS, Chicago, IL). Logarithmic transfor mations were performed on all bacterial count data before analysis. The cecal flui d addition and SCFA addition experiments were analyzed using analysis of va riance with repeated measures. Significance levels were set at p=0.05 for all experimental data. Results Effects on the Horse All animals successfully completed all tr eatments during the course of the study, with no complications or observed side-effect s. Fecal consistency remained normal for all horses during the treatment periods, and there was no apparent softening (to “cow pie” consistency) or loss of shape from the normal fecal ball appearance. There was no observable effect to unbiased individuals of antibiotic treatment or placebo on fecal character or consistency for a ny horse during any experiment. No changes were reported by caretakers regarding the horses’ appetite during the antibiotic trial periods. Body weight and condition scores remained constant throughout the trial period.

PAGE 129

103 Cecal contents tended to be thicker (oatmeal consistency), bright green (depending on the amount of pasture availabl e at the time of the year), with an increased visible fiber content, before any of the treatments (i.e. first treatment pre-treatment sample for each horse). It was noted that this normally thic k consistency tended to become more watery, darker colored tending toward brown, with less particulate matter, after an experimental treatment, especially the oxyt etracycline treatment. The odor of the contents also changed, from the characteristic sharp, acrid odor with a hint of background sweetness, to a distinctly unpleasant or foul smell. Though the fact that a horse was being treated with antibiotics vs. the control treatment (for all three drugs tested) could be determined by observing the cecal contents, it was not discer nable by evaluating the horse, its demeanor, or its feces. Effects on Cecal pH The baseline (pre-treatment) pH values measured for equine cecal contents were in agreement with other inve stigators (mean of 6.6).155;166;170;204 There was no statistically significant effect of treatment on pH of the cecal liquor for any of the antibiotics studied as compared to the control treatment. Data fr om all horses and all treatments with means is summarized in Tables 3-7 through 3-10. An increase in cecal liquor pH was only seen with the oxytetracycline treatment and it was not significantly different than the baseline range. Measured pH values during treatment periods were also within the published ranges of normal horses.

PAGE 130

104 Table 3-7. Cecal liquor pH of cannulated horses before and after 4 days of control (no) antibiotic treatment HORSE PRE POST CHANGE MEAN PRE 1 6.10 6.08 -0.02 6.34 2 6.26 6.09 -0.18 MEAN POST 3 6.17 5.74 -0.43 6.10 4 6.21 6.11 -0.10 MEAN CHANGE 5 6.94 6.49 -0.45 -0.24 Table 3-8. Cecal liquor pH of cannulated ho rses treated with intramuscular ceftiofur sodium at 2 mg/kg twice daily, before and after 4 days of treatment HORSE PRE POST CHANGE MEAN PRE 1 6.56 6.68 0.12 6.74 2 6.56 7.00 0.44 MEAN POST 3 6.65 6.61 -0.05 6.84 4 6.80 6.74 -0.06 MEAN CHANGE 5 7.13 7.15 0.02 0.09 Table 3-9. Cecal liquor pH of cannulated hors es treated with intravenous oxytetracycline at 10 mg/kg once daily, before a nd after 4 days of treatment HORSE PRE POST CHANGE MEAN PRE 1 6.33 6.40 0.07 6.61 2 6.74 7.15 0.41 MEAN POST 3 6.86 7.07 0.22 6.92 4 6.41 6.98 0.57 MEAN CHANGE 5 6.74 7.00 0.26 0.31 Table 3-10. Cecal liquor pH of cannulated horses treated with oral trimethoprimsulfamethoxazole at 30 mg/kg twice da ily, before and after 4 days of treatment HORSE PRE POST CHANGE MEAN PRE 1 6.86 6.37 -0.49 6.73 2 6.56 6.17 -0.39 MEAN POST 3 6.31 5.97 -0.34 6.39 4 7.47 6.82 -0.65 MEAN CHANGE 5 6.43 6.63 0.21 -0.33 Effects on Cecal Protozoal Counts Results are shown in Tables 3-11 through 3-14. In each experiment there appears to be one outlier in terms of pre-treatment protozoal counts or response to the treatments, though it was not always the same horse.

PAGE 131

105 Table 3-11. Total protozoal counts per ml of cecal contents from cannulated horses before and after 4 days of control (no) antibiotic treatment Horse PRE POST Absolute Change % Change 1 966 2984 2018 208.9 2 1666 884 -782 -46.9 3 1634 516 -1118 -68.4 4 2066 2250 184 8.9 5 1334 8100 6766 507.2 MEAN 1533 2947 1414 Table 3-12. Total protozoal counts per ml of cecal contents from cannulated horses treated with intramuscular ceftiofur sodi um at 2 mg/kg twice daily, before and after 4 days of treatment Horse PRE POST Absolute Change % Change 1 7266 4766 -2500 -34.4 2 1534 4966 3432 223.7 3 1616 4534 2918 180.6 4 7884 4916 -2968 -37.7 5 11950 9500 -2450 -20.5 MEAN 6050 5736 314 Table 3-13. Total protozoal counts per ml of cecal contents from cannulated horses treated with intravenous oxytetracyclin e at 10 mg/kg once daily, before and after 4 days of treatment Horse PRE POST Absolute Change % Change 1 850 1966 1116 131.3 2 2584 3400 816 31.6 3 8834 10934 2100 23.8 4 716 6566 5850 817.0 5 10150 6934 -3216 -31.7 MEAN 4627 5960 1333 Table 3-14. Total protozoal counts per ml of cecal contents from cannulated horses treated with oral trimethoprim-sulfa methoxazole at 30 mg/kg twice daily, before and after 4 days of treatment Horse PRE POST Absolute Change % Change 1 3884 2800 -1084 -27.9 2 616 5684 5068 822.7 3 3838 1900 -1938 -50.5 4 4700 2684 -2016 -42.9 5 8566 13650 5084 59.4 MEAN 4321 5344 1023

PAGE 132

106 Both the percent change and the absolute change were examined to account for initial between horse differences in the numbers of protozoa. There were no statistically significant differences in either the absolu te means or the percentage change means compared to the control treatment for a ny antibiotic treatment (p=0.500 to 0.893). One outlying individual was removed from the analys es and the results ( no significant effect of treatments) were similar. Values obtai ned for pre and post-treatment means were all within the range of published values fo r equine cecal protozoal organisms. Effects on Cecal SCFA Quantities and Proportions The effects of antibiotic treatment on equi ne cecal SCFA profile s are detailed in Tables 3-15 through 3-18. There were no si gnificant differences in the absolute concentration of total SCFAs measured for a ny of the antibiotic treatments (as compared to the control treatment), but there were sign ificant differences in specific individual SCFA concentrations for all three antibiotic treatments. The individual SCFAs were examined using the percent change from pretreatment values compared to the control treatment. Several individual acids for different treatments were significant or approached significance. Fo r the ceftiofur treatment—ace tate decreased p=0.080, for the trimethoprim-sulfamethoxazole treatment— acetate increased p=0.080 and isovalerate decreased p=0.043, and for the oxytetracyclin e treatment—propionate increased p=0.043, isobutyrate increased p=0.043, butyrate in creased p=0.043, isovalerate increased p=0.043, and valerate increased p=0.080. To tal SCFA percentage change values approached significance in the ceftiofur (p=0.080) and trimethoprim-sulfamethoxazole (p=0.080) groups when compared to th e control group percentage change.

PAGE 133

107Table 3-15. Cecal liquor concentrations of individual and total SCFAs from cannulated horses, before and after 4 days of control (no) treatment Horse Treatment Time A P IB B IVA V EB TOTAL Concentration in mM 1 1 pre 5.04 6.67 9.94 7.85 2.57 8.01 5.96 46.04 1 1 post 6.30 5.87 9.03 8.73 3.32 10.15 9.33 52.72 2 1 pre 4.61 6.34 6.19 15.86 2.85 11.27 7.21 54.34 2 1 post 4.44 5.97 6.39 13.72 3.64 8.81 10.04 53.01 3 1 pre 3.47 1.35 4.18 3.41 2.53 2.27 11.64 28.85 3 1 post 4.30 2.70 6.13 6.18 3.24 3.86 10.78 37.19 4 1 pre 3.84 2.15 4.49 4.82 3.02 2.93 9.70 30.95 4 1 post 5.77 4.16 7.42 7.73 4.46 4.63 8.61 42.78 5 1 pre 3.20 1.63 4.55 3.41 3.22 2.53 8.27 26.81 5 1 post 4.04 2.72 5.57 5.46 3.33 4.14 9.31 34.57 MEAN PRE 4.03 4.03 3.63 5.87 7.07 2.84 5.40 8.55 MEAN POST 4.97 4.97 4.28 6.91 8.36 3.60 6.32 9.61 % Change of Mean 23.17% 18.06% 17.68% 18.29% 26.84% 16.97% 12.39% 17.81% Mean of % Changes 24.35% 48.51% 25.68% 39.89% 27.21% 39.32% 17.95% 21.64% A = acetate, P = propionate, IB = isobut yrate, B = butyrate, IVA = isovalerate V = valerate, EB = ethyl butyrate

PAGE 134

108Table 3-16. Cecal liquor concentrations of individual and total SCFAs from cannulated horses treated with intramuscular ceftio fur sodium at 2 mg/kg twice daily, before and after 4 days of treatment Horse Treatment Time A P IB B IVA V EB TOTAL Concentration in mM 1 2 pre 6.67 5.62 8.39 12.55 3.72 11.41 9.30 57.66 1 2 post 6.11 4.39 6.70 11.14 3.14 7.99 8.62 48.09 2 2 pre 4.75 7.52 6.40 15.07 2.99 9.88 17.09 63.69 2 2 post 4.34 5.84 5.94 12.08 3.69 8.35 7.12 47.36 3 2 pre 4.83 3.01 6.83 6.04 4.10 4.00 7.02 35.84 3 2 post 4.85 5.09 8.31 10.65 4.50 6.40 7.91 47.70 4 2 pre 4.38 5.60 7.22 10.96 3.83 6.49 8.49 46.98 4 2 post 4.84 7.15 8.10 13.88 5.04 8.11 9.47 56.58 5 2 pre 4.48 4.73 7.40 9.56 4.96 5.79 9.60 46.51 5 2 post 4.51 4.49 6.93 6.91 3.59 4.08 8.81 39.32 MEAN PRE 5.02 5.02 5.30 7.25 10.83 3.92 7.51 10.30 MEAN POST 4.93 4.93 5.39 7.19 10.93 3.99 6.99 8.39 % Change of Mean 1.84% 1.81% 0.75% 0.89% 1.80% 7.02% 18.57% 4.63% Mean of % Changes 2.40% 28.34% 6.93% 33.64% 11.11% 21.64% 7.35% 0.818% p value (compared to control) 0.080 0.686 0.345 0.893 0.225 0.345 0.686 0.080 A = acetate, P = propionate, IB = isobut yrate, B = butyrate, IVA = isovalerate V = valerate, EB = ethyl butyrate

PAGE 135

109Table 3-17. Cecal liquor concentrations of individual and total SCFAs from cannul ated horses treated with intravenous oxytetracycline at 10 mg/kg once daily, be fore and after 4 days of treatment Horse Treatment Time A P IB B IVA V EB TOTAL Concentration in mM 1 3 pre 5.86 4.73 8.10 11.31 3.78 8.48 7.71 49.96 1 3 post 5.12 3.68 5.09 8.40 2.70 7.98 8.67 41.65 2 3 pre 4.29 8.40 6.59 16.41 3.23 10.38 7.72 57.02 2 3 post 4.78 8.65 7.71 17.02 4.43 10.26 6.56 59.42 3 3 pre 4.67 4.55 7.40 9.24 3.56 6.00 9.27 44.71 3 3 post 4.52 5.12 8.14 11.00 5.32 6.37 7.28 47.75 4 3 pre 4.09 3.28 6.12 7.02 3.14 4.20 8.39 36.23 4 3 post 4.46 4.79 7.35 9.33 4.02 5.97 6.84 42.78 5 3 pre 4.09 1.86 5.38 3.51 3.63 3.10 7.72 29.30 5 3 post 4.16 4.64 6.38 8.81 3.44 5.73 6.85 40.01 MEAN PRE 4.60 4.60 4.56 6.72 9.50 3.47 6.43 8.16 MEAN POST 4.61 4.61 5.38 6.94 10.91 3.98 7.26 7.24 % Change of Mean 0.19% 17.87% 3.19% 14.91% 14.78% 12.89% 11.28% 6.62% Mean of % Changes 1.09% 9.50% 0.04% 8.83% 4.31% 1.99% 9.93% 6.49% p value (compared to control) 0.043 0.043 0.043 0.043 0.043 0.080 0.345 0.138 A = acetate, P = propionate, IB = isobut yrate, B = butyrate, IVA = isovalerate V = valerate, EB = ethyl butyrate

PAGE 136

110Table 3-18. Cecal liquor concentrations of individual and total SCFAs from cannulated horses treated with oral trimethoprimsulfamethoxazole at 30 mg/kg twice daily, before and after 4 days of treatment Horse Treatment Time A P IB B IVA V EB TOTAL Concentration in mM 1 4 pre 3.41 5.50 4.74 12.40 2.55 9.56 7.90 46.06 1 4 post 4.76 6.09 7.86 12.26 3.89 10.67 11.00 56.52 2 4 pre 3.74 3.76 5.85 9.33 3.77 5.57 7.40 39.42 2 4 post 4.18 2.46 4.83 5.45 2.84 4.13 10.16 34.04 3 4 pre 5.49 4.89 9.00 9.86 6.50 5.46 5.28 46.48 3 4 post 4.47 3.79 7.27 6.57 4.58 5.21 7.41 39.29 4 4 pre 4.66 4.39 7.42 8.94 4.30 5.56 9.85 45.14 4 4 post 4.49 2.49 5.48 4.93 3.66 3.80 10.87 35.72 5 4 pre 4.23 3.64 6.02 6.90 3.19 5.04 7.76 36.78 5 4 post 4.30 4.10 6.52 8.00 4.41 4.37 6.99 38.69 MEAN PRE 4.30 4.30 4.44 6.61 9.49 4.06 6.24 7.64 MEAN POST 4.44 4.44 3.79 6.39 7.44 3.87 5.64 9.29 % Change of Mean 3.15% 14.71% 3.26% 21.55% 4.61% 9.68% 21.56% 4.49% Mean of % Changes 6.75% 14.35% 5.18% 14.90% 9.91% 9.61% 1.74% 7.38% p value (compared to control) 0.080 0.225 0.225 0.138 0.043 0.138 0.138 0.080 A = acetate, P = propionate, IB = isobut yrate, B = butyrate, IVA = isovalerate V = valerate, EB = ethyl butyrate

PAGE 137

111 Effects on Cecal Anaerobic Bacteria The results of experiments examining the effects of antibiotic treatment on culturable cecal anaerobes are detailed in Tables 3-19 through 3-22. Unfortunately, due to equipment malfunction, cecal anaerobe coun ts from only one horse in the control group were available. All other treatment groups had values for all five horses in them. This deficiency of data in the control group makes statistical comparison difficult, as one individual is being used for comparison of all treatments. Despite the missing information, some trends were evident (Figur e 3-6). No significance was obtained with any treatment percentage change from pre-treat ment values compared to the control horse (p=0.138 to 0.500). Subjectively, there was a trend for the TM PS treatment to decrease the mean counts for all horses from pre-tr eatment values as compared to the NAX and TET treatments which did not show any obvious trend. The pre-treatment counts for total anaerobes also were much higher in all treatments compar ed to the control; however, this may be due to the influence of only one individual in the control group thus making comparison inaccurate. Table 3-19. Mean counts of culturable anaerobi c bacteria from seri al dilutions of raw equine cecal liquor, from 5 cannulated horses, before and after 4 days of control (no) treatment. The diluti on shaded in green was chosen for comparison. Dilution Control PRE Control POST 10-1 TNTC n=1 TNTC n=1 10-2 TNTC n=1 TNTC n=1 10-3 20.5 n=1 47.8 n=1 10-4 5.5 n=1 8.3 n=1 10-5 0.5 n=1 1.8 n=1 10-6 0 n=1 0 n=1 10-7 0 n=1 0 n=1 10-8 0 n=1 0 n=1

PAGE 138

112 Table 3-20. Mean counts of culturable anaerobi c bacteria from seri al dilutions of raw equine cecal liquor, from 5 cannulated horses treated with intramuscular ceftiofur sodium at 2 mg/kg twice daily, be fore and after 4 days of treatment. The dilution shaded in green was chosen for comparison. Dilution Ceftiofur PRE Ceftiofur POST 10-1 TNTC n=5 TNTC n=5 10-2 TNTC n=5 TNTC n=5 10-3 40 n=3 57.1 n=3 10-4 14.8 n=5 17.9 n=5 10-5 2 n=5 1.9 n=5 10-6 0.2 n=5 0.2 n=5 10-7 0 n=5 0.1 n=5 10-8 0 n=5 0 n=5 Table 3-21. Mean counts of culturable anaerobi c bacteria from seri al dilutions of raw equine cecal liquor, from 5 cannulated horses treated with intravenous oxytetracycline at 10 mg/kg once daily, befo re and after 4 days of treatment. The dilution shaded in green was chosen for comparison. Dilution Oxytetracycline PRE Oxytetracycline POST 10-1 TNTC n=5 TNTC n=5 10-2 TNTC n=5 TNTC n=5 10-3 35.6 n=4 17.8 n=3 10-4 7.9 n=5 10.8 n=5 10-5 1.1 n=5 1.2 n=5 10-6 0 n=5 0.1 n=5 10-7 0 n=5 0 n=5 10-8 0 n=5 0 n=5 Table 3-22. Mean counts of culturable anaerobi c bacteria from seri al dilutions of raw equine cecal liquor, from cannulated hor ses treated with oral trimethoprimsulfamethoxazole at 30 mg/kg twice da ily, before and after 4 days of treatment. The dilution shaded in green was chosen for comparison. Dilution Trimethoprim-Sulfa PRE Trimethoprim-Sulfa POST 10-1 TNTC n=5 TNTC n=5 10-2 TNTC n=5 107.5 n=1 10-3 80.6 n=5 26.3 n=4 10-4 16.7 n=5 16.8 n=5 10-5 1.1 n=5 2.4 n=5 10-6 0.4 n=5 0.5 n=5 10-7 0 n=5 0 n=5 10-8 0 n=5 0 n=5

PAGE 139

113 0 20 40 60 80 100 120 140Contro l Cefti o fur Oxytet TMPSThousandsCFU / ml Cecal Liquor Pre-Treatment Post-Treatment Figure 3-6. Mean cecal anaerob ic culture counts expressed as CFU / ml of liquor from five horses before and after treatment with control (no treatment), ceftiofur, oxytetracycline, or trimethoprim-sulfamethoxazole. In vitro Effects of SCFAs on Anaerobic Growth of Salmonella Figures 3-7 through 3-10 demonstrate no significant differences between the plasmid and spv gene containing strain 3306 versus the same plasmid cured (and spv deficient) strain 3337. Data for all short-chain fatty acid solution experiments is expressed as log CFU/ml with 95%CI error bars.

PAGE 140

114 1.00E+03 1.00E+04 1.00E+05 1.00E+06 1.00E+07 1.00E+08 1.00E+09TI M E 0 TIME 2h TIM E 4h TIME 6h TIM E 8 h T IME 1 0 hLog CFU/ml x3306 Control NaCl 30mM x3306 Control NaCl 100mM x3337 Control NaCl 30mM x3337 Control NaCl 100mM Figure 3-7. The effect of LB broth with sodi um chloride (control treatment) added at 30mM or 100mM on anaerobic growth of S. Typhimurium 3306 vs. 3337. All solutions were pH 6.5. Error bars represent 95%CI for two replicates of the experiment. 1.00E+03 1.00E+04 1.00E+05 1.00E+06 1.00E+07 1.00E+08 1.00E+09T I M E 0 TIME 2h TIM E 4h TIME 6h TIM E 8 h TI ME 1 0 hLog CFU/ml x3306 Acetate 30mM x3306 Acetate 100mM x3337 Acetate 30mM x3337 Acetate 100mM Figure 3-8. The effect of LB broth with sodium acetate added at 30mM or 100mM on anaerobic growth of S. Typhimurium 3306 vs. 3337. All solutions were pH 6.5. Error bars represent 95%CI for two replicates of the experiment.

PAGE 141

115 1.00E+03 1.00E+04 1.00E+05 1.00E+06 1.00E+07 1.00E+08 1.00E+09TIME 0 TI M E 2 h TIME 4 h TI M E 6 h TIME 8 h TI M E 10hLog CFU/ml x3306 Butyrate 30mM x3306 Butyrate 100mM x3337 Butyrate 30mM x3337 Butyrate 100mM Figure 3-9. The effect of LB broth with sodium butyrate added at 30mM or 100mM on anaerobic growth of S. Typhimurium 3306 vs. 3337. All solutions were pH 6.5. Error bars represent 95%CI for two replicates of the experiment. 1.00E+02 1.00E+03 1.00E+04 1.00E+05 1.00E+06 1.00E+07 1.00E+08TIME 0 TIME 2h TIM E 4h TIME 6h TIM E 8h TI M E 1 0hLog CFU/ml x3306 Propionate 30mM x3306 Propionate 100mM x3337 Propionate 30mM x3337 Propionate 100mM Figure 3-10. The effect of LB broth with sodium propionate adde d at 30mM or 100mM on anaerobic growth of S. Typhimurium 3306 vs. 3337. All solutions were pH 6.5. Error bars represent 95%CI fo r two replicates of the experiment.

PAGE 142

116 Effect of Acetate There was a dose-dependent inhibitory eff ect of acetate on the anaerobic growth of Salmonella in a nutritionally rich medium (LB broth). Figure 3-11 shows the growth curves of S. Typhimurium in LB broth with 30 or 100mM acetate solutions as compared to equiosmolar NaCl control solutions. Af ter 10 h of growth, the levels of bacteria growing in the 100mM acetate solution were a fu ll log fewer than those growing in either the 30 or 100mM NaCl, or the 30mM acetate solu tion; however, they were still growing. The 30mM acetate solution did not inhibit Salmonella significantly until the 6 h time point, and these differences were no l onger significant at the 10 h time point. 1.00E+04 1.00E+05 1.00E+06 1.00E+07 1.00E+08 1.00E+09TIM E 0 TI ME 2 h TIM E 4 h TIM E 6 h TIM E 8h TIM E 10hLog CFU/ml Control NaCl 30mM Control NaCl 100mM Acetate 30mM Acetate 100mM Figure 3-11. The effect of LB broth with sodium acetate at 30mM or 100mM compared to NaCl at 30mM or 100mM on anaerobic growth of S. Typhimurium. All solutions were pH 6.5. Error bars repres ent 95%CI for four replicates of the experiment. Effect of the Plasmid and spv Genes on Acetate Response There was no statistically significant effect of the presence of the virulence plasmid with the spv genes on the growth of Salmonella in the presence of acetate at 30mM or

PAGE 143

117 100mM concentrations. Dissimilar bacteria l starting concentrati ons between the two experiments made the growth curves appear different. Effect of Butyrate There was a dose-dependent inhibitory eff ect of butyrate on th e anaerobic growth of Salmonella in a nutritionally rich medium (LB broth). Figure 3-12 shows the growth curves of S. Typhimurium in LB broth with 30 or 100mM butyrate solutions as compared to equiosmolar NaCl control solutions. 1.00E+04 1.00E+05 1.00E+06 1.00E+07 1.00E+08 1.00E+09TIM E 0 TI ME 2 h T I ME 4h TI ME 6h TIME 8h TIME 10hLog CFU/ml Control NaCl 30mM Control NaCl 100mM Butyrate 30mM Butyrate 100mM Figure 3-12. The effect of LB broth with sodium butyrate at 30mM or 100mM compared to NaCl at 30mM or 100mM on anaerobic growth of S. Typhimurium. All solutions were pH 6.5. Error bars repres ent 95%CI for four replicates of the experiment. Effect of the Plasmid and spv Genes on Butyrate Response There was no statistically significant effect of the presence of the virulence plasmid with the spv genes on the growth of Salmonella in the presence of butyrate at 30mM or 100mM concentrations. Dissimilar bacteria l starting concentrati ons between the two experiments made the growth curves appear different.

PAGE 144

118 Effect of Propionate There was a dose-dependent inhibitory e ffect of propionate on the anaerobic growth of Salmonella in a nutritionally rich medium (LB broth). Figure 3-13 shows the growth curves of S. Typhimurium in LB broth with 30 or 100mM propionate solutions as compared to equiosmolar NaCl control solutions. 1.00E+04 1.00E+05 1.00E+06 1.00E+07 1.00E+08 1.00E+09TI M E 0 TI ME 2 h TIME 4h TI ME 6 h TIME 8h TI ME 1 0 hLog CFU/ml Control NaCl 30mM Control NaCl 100mM Propionate 30mM Propionate 100mM Figure 3-13. The effect of LB broth w ith sodium propionate at 30mM or 100mM compared to NaCl at 30mM or 100mM on anaerobic growth of S. Typhimurium. All solutions were pH 6.5. Error bars represent 95%CI for four replicates of the experiment. Effect of the Virulence Plasmid and spv Genes on Propionate Response There was no statistically significant effect of the presence of the virulence plasmid with the spv genes on the growth of Salmonella in the presence of propionate at 30mM or 100mM concentrations. Dissimilar bacteria l starting concentrati ons between the two experiments made the growth cu rves appear different. Both 3306 and 3337 in 100mM sodium propionate remained static over the 10 h experimental period. Extending the

PAGE 145

119 culture times to 18-24 h may have revealed very slow multiplication rates or eventual death of the cells. In vitro Effects of Cecal Liquor from Antibiotic-treated Horses on Anaerobic Growth of Salmonella Figures 3-14 and 3-15 show the growth curves for S. Typhimurium exposed to pooled samples of 10% filter-s terilized cecal contents from five horses treated with the same antibiotic (or control treatment). Gr owth in LB broth (Figure 3-14) and M9 minimal medium with 20% glucose (Figure 315) is evaluated. Wh ether the medium was nutritionally rich or limiti ng had no effect, and there was no significant difference between pooled additives for any treatment for either medium. 1.00E+06 1.00E+07 1.00E+08 1.00E+09 1.00E+10T I ME 0 TIM E 2h T I M E 4h TI M E 6h TIM E 8h T I M E 10h TI M E 14hLog CFU/ml LB Broth Only No Antibiotic Treatment NAX Treatment TET Treatment TMPS Treatment Figure 3-14. The effect of LB broth with 10% added filter-sterilized cecal contents pooled from five horses by treatm ent on the anaerobic growth of S. Typhimurium. The horses were treated w ith control (no treatment), ceftiofur (NAX), oxytetracycline (TET), or trim ethoprim-sulfamethoxazole (TMPS). Due to inadequate dilution of the starting bacterial inoculum (106-107 CFU/ml), the growth curves in Figures 3-14 through 316 demonstrate only a small portion of the

PAGE 146

120 logarithmic phase of bacterial growth. Di fferences in growth rate would be most apparent during this phase. The issue wa s corrected by increasi ng the dilution of the initial inoculum to lower than 105 CFU/ml in further experiments. For all further experiments examining the individual horse s, between horse di fferences were not observed, and means of the five horses were compared between treatments. There were no apparent inhibitory effects of the pooled cecal fluid additions in LB broth (Figure 314); however, some treatment additives subjectively appeared to stimulate growth rates in M9 relative to the control (Figure 3-15). 1.E+05 1.E+06 1.E+07 1.E+08TIME 0 TIME 2 h TIM E 4 h TIM E 6 h TI ME 8 h TI ME 10 h TI ME 12 h TI ME 24 hLog CFU/ml M9 Only No Antibiotic Treatment NAX Treatment TET Treatment TMPS Treatment Figure 3-15. The effect of M9 minimal me dium (+ glucose) with 10% added filtersterilized cecal contents pooled from five horses by treatment on the anaerobic growth of S. Typhimurium. The horses were treated with control (no treatment), ceftiofur (NAX), oxytetra cycline (TET), or trimethoprimsulfamethoxazole (TMPS).

PAGE 147

121 1.00E+05 1.00E+06 1.00E+07TIME 0h TIM E 2 h TIME 4 h TI M E 6h* TIME 8 h TIM E 1 0hLog CFU/ml Control M9 Minimal Media No Antibiotic Treatment NAX Treatment Figure 3-16. The effect of M9 minimal me dium (+ glucose) with 10% added filtersterilized cecal contents from antib iotic-treated horses on the anaerobic growth of S. Typhimurium. Data points are the mean of 5 individual horses treated with control (no treatment) or ceftiofur (NAX). Time 6h is a missing data point. When filter-sterilized cecal contents from horses treated with no treatment (control), ceftiofur sodium (NAX), and tr imethoprim-sulfamethoxazole (TMPS) were added in a 10% concentration to anaerobic salmonella cultures in M9 minimal medium (Figures 3-16 and 3-17), they demonstrated increased growth rates relative to M9 alone or M9 with 10% TET treated horse cecal flui d (p=0.001 to 0.005). The stationary phase cell densities appeared similar between all group s, including the controls. There was also no difference between the M9 only control t ubes versus the TET treated additive tubes, where the other 3 antibiotic treatments in creased growth rates during the logarithmic phase.

PAGE 148

122 1.00E+03 1.00E+04 1.00E+05 1.00E+06 1.00E+07TIM E 0 TI M E 2 h TIME 4h TIM E 6h TI M E 8h TI ME 1 0 hLog CFU/ml Control M9 Minimal Media TMPS Treatment TET Treatment Figure 3-17. The effect of M9 minimal me dium (+ glucose) with 10% added filtersterilized cecal contents from antib iotic-treated horses on the anaerobic growth of S Typhimurium. Data points are the mean of 5 individual horses treated with oxytetracycline (TET) or trimethoprim-sulfamethoxazole (TMPS). Effect of the Virulence Plasmid and spv Genes on Anaerobic Growth of Salmonella Exposed to Cecal Liquor from Antibiotic-Treated Horses Once again, no significant difference c ould be detected between strains 3306 and 3337 in any of the experiments (data not shown), so data from each trial were combined for the two strains and considered replicates The absence of variation between these strains is not surprising, as no intestinal gr owth function has been attributed to the virulence plasmid or spv genes in any host species. Discussion Antibiotic administration had minimal effect s on most of the dependent variables measured in this study. One obvious area of contention is the dur ation of antibiotic treatment given to these experimental animal s. In hindsight the four-day duration of

PAGE 149

123 therapy may have been too short to alte r the intestinal microenvironment to any statistically significant degree. The selecti on of four days of treatment was based on anecdotal and experimental data indicating th at enteric complications associated with antibiotic usage in horses commonly occur within this period. In clin ical equine practice it is not uncommon to treat horses with antib iotics for variable periods, ranging upwards from three days. Antibiotic-associated diarrhea frequently occurs within the initial three days of therapy and there are reports of horses developing se vere and fatal diarrhea after a single dose of antibiotic.205 Several studies examining the effect of antibiotic treatment on development of diarrhea in horses206 and humans,207 or on fecal bacterial counts in horses49 and humans,85 have concluded that the devel opment of diarrhea, trough bacterial and SCFA levels82 and minimums in microflora -associated characteristics86 were usually achieved within three days of commencing oral or parenteral thera py. This supported the selection of a four-day treatment course as any changes in fecal bacterial or SCFA measurements, or biochemical derangements would likely be refl ected by the cecal microenvironment. Carryover effects of antibiotic treatment on cecal microflora were minimized using a randomized block experimental design with a relatively long in ter-treatment period (>30 days). The cecal flora and its biochemical environment are in a constant state of flux, adapting in response to host, diet, season, an d other factors. It is logical to assume that antibiotic treatment of the host is a significant event in shaping that adaptive response. Whether or not the microenvironment can actually reconstitu te to a normal or a pre-treatment state after an event such as an tibiotic therapy was not determined in this

PAGE 150

124 study. Given the minimal treatment effects in the present study it was unlikely that significant carryover effects occurred. None of the horses in the study devel oped overt complications of antibiotic treatment, including the development of dia rrhea. There was also no apparent reduction in appetite during the treatment period. It has been reported that antibiotic therapy, especially oral therapy, may cause part ial or complete anorexia in horses,61 although this is not a reported complication of any of the antibiotics used in this study. It has been suggested that the risk of an tibiotic-associated diarrhea ma y be potentiated by a reduction in feed intake.208 The effect of feed intake on antibi otic-mediated changes in intestinal flora could be a worthwhile focus of future studies. In the present study it was possible that transient diarrhea could have been passed unobserved, but there was no residual evidence of this (wet or soiled tail, perineum, or hocks). Several investigators have reporte d antibiotic-mediated bi ochemical changes in fecal and cecal contents in the abse nce of diarrhea in several species.49;82-85 Consequently, the absence of diarrhea should not by itself provide an explanation for the minimal observed changes in cecal microflora. The anticipated effect of antibiotics with activity against anaerobic bacteria would have been an increase in pH. This result s hould be coincident with a significant reduction in production of SCFAs. Ne ither effect was observed unde r the present experimental conditions, with the exception of some changes in individua l SCFA concentrations. The time of concentrate feeding and cecal sampli ng was consistent each day in order to minimize normal temporal and dietary infl uences on the cecal pH. Under normal conditions it is likely that cecal pH varies albeit within a rela tively small range.

PAGE 151

125 Continuous pH monitoring or more frequent sa mpling might have yiel ded different data, but this was considered unlikely. Furtherm ore, techniques used to continuously or repeatedly record luminal pH are likely to lead to contamination of the anaerobic environment with air. The significance of luminal pH may not be as important as previously thought, especially in the case of attaching and invasive pathogens such as Salmonella spp It was recently demonstrated that the absorption of SCFAs from the intestinal lumen was relatively independent of the luminal pH.209 Local pH dictates the ionization status of these acids, and therefore their ability to cr oss cell membranes. However, the pH at the apical cell surface is near neutral, and remains so in spite of major changes in lumen pH. The regulation of this pH layer is exquisite ly dependent on bicarbon ate secretion and may explain how invasive pathogens can biochemically recogniz e a change in location and use that as a signal to attach or invade. Antibiotic therapy has the potential to disrupt non-bacterial members of the autochthonous flora, namely the protozoa. The most commonly studied ecosystem with respect to autochthonous protoz oa is the ruminant forestomach. The species, numbers, and functions of protozoa residing within the cecum and large colon of horses are significantly different th an that of ruminants.154;164;210 Along with bacteria, protozoa are also involved in the terminal digestion and fermentation of polysaccharide foodstuffs, as they function to ferment polysaccharides (m inimal contribution to the total), store starches—thus protecting them from bacter ial fermentation, as well as regulating the number of resident fungi and ba cteria through predatory actions.211 Indirect effects of

PAGE 152

126 antibiotic therapy could also regulate total numbers of commensal protozoa, through changes in local pH, or metabolites such as SCFAs. Adam reported wide temporal variations in the cecal protozoa of horses even within an individual on a consistent diet composition201 and this was corroborated by Moore and Dehority.210 Reported values for cecal protozoa in the normal, non-fistulated horse range from 4 x 103 protozoa per ml contents 154 to 55 x 103 per ml.210 This intraand interanimal variability in protozoal numbers was also noted in the present study and made recognition of potential treatment effects di fficult. Several inve stigators have found divergent values for the same characteristi c being studied, depending on the whether the animal had ceco-cutaneous access, or was intact and sacrificed.166 Diet type has also been reported to alter the relative proportions of protozoal species in the cecum of the equine, but not overall numbers.164 This is in contrast to th e equine large intestine, where commensal protozoal numbers were typically 2-3 times greater than in the cecum and their numbers changed significantly in respons e to increasing the amount of fermentable carbohydrate in the diet.210 The variability in protozoal numbers in the present study, coupled with the relatively low number of experimental animal s, made investigation of treatment effects difficult. The effect of antibiotic therapy on equine cecal protozoal numbers has not been previously reported, so no assumptions were made prior to experimentation. When examining data from ruminants, it might be hypothesized that antibiotic therapy with an agent encompassing a protozoal spectrum su ch as metronidazole, the sulphonamides, trimethoprim, the lincosamides,153 or aminoglycosides212 would decrease relative numbers of commensal protozoa or change the demographi cs of their populations.

PAGE 153

127 However, this may be irrelevant in horses as Moore and Dehority concluded that the commensal protozoa in horse s contribute insignificantly to total functional hindgut fermentation.210 Significant antibiotic-induced changes in SC FA concentrations were seen in the present study, but these changes were not as great as anticipated ba sed on similar studies in human beings. Previous data in horses unf ortunately are restricted to studies involving a single antibiotic treatment making direct comparisons difficult.170;174 There are several technical reasons that could explain the minimal antibiotic-induced changes. These include a relatively insensitive measuring technique, deterioration of the volatile acids during sample processing and storage, or inappropriate model application (i.e., cannulated animals vs. intact). Measurement of SCFA concentrations at a single point in time does not directly reflect drug-induced chan ges in production as fluctuations in cecal volume, as well as local SCFA absorption or conversion could dist ort significant changes associated with treatment. Post-prandial vo lume of the equine cecum or large colon has been shown to increase from one to four times compared to the pre-fed state,213 so dilution must be considered when trying to determine production rates or total cecal SCFAs. A constant time between feeding and sampling, coupled with a constant diet in the present study should have minimized poi nt-to-point fluctuations in volume, independent of treatment. Short-chain fatty aci d relative proportions to each other as a percentage of the total (molar ratios) may be a better indicator, as cecal volume can only be estimated in the living animal. Unfortuna tely molar ratios were also not significantly different from controls in antibiotic treated animals.

PAGE 154

128 Oxytetracycline administration had the gr eatest impact on SCFA concentrations. The association between the tetracycline family of antibiotics and diarrhea in horses has received enormous attention in the veterina ry literature. The incidence of diarrhea appears to be greatest when th e drug is administered orally.55 Recently it has been suggested that the incidence of diarrh ea after intravenous administration of oxytetracycline has been ove r-stated (personal communica tion, G.D. Lester, Murdoch University, Western Australia). Biliary ex cretion of oxytetracyc line after parenteral administration is considered to be minimal.170 This route of ex cretion could however become important when high doses of the drug are used. Interestingly, early reports of the association between tetracycline and di arrhea were based on e xperimental scenarios involving very high parent eral doses of the drug. The fact that ceftiofur an d trimethoprim-sulfamethoxazole had minimal significant impact on total and individual SCFA levels leaves open the pa thophysiology regarding diarrhea caused by these compounds. There is recent evidence, albeit anecdotal, that the diarrhea reportedly associated with ceftiofur may be mediated by Clostridium difficile (personal communication, G.D. Lester, Murdoc h University, Western Australia). The relationship between trimethoprim-sulfamethoxa zole preparations and diarrhea may be influenced by alterations in feed intake.208 Continued investigation into the area of antibiotic effects on the gastrointestinal mi croenvironment could include more sensitive identification and quantification of natu rally occurring and locally active compounds such as SCFAs, bacterial by-products of metabolism, bacterial breakdown compounds such as endotoxin, as well as fluxes in th e resident commensal popul ations of organisms such as yeasts, bacteria, and protozoa.

PAGE 155

129 There was no significant effect of antibio tic treatment on cecal anaerobic bacterial numbers. There was however trends in the data to suggest that trimethoprimsulfamethoxazole may have reduced bacterial nu mbers. The medium used to culture and quantitatively examine the flora in biologic anaerobic samples is typically nutritionally rich and non-selective in order to maximize r ecovery of the more fastidious organisms. A selective antibiotic (gentamicin) was include d in the anaerobic cu lture media in order to reflect a true obligate an aerobic bacterial count. Anaer obic bacteria, due to their decreased trans-membrane electrical poten tial, are intrinsically resistant to the aminoglycosides, as the drug cannot be transported into the cell212 (personal communication, M. Cox, Anaerobe Systems, San Jose, CA). This happens to a minor degree in facultative anaerobes undergoing anae robic respiration. F acultative anaerobes may contribute somewhat to carbohydrate fe rmentation in the normal animal, though they may play a larger part in hyper-fermentation disorders, such as car bohydrate overload in horses and rumen acidosis in cattle. Faculta tive anaerobes are also a small minority of the cecal microflora in the horse, as one investigator reported that the overwhelming majority of bacteria in the cecum of the horse were strict anaerobes (> 80%).154 Few historical quantitative reports on the total anaerobic flora of the rumen or cecum have addressed this issue of b acterial sub-populations, and most used non-selective media under completely anaerobic conditions, which does not differentiate between species of varying oxygen tolerance. de Fombelle repor ted that large intestinal (cecum and right ventral colon) anaerobic bacter ia in cannulated ponies were present at approximately 108 CFU/ml, while the aero-anaerobic p opulation was a little more than 107 CFU/ml. Since the aero-tolerant anaerobes w ould be culture-included in th e population of the strict

PAGE 156

130 anaerobes, according to this data, th ey should number approximately 9 x 107 CFU/ml.169 Reported anaerobic bacterial strain s isolated from the cecum of Equidae include Butyrivibrios Streptococcus Bacteroides Lactobacillus Selenomonas Eubacteria Proprionibacteria Staphylococcus Veillonella Clostridium and Bifidobacter .51;166;171 Oxygen tolerant strains of stre ptococci, lactobacilli, and stap hylococci have been isolated from the horse cecum.163 In 1973, Kern et al. reported that a larger proportion of the bacteria isolated from the cecum of ponies were facultative anaerobes (as compared to the rumen of steers). The authors hypothesi zed that either less active fermentation occurred in the pony cecum or that th e availability of oxygen was increased.164 If the latter is indeed true, it could be anticipated that fistulated or cannulated animals will have even lower average counts of anaerobes and greater than average counts of aerotolerant species, depending on the integrity of the cannula. Discrepancies between direct smear counts a nd viable or culturable organisms have been reported in two studies.163;164 This is likely related to the fastidiousness or oxygen tolerance of the organisms, and was an equall y important concern in this work. The fact that a partially selective me dium was used in conjunction with chronically cannulated animals, can easily explain why the actual num bers of culturable anae robes in this study were consistently lower than other repor ted values for the equine cecum. Also, differences in measurement and reporting units (e.g., CFU per volume of contents vs. weight of contents) further complicate comp arison between studies. Furthermore, our experimental collection technique does not address an important population of bacteria that are adherent to the mucosa.214 This adherent population of bacteria was included in a study of equine cecum, but the authors made no distinction between luminal bacterial

PAGE 157

131 numbers and adherent numbers in their results.166 Ideally, studies where the animals were sacrificed and the gastrointestinal tr act was immediately opened, and contents as well as mucosal scrapings were collected may give a more accurate representation of the true numbers of bacteria. Our data failed to demonstrate marked effects of antibiotics on SCFA concentrations in vivo Consequently, it is difficult to directly implicate changes in SCFA concentrations in the pathogenesis of an tibiotic-associated diarrhea. Nevertheless, manipulation of luminal SCFA concentrati ons may still be an important focus of treatment and prevention of salm onella infection, particularly in high-risk patients. It is apparent from our data that acetate, propionate and butyrate inhibit anaerobic growth of Salmonella in a concentration-dependent manne r. Propionate and butyrate at 100mM concentrations were most inhibitory, followed by propiona te at 30mM and acetate at 100mM. Further investigation with more appropriate physiol ogic concentrations is now warranted. Once an effective breakpoint is determined; methods of manipulating the endogenous ecosystem to approximate those va lues become an important next step. Also, this study only looked at single SCFA s, it is quite probable that combination cocktails may yield more significant results. We initially hypothesized that certain SCFAs would be inhi bitory to growth of the Salmonella and that antibiotics would, by reduci ng anaerobic bacterial numbers, be associated with reduced concentrations of SCFAs. It was anticipated that if this hypothesis were true that addi tion of pathogenic bacteria to sterile cecal liquor from antibiotic-treated horses woul d demonstrate enhanced growth when compared to cecal liquor collected from horses that had not b een treated. Unfortunately, data reported

PAGE 158

132 above confirmed that the effect of antibiotics on cecal contents was minimal. Given this the finding that Salmonella grew equally well in the unt reated control, trimethoprimsulfamethoxazole treated, and ceftiofur trea ted cecal liquor was not unexpected. These data were consistent between plasmid/ spv bearing and plasmid/ spv deficient strains of Salmonella What was not expected was the findi ng that salmonella growth appeared to be slowed when added to the cecal liquor of horses treated with oxytetracycline. There may have been residual active drug in the samples, although Horspool determined that less than 17% of the amount of oxytetracyclin e administered intravenously reaches the cecal liquor.170 It is also possible that changes in other organic acids which were not measured, such as lactate, could also ha ve influenced bacterial growth rates. Another interesting finding was that the growth of Salmonella in all cecal liquor supplemented media, with the exception of th at collected from oxyt etracycline-treated horses, was greater than that in the M9 me dium, which is nutritionally complete for Salmonella spp. This indicated that sterile cecal fluid (even after antibiotic treatment) has intrinsic nutritive value relative to the minimal medium alone.

PAGE 159

133 CHAPTER 4 SUMMARY, CONCLUSIONS, AND FUTURE DIRECTIONS There is a dearth of information regarding the molecular characteristics of salmonellae associated with infection of hor ses. The initial aim of the study was to contrast two populations of horses with respec t to molecular characteristics, specifically the presence of bacterial plasmids and spv genes. The populations investigated included a symptomatic hospitalized group at the Vete rinary Medical Teaching Hospital at the University of Florida and a population of asym ptomatic animals from horse properties in North Central Florida. We initially hypothe sized that salmonellae recovered from asymptomatic horses would be less likely to contain virulenc e genes than those associated with clinical disease. Unfort unately, no salmonellae were isolated from asymptomatic horses. We considered seve ral possibilities why th is occurred, including the use of a relatively insensitive method of detection, the probl em of intermittent shedding, and the interference of additional fecal compounds on bact erial growth. The technique used in the present study of enrich ment and culture remains the gold standard of diagnosis, although more sensitive methods su ch as PCR are available. Our need for live bacteria for additional molecular and sensitivity testing made utilization of PCR impractical for our purposes.106 Our ability to recover the bacteria from fecal samples was validated on numerous occasions through positive culture of hospitalized animals. The most likely explanation for the failure to recover salmonellae from asymptomatic animals is that horses in this region have an extremely low carriage rate of salmonellae.

PAGE 160

134 Given the wide variation in reported rates of bacterial shedding this observation should not be considered novel. The focus of the study shifted to a prim arily descriptive assessment of the molecular characteristics of bacteria recovered from symptomatic animals. In addition we examined some clinical and laboratory fa ctors that were associated with systemic spread and mortality. The demographics of affected animals appear ed to reflect that of the entire hospital although this comparison was not investigated st atistically due to a difficulty in obtaining accurate data of the hospital population for this period. The diseased population was comprised of large number of juvenile animal s; interestingly the case outcome was worse in horses less the 4 years of age. A likel y explanation was that neonatal foals, when affected with Salmonella frequently became bacteremic w ith secondary sites of infection. More than 70% of the isolates were obtained from fecal samples, which is consistent with the entero-environmental life cycle of these bacteria.215 Multiple serovars were isolated from six animals, which demonstr ate the fact that different serovars of these organisms can simultaneously co-infect a single host. Carriage of virulence plasmids, as defined by the presence of spv genes, was restricted to certain Group B serovars. These isolates included the serovars Typhimurium, Typhimurium var. Copenhage n, and 4,5,12:i-monophasic, an antigenically close relation to S. Typhimurium. The virulence plasmid and spv genes have been shown to be important in the calf model of salmonella pathogenesis.10 Isolates recovered systemically were more likely to be spv -gene positive. It was al so not surprising that the presence of virulence plasmids also conferre d a higher case mortalit y. These findings are

PAGE 161

135 consistent with data collected in other species. Our data set was not sufficiently robust to allow us to retrospectively examine the relationship between spv genes and severity of enteric disease. A prospectiv e clinical study examining the di sease severity with respect to these genes would be required to investigate this association. More than 64% of the clinical isolates examined carried a large (> 20-kb) plasmid, but only 20% of these isolates carried a virulence plasmid. Speculation as to what the other plasmids could potentially be led to the preliminary investigation of antibiotic resistance determinants. Three isolates were selected on the basis of appropriate antibiotic sensitivities and yielded transference of ceftiofu r, cefazolin, and ampicillin resistance. Likely the majority of these large plasmids contain R determin ants, but further examinations would be required in order to increase the accuracy of this statement. Resistance to common antimicrobial agents is emerging issue of concern regarding Salmonella in human medicine. Monitoring of resistance levels in endemic strains of pathogens such as Salmonella allows clinicians to remain aware of the selective pre ssures placed on these organisms. Four serovars of Salmonella accounted for the 22 mu ltidrug-resistant isolates identified in this study (resistant to 8 drugs out of the 12 cl inically relevant drugs tested). S. Java (10), S. Typhimurium var. Copenhagen (5), S. Javiana (2), and S. Newport (1). This may be more a refl ection of the population being studied (sick, surgical, and hospitalized patients) as they are the population most lik ely to be receiving these drugs. The second major phase of the study was to investigate the relationship between antimicrobial therapy and salmonella gr owth. We applied a combination of in vivo and in

PAGE 162

136 vitro techniques to investigate this, as we wa nted to avoid animal inoculation with Salmonella The initial stage involved examin ing the effect of antibiotics on cecal microflora using Thoroughbred horses with cecal cannulae. Dependent variables examined included cecal anaerobe counts, cec al pH, and cecal SCFA concentrations. There were no significant diffe rences detected in total an aerobe counts with respect to antibiotic treatment, but several signifi cant changes in SCFA concentrations were observed. A relationship between these va riables is assumed, but our methods and experimental numbers were likely insufficient to detect significant differences in bacterial counts. Newer techniques i nvolving quantitative PCR with species-specific primers are much more sensitive as they do not require the organisms to be viable.216;217 Antibiotics were shown to have a significant effect on several SCFA concentrations, but in general the magnit udes of the changes observed were mild. Changes in the control group raised some con cerns regarding the preparation used in the study. Likely our data would have been st rengthened by a combination of increased numbers of experimental animals and/or repeated sampling of contents to reduce variability. There are problem s associated with single point sampling, most importantly is the impact of the unmeasured variable in testinal volume on concentration. Effort was made in the experimental design to minimize any impact of diet t ype and timing relative to collection, but changes in intestinal vol ume induced by antibiotic treatment would not have been detected. Oxytetracycline had the great est impact on SCFA concentration. The relationship between the tetracycline family of antibiotic s and diarrhea is cont roversial, but the literature indicates that diarrhea can occur in response to treatment.52;53;55 Our data

PAGE 163

137 indicate that there is an eff ect of oxytetracycline on cecal flor a, albeit of small magnitude. This is most likely due to the limited biliary excretion of the drug. We then determined that SCFA were inhibitory to equine isolates of Salmonella using in vitro techniques. There was a concentrat ion-dependant inhibition of anaerobic salmonella growth by SCFA solutions. This was independent of spv gene presence and further supports the hypothesis that the spv genes are not involved in the enteric phase of the disease in horses. Propionate and butyr ate at 100mM concentrations were most inhibitory, followed by propionate at 30mM and acetate at 100mM. These concentrations are considered supraphysiol ogic, but demonstration of in vitro inhibition provides the basis for additional in vivo investigations. The focus of additional studies would likely be to identify supplements that could be ut ilized not only therap eutically, but also prophylactically during periods of increased pathogen susceptibility. SCFA-treated starches fed to rats increased th e colonic levels of those acids.193 Treated feedstuffs or supplements would be readily accepted by horse owners and veterinarians interested in minimizing the side-effects of antimicrobial therapy. Based on the data collected during the fi rst phase of the antibiotic study it was unlikely that we would be able to demonstr ate significant enhancem ent or inhibition of growth with the liquor from antibiotic treated horses. Indeed, there were no differences between treatments for ceftiofur, trimet hoprim-sulfamethoxazole, and control (no treatment) on the anaero bic growth rates of Salmonella when they were exposed to 10% solutions of sterilized cecal liquor from antib iotic-treated horses. There were however two interesting, if not unexpected find ings. The cecal liquor collected from oxytetracycline-treated anim als appeared to inhibit in vitro salmonella growth relative to

PAGE 164

138 that of control, ceftiofur, or trimethoprim -sulfamethoxazole treated cecal liquor. One likely explanation may be that residual antibiotic may have suppressed growth in this treatment group. Of additional interest was that control cecal liquor enhanced growth over glucose-supplemented media indicating that growth of bacteria is facilitated by factors within the cecal liquor. Based on these results, future work investigating the indirect impact of therapeutic antimicrobial therapies on commensal gastroin testinal flora in humans and horses should be encouraged. Antibiotic-associated dia rrheas, especially those attributable to Salmonella spp. are important causes of morbidity and mortality worldwide, and efforts to understand the pathophysiology behind antibi otic-host-pathogen interactions may lead to novel and economic preventative modalities or therapeutics.

PAGE 165

139 APPENDIX A SALMONELLA EPIDEMIOLOGY DATA COLLECTION SHEET Page 1: Salmonella epidemiology survey

PAGE 166

140 Page 2: Salmonella epidemiology survey

PAGE 167

141 APPENDIX B INDEX OF SUPPLIERS AND CONTACT INFORMATION Corning Incorporated Life Sciences 45 Nagog Park Acton, MA 01720 Tel: 978-635-2200 Tel: 800-492-1110 Fax: 978-635-2476 http://www.corning.com/lifesciences/USCanada/en/ Bio-Rad Laboratories 1000 Alfred Nobel Drive Hercules, CA 94547 Tel: 510-724-7000 Fax: 510-741-5817 http://www.biorad.com JA Webster Veterinary Supply 86 Leominster Road Sterling, MA 01564-2198 Tel: 800-225-7911 Tel: 978-422-8211 Fax: 978-422-8959 http://www.jawebster.com/index.html Supelco Chromatography Tel 800-247-6628 http://www.sigmaaldrich.com/Brands/Su pelco_Home.html MJ Research, Inc. 5350 Capital Court, #102 Reno, NV 89502 Tel: 888-652-9253 (888-MJCYCLE) 888-735-8437 (888-PELTIER) http://www.mjresearch.com Sheldon Manufacturing 300 N. 26th Avenue Cornelius, OR 97113 Tel: 800-322-4897 503-640-3000 http://www.shellab.com Anaerobe Systems 15906 Concord Circle Morgan Hill, CA 95037 Tel: (408) 782-7557 Fax: (408) 782-3031 www.anaerobesystems.com Sigma-Aldrich Chemical Corp. 3050 Spruce Street St. Louis, MO 63103 Tel: 800-521-8956 http://www.sigmaaldrich.com bioMrieux, Inc. 100 Rodolphe Street Durham, NC 27712 Tel: (919) 620 20 00 Fax: (919) 620 22 11 http://www.biomerieux-usa.com Fisher Scientific International Inc. Hampton, NH 03842 Tel: (603) 926-5911 Fax: (603) 929-2379 www.fisherscientific.com

PAGE 168

142 Remel Inc. 12076 Santa Fe Drive P.O. Box 14428 Lenexa, KS 66215 www.remel.com SPSS Inc. 233 S. Wacker Drive 11th Floor Chicago, IL 60606 Tel: (312) 651-3000 http://www.spss.com Microsoft Corporation One Microsoft Way Redmond, WA 98052-6399 http://www.microsoft.com Perkin Elmer Corporation Life and Analytical Sciences Division 549 Albany Street Boston, MA 02118, USA Tel: (617) 482-9595 Customer Care: 800-762-4000 (USA) http://www.perkinelmer.com Invitrogen Corporation (formerly Gibco BRL) 1600 Faraday Avenue Carlsbad, California 92008 Tel: (760) 603-7200 Fax: (760) 602-6500 http://www.gibcobrl.com Qiagen Inc. 28159 Avenue Stanford Valencia, CA 91355 Tel: 800-426-8157 Fax: 800-718-2056 http://www.qiagen.com UVP Incorporated BioImaging Systems Group 2066 W. 11th Street Upland, CA 91786 Tel: 800-452-6788 or 909-946-3197 Fax: 909-946-3597 http://www.uvp.com TREK Diagnostic Systems, Inc. 982 Keynote Circle, Suite 6 Cleveland, Ohio 44131 USA Tel: 216-351-TREK (8735) Toll Free: 800-871-8909 Fax: 216-351-5456 Technical Support: 1.800.642.7029 www.trekds.com

PAGE 169

143 APPENDIX C SALMONELLA ISOLATE INDEX Table C-1. Index of Salmonella isolates by group and se rovar, with plasmid and spv gene status Isolate Identification Number (Case) Plasmid spv Group Serovar 1 Unknown 2 Unknown 3 B Typhimurium 4 E Anatum 5 C2 Newport 6 B Typhimurium 7 B Typhimurium 8 B Typhimurium 9 B Java 10 B 4,5,12:i-monophasic 11 B 4,5,12:i-monophasic 12 C2 Newport 13 B Java 14 C2 Newport 15 B Java 16 B Java 17 B Java 18 B Agona 19 B Typhimurium var. Copenhagen 20 B Java 21 B 4,5,12:i-monophasic 22 B Typhimurium var. Copenhagen 23 B Typhimurium 24 C1 Hartford 25 B Typhimurium 26 B Typhimurium var. Copenhagen 27 B Java 28 B Java 29 B Typhimurium var. Copenhagen 30 B Java 31 D Javiana

PAGE 170

144 Table C-1. Continued Isolate Identification Number (Case) Plasmid spv Group Serovar 32 C1 Braenderup 33 B Java 34 C2 Newport 35 E Anatum 36 F Rubishlaw 37 C2 Muenchen 38 D Javiana 39 C2 Newport 40 B Typhimurium var. Copenhagen 41 B Java 42 --Unknown 43 B Typhimurium var. Copenhagen 44 B Typhimurium var. Copenhagen 45 E London 46 C2 Muenchen 47 C2 Newport 48 B Typhimurium 49 B Typhimurium 50 B SaintPaul 51 C1 Mbandaka 52 C2 Newport 53 D Miami 54 B SaintPaul 55 B Java 56 C2 Muenchen 57 B Java 58 D Miami 59 B Java 60 B SaintPaul 61 B Java 62 D Miami 63 E London 64 E Meleagridis 65 D Javiana 66 C2 Newport 67 D Miami 68 C2 Muenchen 69 D Javiana

PAGE 171

145 Table C-1. Continued Isolate Identification Number (Case) Plasmid spv Group Serovar 70 B Java 71 D Miami 72 B SaintPaul 73 C1 Hartford 74 C2 Muenchen 75 B Java 76 C1 Infantis 77 D Miami 78 B Java 79 D Javiana 80 C2 Muenchen 81 B Multiple B Serovars 82 B SaintPaul 83 B Reading 84 E Anatum 85 B Java 86 B Java 87 C2 Newport 88 B Java 89 B Typhimurium var. Copenhagen 90 C2 Tallahassee 91 C2 Newport 92 B Java 93 D Miami 94 E Newington 95 C2 Newport 96 B Java 97 B Java 98 B Java 99 C2 Newport 100 E Anatum 101 B SaintPaul 102 C2 Newport 103 E Newington 104 C1 Mbandaka

PAGE 172

146 APPENDIX D SALMONELLA DATABASE CASE DESCRIPTIVE INFORMATION This appendix contains the complete data base of descriptive information obtained regarding each clin ical isolate of Salmonella examined in this study, displayed in Tables D-1 through D-3. Isolate coding has been a ssigned to conceal identifying client and patient information. Blank cells indicate th at the record was missing or otherwise not available and annotation stating that the info rmation could not be determined from the record and patient history is shown in cases where the record was reviewed.

PAGE 173

147Table D-1. Salmonella case descriptive info rmation: breed, age, sex, presenting compla int, risk factors for salmonellosis, spe cimen origin, and salmonella group(s) and serovar(s). Blank cells indicate missing or unavailable records. Case ID Breed Age Sex Presenting Complaint Clinical Syndrome Risk Factors Specimen Origin Salmonella Group Serovar 2nd Salmonella Group 2nd Serovar 1 QH 14y M Febrile, colic Colic No Diarrhea Colic Feces C2 unknown N/A N/A 2 Unknown 13y F Diarrhea Feces B unknown N/A N/A 3 TB 9y F Diarrhea, weight loss, late in foal Diarrhea, Weight Loss Pregnant Feces B typhimurium N/A N/A 4 Paso Fino 3y M Diarrhea, colic SC impaction. Ex Lap, did badly post-op, severe endotoxemia, died spontaneously Colic Diarrhea Colic, Anesthesia / Surgery Feces E anatum N/A N/A 5 Arab 12y M Colic Diarrhea Colic Feces C2 newport N/A N/A 6 Standardbred 2y F Diarrhea, fever Diarrhea, Fever Feces B typhimurium N/A N/A 7 TB 2wks M Fever, swollen joint, diarrhea Synovitis, Fever, Diarrhea Systemic Disease Joint Fluid B typhimurium N/A N/A 8 Paso Fino 2y F Chronic diarrhea (6 weeks) Diarrhea Chronic None Determined Feces B typhimurium N/A N/A 9 Standardbred 2y F Colic, fever, reflux, anterior enteritis, previously admitted (6 d. prior) to VMTH for scintigraphy and CSF tap. Exploratory laparotomy showed 3m proximal j ejunum thickened w/serosal ecchymoses Colic No Diarrhea, Fever Colic, Anesthesia / Surgery Duodenum Necropsy B j ava N/A N/A 10 TB 2d F Diarrhea, DOA Diarrhea Neonate SI Necropsy B 4,5,12 : imonophasic N/A N/A 11 TB 1m M Colic, GDUD, S. equi equi abscess, collapsing trachea Colic No Diarrhea Systemic Disease (non-colic), Colic LI Necropsy B 4,5,12 : imonophasic N/A N/A

PAGE 174

148 Table D-1. Continued Case ID Breed Age Sex Presenting Complaint Clinical Syndrome Risk Factors Specimen Origin Salmonella Group Serovar 2nd Salmonella Group 2nd Serovar 12 Paint 1y M Diarrhea, foaming at mouth, coughing since show 1m prior, currently on "cough medicine", given Probios 1d prior to admission Diarrhea Systemic Disease (non-colic) SI, SC, Lymph Node Necropsy C2 newport N/A N/A 13 Warmblood 5y M Shipped from Belgium to NY, NY to MD, MD to FL in the week prior to presentation. Developed fever and respiratory disease Diarrhea, Fever Shipping, Systemic Disease (non-colic) Feces B j ava N/A N/A 14 QH 1mo M Diarrhea, bucket-fed orphan foal (mare died 8h postpartum) Diarrhea Antibiotic Administration Feces C2 newport N/A N/A 15 QH 7y M Colic cecal impaction, surgical correction, post-op diarrhea Post-Op Colic Diarrhea Colic, Anesthesia / Surgery Feces B j ava N/A N/A 16 TB 5m F Pneumonia Diarrhea, Pneumonia Antibiotic Administration, Systemic Disease Feces B j ava N/A N/A 17 Mix Breed 20y M Diarrhea, post-o p surgery for incisor removal, colic episode 4d prior to surgery. Has had repeated leukopenic episodes since desmotomy surgery 2y prior Post-Op (noncolic) Diarrhea Anesthesia / Surgery, Colic Feces B j ava N/A N/A 18 Unknown Necropsy Necropsy unknown COD None Determined LI Necropsy B agona N/A N/A 19 TB 1y F Shipped from KY 9d. prior to presentation. Post-op stapling bilateral carpi, developed diarrhea and fever Post-Op (noncolic) Diarrhea Shipping, Anesthesia / Surgery Feces B typhimurium (copenhagen) N/A N/A

PAGE 175

149 Table D-1. Continued Case ID Breed Age Sex Presenting Complaint Clinical Syndrome Risk Factors Specimen Origin Salmonella Group Serovar 2nd Salmonella Group 2nd Serovar 20 TB 14y M Fever, diarrhea, post-op epiploic foramen entrapment, SI resection Post-Op Colic Diarrhea, Fever Colic, Anesthesia / Surgery Feces B j ava N/A N/A 21 TB 1m M Diarrhea, fever Diarrhea, Fever None Determined Feces B 4,5,12 : imonophasic N/A N/A 22 TB 1.5y M Fracture distal humerus, anesthesia for radiographs Diarrhea Anesthesia, Systemic Disease Feces B typhimurium (copenhagen) N/A N/A 23 TB 1y M Colic, nephrosplenic entrapment and colon torsion, died of colon rupture 8d post surgery Post-Op Colic Diarrhea Anesthesia / Surgery, Colic Feces B typhimurium N/A N/A 24 TB 2m F Bilateral stifle septic arthritis / osteomyelitis Synovitis, Osteomyelitis, Diarrhea Antibiotic Administration, Systemic Disease Joint Fluid C1 hartford N/A N/A 25 TB 1m F Shipped from IN 5d before presentation, pneumonia, polysynovitis, pericarditis, sepsis, fever Bacteremia, Diarrhea Shipping Duodenum Necropsy B typhimurium N/A N/A 26 Paso Fino 1.5m F Chronic diarrhea. Previously admitted to VMTH 6d earlier for diarrhea (presumed sand induced). Salmonella recovered from physis at necropsy (isolate not saved) Diarrhea Chronic Antibiotic Administration Feces B typhimurium (copenhagen) N/A N/A 27 TB 3m M Dysphagic since guttural pouch fistulation surgery 1month prior. Aspiration pneumonia. Pneumonia, Diarrhea Anesthesia / Surgery, Systemic Disease Feces B j ava N/A N/A

PAGE 176

150 Table D-1. Continued Case ID Breed Age Sex Presenting Complaint Clinical Syndrome Risk Factors Specimen Origin Salmonella Group Serovar 2nd Salmonella Group 2nd Serovar 28 TB 1m M Umbilical herniorrhapy, incisional infection, hernia recurrence, repeat herniorrhapy Post-Op (noncolic) Diarrhea Anesthesia / Surgery Feces B j ava N/A N/A 29 Arab 2y M Splinters discovered in coronary band (puncture wound) developed colitis and laminitis during hospitalization Diarrhea, Laminitis Antibiotic Administration, Systemic Disease (non-colic), Colic Feces B j ava N/A N/A 30 QH 7y M Chronic stress colicker per owner, presented for colic / colon torsion / proximal enteritis. Rancid gastric contents pre-surgery and watery SI contents at surgery Post-Op Colic Diarrhea, Chronic Colic Colic, Anesthesia / Surgery Feces B unknown NVSL sample contaminated N/A N/A 31 Warmblood 9y M Diarrhea, diagnosed Salmonella positive 2y previous Diarrhea None Determined Feces D j aviana N/A N/A 32 TB 2y M Diarrhea, fever, given Probios 2d before presentation as therapy Diarrhea, Fever Antibiotic Administration Feces C1 braenderup N/A N/A 33 Mix Breed 14y F SI entrapment unidentified Colic No Diarrhea Colic SI Necropsy B j ava N/A N/A 34 QH 11y F Colic large colon impaction Colic Diarrhea Colic Feces C2 newport N/A N/A 35 QH 14y F Diarrhea, presented to VMTH 7d. prior for large colon displacement, several passages of NG tube and developed diarrhea. Given Probios day of admission. Colic Diarrhea Colic Feces E anatum N/A N/A

PAGE 177

151 Table D-1. Continued Case ID Breed Age Sex Presenting Complaint Clinical Syndrome Risk Factors Specimen Origin Salmonella Group Serovar 2nd Salmonella Group 2nd Serovar 36 Miniature Horse 7y M Fever, diarrhea, anorexia Diarrhea, Fever None Determined Feces F rubishlaw C2 muenchen 37 QH 3m M Colic (SI volvulus), chronic diarrhea for 3 mo. Colic Diarrhea, Chronic Diarrhea Anesthesia / Surgery, Colic Feces C2 muenchen B Multiple serovars (NVSL sample) 38 Arab 20y F Chronic colic Chronic Colic No Diarrhea Colic Feces D j aviana N/A N/A 39 Miniature Horse 6y F Diarrhea, fever. "Every summer gets intermittent diarrhea." Diarrhea, Fever None Determined Feces C2 newport C1 hartford 40 Paso Fino 2y F Diarrhea. Admitted to VMTH 2 weeks prior with mild pleuropneumonia, treated with antibiotics Diarrhea Antibiotic Administration, Systemic Disease Feces B typhimurium (copenhagen) N/A N/A 41 Draft Breed 6y M Colic (ileal impaction), corrective surgery Post-Op Colic Diarrhea Colic, Anesthesia / Surgery Feces B j ava N/A N/A 42 QH 4m M Presented for B. bronchiseptica pneumonia approx. 2 weeks prior. Represented for diarrhea Diarrhea Systemic Disease (non-colic) Feces B unknown (NVSL sample contaminated) N/A N/A 43 Paso Fino 4m M Diarrhea. Diagnosed Salmonella positive horse stalled next to mare and foal at farm. Diarrhea started 4d after treatment for respiratory disease initiated. Diarrhea Antibiotic Administration, Exposure to known Salmonella + Horse Feces B typhimurium (copenhagen) N/A N/A 44 Standardbred 3y M Fever, pneumonia. Developed small colon and cecal impaction 2d after presentation. Colic Diarrhea, Fever Colic, Antibiotic Administration, Systemic Disease (non-colic) Feces B typhimurium (copenhagen) N/A N/A

PAGE 178

152 Table D-1. Continued Case ID Breed Age Sex Presenting Complaint Clinical Syndrome Risk Factors Specimen Origin Salmonella Group Serovar 2nd Salmonella Group 2nd Serovar 45 Pony 3y F Colic PF impaction Colic Diarrhea Colic Feces E london N/A N/A 46 Unknown 38y LI Necropsy C2 muenchen N/A N/A 47 TB 1.5y M Colic, diarrhea, fever. History of diarrhea before presentation. Received plasma during treatment Post-Op Colic Diarrhea, Fever Anesthesia / Surgery, Colic Feces C2 newport N/A N/A 48 TB 2m F Diarrhea, severe GDUD, perforated duodenum Diarrhea Colic SI Necropsy B typhimurium N/A N/A 49 Paint 1y M Laminitis, castration surgery 7d. prior, developed diarrhea and right dorsal colitis, then laminitis Diarrhea, Laminitis Anesthesia / Surgery, Colic Lung Necropsy B typhimurium N/A N/A 50 TB 3y M Presented to VMTH with leukopenia & diarrhea, transfaunation x 2, started iodochlorhydroxyquin therapy, given Probios Diarrhea Chronic Antibiotic Administration Feces B saint paul N/A N/A 51 QH 8y M Diarrhea Diarrhea None Determined Feces C1 mbandaka N/A N/A 52 TB 3m F Diarrhea, fever. Treated with metronidazole for diarrhea before admission Diarrhea, Fever Antibiotic Administration Feces C2 newport D miami 53 Paint 22y M Shipped from NC 2 weeks previously. Anorexia, cecal impaction. Colic Diarrhea Colic, Shipping Feces D miami N/A N/A 54 QH 2y F Diarrhea, fever Diarrhea, Fever Antibiotic Administration Feces B saint paul N/A N/A

PAGE 179

153 Table D-1. Continued Case ID Breed Age Sex Presenting Complaint Clinical Syndrome Risk Factors Specimen Origin Salmonella Group Serovar 2nd Salmonella Group 2nd Serovar 55 TB 1y F Surgery at VMTH for corneal stromal abscess, sent home on doxycycline. Diarrhea & fever 2d after discharge, treated with ceftiofur and metronidazole after fever started Post-Op (noncolic) Diarrhea Antibiotic Administration, Surgery/Anesthesia, Systemic Disease SI Necropsy B j ava N/A N/A 56 QH 16y M Chronic diarrhea, lymphocytic-plasmacytic enteritis, referred from AL for unresponsive diarrhea and hypoproteinemia, rectal biopsy negative for Salmonella Diarrhea Chronic Colic Feces C2 muenchen N/A N/A 57 Paso Fino 1y M Colic, weight loss, GDUD, treated with erythromycin for suspected ileus Colic Diarrhea Colic, Antibiotic Administration Feces B j ava N/A N/A 58 TB 3m F Diarrhea, fever, treated with metronidazole for diarrhea before admission Diarrhea, Fever Antibiotic Administration Feces D miami C2 newport 59 Arab 3y M Chronic colic, current episode, SI resection 1215ft, Salmonella isolated from SI at surgery. TMPS given 3m prior just before first colic episode occurred Post-Op Colic No Diarrhea, Chronic Colic Colic, Anesthesia / Surgery, Antibiotic Administration Feces / SI B j ava N/A N/A 60 Unknown 3y M Surgical patient from referral hospital Anesthesia / Surgery Feces B saint paul N/A N/A 61 Miniature Horse 3m M Congenital cataracts, cataract surgery, history of pneumonia prior to surgery Post-Op (noncolic) Diarrhea Anesthesia / Surgery, Systemic Disease (non-colic) Feces B j ava N/A N/A

PAGE 180

154 Table D-1. Continued Case ID Breed Age Sex Presenting Complaint Clinical Syndrome Risk Factors Specimen Origin Salmonella Group Serovar 2nd Salmonella Group 2nd Serovar 62 TB 5m F Diarrhea, pneumonia Diarrhea, Pneumonia Systemic Disease (non-colic), Antibiotic Administration Feces D miami N/A N/A 63 TB 3m M Diarrhea, fever, treated with penicillin G for fever Diarrhea, Fever Antibiotic Administration Feces E london N/A N/A 64 Unknown 6y M Biopsy Rectal E meleagridis N/A N/A 65 QH 3y M Colic – right dorsal displacement Post-Op Colic Diarrhea Colic, Anesthesia / Surgery Feces D j aviana N/A N/A 66 TB 3m F Hepatic encephalopathy, endocarditis, fever, strangles recovery Fever HE Systemic Disease (non-colic) Feces C2 newport N/A N/A 67 Unknown 6y F Fever, diarrhea Diarrhea Fever Feces D miami N/A N/A 68 Pony 3m M Fever, diarrhea, respiratory infection Diarrhea Fever Antibiotic Administration, Systemic Disease Feces C2 muenchen N/A N/A 69 TB 6y M Colic, large colon impaction / displacement Post-Op Colic Diarrhea Colic, Anesthesia / Surgery Feces D j aviana B j ava 70 TB 6y M Colic, large colon impaction / displacement Post-Op Colic Diarrhea Colic, Anesthesia / Surgery Feces B j ava D j aviana 71 Miniature Horse 3m M Anorexia, depression, diarrhea Diarrhea Feces D miami B j ava 72 TB 3m M Rotavirus +, fever, diarrhea, GDUD, colic Colic Diarrhea, Fever Colic, Systemic Disease Feces B saint paul N/A N/A 73 Miniature Horse 6y F Diarrhea, fever. "Every summer gets intermittent diarrhea" Diarrhea, Fever None Determined Feces C1 hartford C2 newport 74 Paint 1.5y M Chronic diarrhea for 4.5m prior to presentation Diarrhea Chronic None Determined Liver Necropsy C2 muenchen N/A N/A

PAGE 181

155 Table D-1. Continued Case ID Breed Age Sex Presenting Complaint Clinical Syndrome Risk Factors Specimen Origin Salmonella Group Serovar 2nd Salmonella Group 2nd Serovar 75 Miniature Horse 3m M Anorexia, depression, diarrhea Diarrhea None Determined Feces B j ava D miami 76 Paso Fino 4m F Diarrhea, developed laminitis Diarrhea, Laminitis None Determined Feces C1 infantis N/A N/A 77 Appaloosa 2m M Diarrhea, fever, de-wormed the day before diarrhea started Diarrhea Fever Deworming Feces D miami N/A N/A 78 Welsh pony 18y F Gastric impaction (persimmon fruit), gastrotomy surgery Post-Op Colic Diarrhea Colic, Anesthesia / Surgery Feces B j ava N/A N/A 79 Holsteiner 7-8y F Colic, right dorsal displacement with colonic volvulus. Isolate lost. Post-Op Colic Diarrhea Colic, Surgery / Anesthesia Feces D j aviana N/A N/A 80 Miniature Horse 7y M Fever, diarrhea, anorexia Diarrhea, Fever None Determined Feces C2 muenchen F rubishlaw 81 QH 3m M Colic (SI volvulus), chronic diarrhea 3mo. Colic Diarrhea, Chronic Diarrhea Colic, Anesthesia / Surgery Feces B multiple serovars (NVSL sample) C2 muenchen 82 Paso Fino 7m M Colic, ileocecal intussusception -NO surgery – medically treated then euthanatized Colic Diarrhea Colic SI Necropsy B saint paul N/A N/A 83 Unknown M LI Necropsy B reading N/A N/A 84 QH 2y F Chronic colic, protein losing enteropathy Chronic Colic No Diarrhea Colic Feces E anatum N/A N/A 85 TB 5y M Epiploic foramen entrapment with jejunal resection. Leaky anastomosis and peritonitis Post-Op Colic Diarrhea Colic, Anesthesia / Surgery Feces B j ava N/A N/A

PAGE 182

156 Table D-1. Continued Case ID Breed Age Sex Presenting Complaint Clinical Syndrome Risk Factors Specimen Origin Salmonella Group Serovar 2nd Salmonella Group 2nd Serovar 86 TB 7m F Severe GDUD, 5 wks ADR, treated for prior respiratory disease Pneumonia, Diarrhea Antibiotic Administration, Colic, Systemic Disease Feces B j ava N/A N/A 87 TB 3y F Chronic diarrhea, wt. loss. Treated with probiotic, gastrocote, activated charcoal, iodochlorhydroxyquin (1 dose) no response. 4d of VSL-3 (probiotic) and treated with ICH 10g PO once daily for 14d. Neg. rectal biopsy for Salmonella Diarrhea Chronic None Determined Feces C2 newport N/A N/A 88 TB 5y F Post-op gastric cannulation surgery developed pipestream diarrhea, fever, depression. Treated with enrofloxacin for 24d. due to environmental risk Post-Op (noncolic) Diarrhea, Fever Anesthesia / Surgery Feces B j ava N/A N/A 89 Unknown 6m Feces B typhimurium (copenhagen) N/A N/A 90 Paso Fino 3y M Fever, diarrhea developed 5d. after starting antibiotics Diarrhea, Fever Antibiotic Administration Feces C2 tallahassee N/A N/A 91 Oldenburg 6y M Large colon impaction, shipped from NY 1 week prior Colic No Diarrhea Shipping, Colic Feces C2 newport N/A N/A 92 Unknown 10y F Feces B j ava N/A N/A 93 Paso Fino 6y F Diarrhea, colic, pregnant (aborted fetus during hospitalization) Colic Diarrhea Colic, pregnant Feces D miami N/A N/A 94 Unknown 10y M Feces E newington N/A N/A

PAGE 183

157 Table D-1. Continued Case ID Breed Age Sex Presenting Complaint Clinical Syndrome Risk Factors Specimen Origin Salmonella Group Serovar 2nd Salmonella Group 2nd Serovar 95 Miniature Horse 4y M 2wks prior injured hind limb, given phenylbutazone and flunixin (possible overdose on phenylbutazone) Colic Diarrhea Colic, Systemic Disease Feces C2 newport N/A N/A 96 Unknown 26y M Feces B j ava N/A N/A 97 QH 2y M Small colon impaction, rectal mucosal irritation Post-Op Colic Diarrhea Colic, Anesthesia / Surgery Feces B j ava N/A N/A 98 Unknown 5y M Feces B j ava N/A N/A 99 Unknown F Feces C2 newport N/A N/A 100 Unknown Feces E anatum N/A N/A 101 Unknown 1y SI Necropsy B saint paul N/A N/A 102 Unknown Feces C2 newport N/A N/A 103 Unknown Feces E newington N/A N/A 104 Unknown Feces C1 mbandaka N/A N/A 105 Unknown Abscess Abscess D unknown N/A N/A 106 Unknown 1y C1 thompson N/A N/A 107 QH 3m M Feces D j aviana N/A N/A 108 TB 3m M Diarrhea Feces D unknown N/A N/A 109 TB 16y F Colic No Diarrhea Colic Gastric Reflux D unknown N/A N/A 110 Unknown 6m M Feces D unknown N/A N/A 111 TB 8m M Feces D j aviana N/A N/A

PAGE 184

158 Table D-1. Continued Case ID Breed Age Sex Presenting Complaint Clinical Syndrome Risk Factors Specimen Origin Salmonella Group Serovar 2nd Salmonella Group 2nd Serovar 112 Paso Fino 1.5m F Chronic diarrhea. Previously admitted to VMTH 6d earlier for diarrhea (presumed sand induced). Salmonella recovered from physis at necropsy (isolate not saved) Diarrhea Chronic Antibiotic Administration Physis B typhimurium (copenhagen) N/A N/A 113 TB 1m F Shipped from IN 5d before presentation, pneumonia, polysynovitis, pericarditis, sepsis, fever Bacteremia, Diarrhea Shipping Lung Necropsy B typhimurium N/A N/A 114 TB 1m F Shipped from IN 5d before presentation, pneumonia, polysynovitis, pericarditis, sepsis, fever Bacteremia, Diarrhea Shipping Joint Fluid B typhimurium N/A N/A 115 TB 1m F Shipped from IN 5d before presentation, pneumonia, polysynovitis, pericarditis, sepsis, fever Bacteremia, Diarrhea Shipping Blood B typhimurium N/A N/A 116 Unknown 10y M Colic Gastric Reflux E newington N/A N/A 117 Unknown 3m F Feces C2 newport N/A N/A

PAGE 185

159Table D-2. Salmonella case descriptive info rmation: serovar, date sample taken, pr esence of diarrhea, total hospitalization co st, case outcome, hospitalization days, number of positive cultures, hemato logic indices at time of positive culture, and total protein changes during hospitalization. Blank cells i ndicate missing or unavailable records. Case ID Serovar Date Sample Collected Diarrhea? Total Bill Outcome Hospital Days # Positive CulturesOut of Total Total WBC Count Neutrophil s TP at Admission TP at Death or Discharge 1 N/A Sep-99 NO $2,034.30 Lived 11 2/5 3590 1080 6.0 6.1 2 N/A Dec-99 YES Unknown 3 typhimurium Dec-99 YES $3,041.50 Euthanatized 16 5/5 3600 720 6.4 5.5 4 anatum Dec-99 YES $4,898.15 Died 5 3/5 1800 20 7.3 6.0 5 newport Jan-00 YES $3,299.20 Euthanatized 5 3/5 1690 422 6.9 5.1 6 typhimurium Feb-00 YES $3,341.20 Lived 24 2/7 18700 13000 3.9 5.4 7 typhimurium Mar-00 YES $1,111.35 Euthanatized 3 1/1 14900 9830 6.5 8 typhimurium Apr-00 YES $3,487.50 Lived 48 1/5 5130 3620 7.4 6.7 9 j ava Apr-00 NO $4,765.50 Euthanatized 10 1/1 13000 10730 6.4 7.8 10 4,5,12 : imonophasic Apr-00 YES Died 0 2/2 7700 1900 7.1 7.1 11 4,5,12 : imonophasic Apr-00 NO $1,433.10 Euthanatized 10 11300 9550 5.3 5.3 12 newport Apr-00 YES $547.25 Euthanatized 1 2060 0 7.4 7.2 13 j ava May-00 YES Lived 29 3/3 6190 4620 7.5 6.5 14 newport May-00 YES Lived 8 1/1 2700 1260 4.6 5.0 15 j ava May-00 YES $4,887.15 Lived 12 5/5 1750 720 6.2 5.2 16 j ava May-00 YES $1,869.30 Lived 9 3/4 5890 3330 5.9 6.7 17 j ava May-00 YES $2,027.20 Lived 7 5/5 1460 610 6.2 6.2 18 agona May-00 Died 0 19 typhimurium (copenhagen) May-00 YES $3,720.25 Euthanatized 13 3/4 8830 900 6.3 4.2 20 j ava Jun-00 YES $1,610.80 Lived 11 3/5 1050 60 5.8 4.7 21 4,5,12 : imonophasic Jun-00 YES $1,376.15 Lived 6 1/1 3860 2100 6.7 6.2

PAGE 186

160 Table D-2. Continued Case ID Serovar Date Sample Collected Diarrhea? Total Bill Outcome Hospital Days # Positive CulturesOut of Total Total WBC Count Neutrophil s TP at Admission TP at Death or Discharge 22 typhimurium (copenhagen) Jun-00 YES $2,412.25 Euthanatized 10 4/5 5420 380 5.5 4.0 23 typhimurium Jun-00 YES $3,070.15 Died 8 16300 10510 5.6 3.5 24 hartford Jun-00 YES (noted at necropsy) $440.80 Euthanatized 3 1/1 8010 5720 7.2 25 typhimurium Jun-00 YES $325.65 Euthanatized 0 1/1 12000 5400 7.7 7.7 26 typhimurium (copenhagen) Jun-00 YES Euthanatized 13 5/7 16300 6031 7.8 5.0 27 j ava Jun-00 YES Lived 14 5910 3690 7.0 6.2 28 j ava Jun-00 YES $1,735.70 Lived 18 4/5 9010 4770 6.3 6.3 29 j ava Jun-00 YES $8,888.60 Euthanatized 30 1/8 2100 290 7.8 3.6 30 unknown NVSL sample contaminated Jun-00 YES Lived 5 1/1 4100 2110 6.2 6.5 31 j aviana Jul-00 YES $1,000.45 Lived 5 1/5 6360 3700 7.4 6.5 32 braenderup Jul-00 YES $1,412.65 Lived 6 1/4 4370 310 5.5 5.0 33 j ava Jul-00 NO $4,471.80 Euthanatized 2 1/1 12400 6450 5.7 5.9 34 newport Jul-00 YES Lived 1/5 2060 610 5.6 4.8 35 anatum Aug-00 YES Lived 16 4/4 3440 2110 7.6 7.2 36 rubishlaw Aug-00 YES Lived 6 3/5 3070 1700 7.8 7.2 37 muenchen Aug-00 YES Lived 5/5 8610 4700 7.8 6.7 38 j aviana Sep-00 NO $911.50 Lived 4 5/5 2850 1120 7.3 39 newport Sep-00 YES Lived 10 2/5 2890 260 6.6 5.5 40 typhimurium (copenhagen) Sep-00 YES Lived 5 4/5 4260 2600 5.8 4.5 41 j ava Sep-00 YES $6,790.55 Euthanatized 9 5/5 3260 1200 6.8 5.5 42 unknown (NVSL sample contaminated) Oct-00 YES $1,081.70 Died 13 2/5 6330 3150 7.0 5.5

PAGE 187

161 Table D-2. Continued Case ID Serovar Date Sample Collected Diarrhea? Total Bill Outcome Hospital Days # Positive CulturesOut of Total Total WBC Count Neutrophil s TP at Admission TP at Death or Discharge 43 typhimurium (copenhagen) Nov-00 YES Lived 13 2/5 4390 640 7.2 5.8 44 typhimurium (copenhagen) Dec-00 YES Lived 22 1/10 3090 1100 6.0 5.8 45 london Feb-01 YES $834.80 Lived 5 1/1 4130 2190 6.7 6.5 46 muenchen Feb-01 Died 0 1/1 47 newport Mar-01 YES $3,278.65 Euthanatized 6 1/5 6380 3000 5.9 6.0 (w/plasma) 48 typhimurium Apr-01 YES $300.00 Euthanatized 1 1/1 49 typhimurium Apr-01 YES $546.10 Euthanatized 1 1/1 10600 6650 5.6 5.6 50 saint paul May-01 YES $5,312.95 Euthanatized 8 2/10 3560 1100 6.0 6.8 51 mbandaka May-01 YES Lived 13 2/4 3080 230 6.2 5.6 52 newport May-01 YES $930.70 Lived 14 4/5 34400 28500 6.3 5.6 53 miami May-01 YES Lived 6 4/4 3080 990 7.7 7.3 54 saint paul May-01 YES Lived 4 1/5 17500 13700 6.8 6.8 55 j ava Jul-01 YES $1,128.55 Euthanatized 5 1/1 3360 100 7.1 4.7 56 muenchen Jun-01 YES $2,988.20 Lived 11 5/5 3490 2050 3.1 3.4/3.7 57 j ava Jun-01 YES Lived 7 1/3 5090 2400 8.0 6.7 58 miami May-01 YES $930.70 Lived 14 4/5 34400 28500 6.3 5.6 59 j ava Jul-01 NO $5,936.40 Lived 11 2/2 5590 3710 7.1 6.5 60 saint paul Jul-01 Died Unknown Cause 0 61 j ava Jul-01 YES $7,320.40 Lived 26 5/7 4660 1490 7.0 6.0 62 miami Jul-01 YES $1,002.65 Lived 6 1/3 12000 6180 5.8 5.5 63 london Jul-01 YES $2,469.05 Lived 7 3/5 6900 700 6.3 4.7 64 meleagridis Jul-01 Died Unknown Cause 0

PAGE 188

162 Table D-2. Continued Case ID Serovar Date Sample Collected Diarrhea? Total Bill Outcome Hospital Days # Positive CulturesOut of Total Total WBC Count Neutrophil s TP at Admission TP at Death or Discharge 65 j aviana Jul-01 YES $2,616.70 Lived 10 1/1 4990 2850 6.1 6.0 66 newport Aug-01 NO $1,281.50 Euthanatized 2 15300 11020 6.9 6.7 67 miami Aug-01 YES Lived 15 3/5 2220 570 7.8 6.0 68 muenchen Aug-01 YES $1,631.10 Lived 8 2/5 2950 0 5.2 5.8 69 j aviana Aug-01 YES $3,769.85 Lived 8 3/5 4070 2770 6.2 7.1 70 j ava Aug-01 YES $3,769.85 Lived 8 3/5 4070 2770 6.2 7.1 71 miami Aug-01 YES $1,453.15 Lived 5 1/5 3460 470 5.8 7.0 72 saint paul Aug-01 YES $3,152.35 Lived 18 1/5 13400 10180 6.2 6.0 73 hartford Sep-00 YES Lived 10 2/5 2890 260 6.6 5.5 74 muenchen Sep-01 YES $1,112.05 Euthanatized 3 0/5 3740 7400 7.0 7.4 75 j ava Aug-01 YES $1,453.15 Lived 5 1/5 3460 470 5.8 7.0 76 infantis Sep-01 YES $4,245.65 Lived 25 3/5 14700 12080 3.9 6.0 77 miami Sep-01 YES $2,143.00 Euthanatized 7 4/5 5240 50 5.6 3.5 78 j ava Sep-01 YES $9,309.10 Euthanatized 13 1/3 1730 180 6.5 5.2 79 j aviana Oct-00 YES Lived 9 1/1 4230 2440 6.1 7.4 80 muenchen Aug-00 YES Lived 6 3/5 3070 1700 7.8 7.2 81 multiple serovars (NVSL sample) Aug-00 YES Lived 5/5 8610 4700 7.8 6.7 82 saint paul Oct-01 YES $1,382.90 Euthanatized 6 1/1 5360 3240 6.5 6.8 83 reading Nov-01 Died Unknown Cause 84 anatum Nov-01 NO Lived 3 1/5 22100 14000 3.5 3.7 85 j ava Nov-01 YES $6,076.40 Euthanatized 7 1/2 2520 160 6.8 4.5 86 j ava Dec-01 YES $1,661.00 Euthanatized 7 2/2 4030 2760 6.2 6.0 87 newport Jan-02 YES $3,541.15 Lived 36 1/18 15200 12390 6.6 6.5 88 j ava YES Lived 14 89 typhimurium (copenhagen) Nov-01 Unknown

PAGE 189

163 Table D-2. Continued Case ID Serovar Date Sample Collected Diarrhea? Total Bill Outcome Hospital Days # Positive CulturesOut of Total Total WBC Count Neutrophil s TP at Admission TP at Death or Discharge 90 tallahassee Dec-01 YES $1,470.35 Euthanatized 1 2/2 7250 260 5.3 5.6 91 newport Dec-01 NO Lived 4 2/2 4860 3330 6.4 8.2 92 j ava Dec-01 Unknown 93 miami Dec-01 YES Lived 9 1/5 1670 90 6.1 5.4 94 newington Jan-02 Unknown 95 newport Jan-02 YES Lived 13 6/8 7960 5060 5.5 5.5 96 j ava Jan-02 Unknown 97 j ava Jan-02 YES Euthanatized 7 1/5 5710 915 6.2 3.5 98 j ava May-02 Unknown 99 newport Jun-02 Unknown 100 anatum Oct-02 Unknown 101 saint paul Oct-02 Unknown 102 newport Oct-02 Unknown 103 newington Oct-02 Unknown 104 mbandaka Oct-02 Unknown 105 N/A Mar-00 NO Lived 0 1/1 106 thompson Nov-01 Unknown 107 j aviana Jun-99 Unknown 108 N/A Jun-99 YES Unknown 109 N/A Jun-99 NO Unknown 110 N/A Oct-99 Unknown 111 j aviana Aug-99 Unknown 112 typhimurium (copenhagen) Jul-00 YES Euthanatized 13 1/1 16300 6031 7.8 5.0 113 typhimurium Jun-00 YES $325.65 Euthanatized 1 1/1 12000 5400 7.7 7.7 114 typhimurium Jun-00 YES $325.65 Euthanatized 1 1/1 12000 5400 7.7 7.7

PAGE 190

164 Table D-2. Continued Case ID Serovar Date Sample Collected Diarrhea? Total Bill Outcome Hospital Days # Positive CulturesOut of Total Total WBC Count Neutrophil s TP at Admission TP at Death or Discharge 115 typhimurium Jun-00 YES $325.65 Euthanatized 1 1/1 12000 5400 7.7 7.7 116 newington Jan-02 Unknown 117 newport May-02 Unknown MEANS Jan-Mar = 13 Apr-Jun = 42 Jul-Sep = 37 Oct-Dec = 24 22 Unkn 11 No 84 Yes $2623.23 9.25 67.93% 7270.11 4032.12 6.45 5.93

PAGE 191

165Table D-3. Salmonella case descriptive info rmation: serovar, antibiotic therapy prio r to admission and types, antibiotic thera py during hospitalization and types. Drugs in boldface type were used specifically to treat the salmonella infection. Blank cells indic ate missing or unavailable records, and three dashes indicates that nothing could be determined from the record. Case ID Serovar Antibiotic Therapy Prior to Admission? Pre-Hospitalization Antibiotic(s) Antibiotic Therapy During Hospitalization? Hospitalization Antibiotics 1 N/A No --Yes Enrofloxacin Metronidazole 2 N/A 3 typhimurium No --Yes Metronidazole Cefazolin Gentamicin Enrofloxacin 4 anatum No --Yes Gentamicin Cefazolin Metronidazole 5 newport Undetermined Yes Cefazolin Gentamicin Metronidazole 6 typhimurium No --Yes Ceftiofur Enrofloxacin 7 typhimurium Undetermined Yes Amikacin Cefazolin 8 typhimurium No --Yes Metronidazole Enrofloxacin 9 j ava Undetermined Yes Cefazolin Gentamicin 10 4,5,12 : i-monophasic No --No --11 4,5,12 : i-monophasic Yes Trimethoprim Sulfamethoxazole Rifampin Cefazolin Azithromycin Yes Cefazolin Gentamicin 12 newport Undetermined Yes Cefazolin Metronidazole

PAGE 192

166 Table D-3. Continued Case ID Serovar Antibiotic Therapy Prior to Admission? Pre-Hospitalization Antibiotic(s) Antibiotic Therapy During Hospitalization? Hospitalization Antibiotics 13 j ava Undetermined Yes Cefazolin Gentamicin Metronidazole Enrofloxacin Trimethoprim Sulfamethoxazole 14 newport Yes Penicillin Trimethoprim Sulfamethoxazole Yes Gentamicin Metronidazole Ampicillin Enrofloxacin 15 j ava Yes Penicillin Yes Metronidazole Cefazolin Gentamicin Enrofloxacin 16 j ava Yes Yes Cefazolin Gentamicin Metronidazole Ceftiofur 17 j ava Yes Trimethoprim Sulfamethoxazole Metronidazole Yes Metronidazole 18 agona 19 typhimurium (copenhagen) Yes Penicillin Yes Metronidazole Cefazolin Gentamicin Enrofloxacin 20 j ava Yes Gentamicin Cefazolin Metronidazole Yes Gentamicin Cefazolin Metronidazole 21 4,5,12 : i-monophasic Undetermined Yes Ampicillin Metronidazole Amikacin

PAGE 193

167 Table D-3. Continued Case ID Serovar Antibiotic Therapy Prior to Admission? Pre-Hospitalization Antibiotic(s) Antibiotic Therapy During Hospitalization? Hospitalization Antibiotics 22 typhimurium (copenhagen) No --Yes Cefazolin Gentamicin Metronidazole Enrofloxacin 23 typhimurium Undetermined Yes Gentamicin Penicillin Metronidazole Enrofloxacin 24 hartford Yes Amikacin Penicillin No --25 typhimurium No --No --26 typhimurium (copenhagen) Yes Metronidazole Yes Cefazolin Metronidazole Gentamicin Enrofloxacin 27 j ava Yes Trimethoprim Sulfamethoxazole Cefazolin Metronidazole Gentamicin Yes Cefazolin Metronidazole Gentamicin Enrofloxacin 28 j ava Yes Trimethoprim Sulfamethoxazole Yes Cefazolin Gentamicin 29 j ava Yes Gentamicin Yes Cefazolin Gentamicin Enrofloxacin Metronidazole 30 unknown NVSL sample contaminated Undetermined Yes Cefazolin Gentamicin Metronidazole 31 j aviana Undetermined No --32 braenderup Yes Gentamicin Amikacin Yes Metronidazole 33 j ava No --Undetermined Cefazolin Gentamicin

PAGE 194

168 Table D-3. Continued Case ID Serovar Antibiotic Therapy Prior to Admission? Pre-Hospitalization Antibiotic(s) Antibiotic Therapy During Hospitalization? Hospitalization Antibiotics 34 newport Undetermined No --35 anatum Undetermined Yes Metronidazole Ceftiofur 36 rubishlaw No --Yes Cefazolin Gentamicin 37 muenchen Undetermined Yes Penicillin Gentamicin Metronidazole 38 j aviana Undetermined No --39 newport Undetermined Yes Metronidazole 40 typhimurium (copenhagen) Yes Trimethoprim Sulfamethoxazole Metronidazole Penicillin Gentamicin Yes Metronidazole 41 j ava Undetermined Yes Penicillin Gentamicin Metronidazole 42 unknown (NVSL sample contaminated) Yes Trimethoprim Sulfamethoxazole Erythromycin Rifampin Metronidazole Gentamicin Yes Enrofloxacin 43 typhimurium (copenhagen) Yes Trimethoprim Sulfamethoxazole Yes Metronidazole Penicillin Amikacin 44 typhimurium (copenhagen) Yes Gentamicin Penicillin Yes Metronidazole Penicillin Amikacin 45 london Undetermined No --46 muenchen Undetermined 47 newport Undetermined Yes Penicillin Gentamicin Metronidazole

PAGE 195

169 Table D-3. Continued Case ID Serovar Antibiotic Therapy Prior to Admission? Pre-Hospitalization Antibiotic(s) Antibiotic Therapy During Hospitalization? Hospitalization Antibiotics 48 typhimurium Undetermined No --49 typhimurium Yes Tetracycline No --50 saint paul Yes Enrofloxacin Metronidazole Penicillin Yes Enrofloxacin Metronidazole 51 mbandaka Undetermined Yes Enrofloxacin Metronidazole 52 newport Yes Metronidazole Yes Penicillin Gentamicin Metronidazole Rifampin 53 miami Undetermined Yes Gentamicin Penicillin Metronidazole 54 saint paul Yes Metronidazole Yes Metronidazole 55 j ava Yes Doxycycline Yes Gentamicin Metronidazole Penicillin 56 muenchen Yes Metronidazole Yes Enrofloxacin 57 j ava No --No --58 miami Yes Metronidazole Yes Penicillin Gentamicin Metronidazole Rifampin 59 j ava Yes Trimethoprim Sulfamethoxazole Yes Penicillin Gentamicin Enrofloxacin 60 saint paul 61 j ava Yes Ceftiofur Yes Trimethoprim Sulfamethoxazole Penicillin Ceftiofur Amikacin

PAGE 196

170 Table D-3. Continued Case ID Serovar Antibiotic Therapy Prior to Admission? Pre-Hospitalization Antibiotic(s) Antibiotic Therapy During Hospitalization? Hospitalization Antibiotics 62 miami Yes Penicillin Gentamicin Rifampin Yes Penicillin Gentamicin Metronidazole Enrofloxacin 63 london Yes Penicillin Yes Metronidazole Gentamicin Penicillin 64 meleagridis 65 j aviana No --Yes Gentamicin Penicillin 66 newport Yes Penicillin Gentamicin Yes Penicillin Metronidazole Ceftiofur 67 miami Yes Penicillin Gentamicin Metronidazole Enrofloxacin 68 muenchen Yes Yes Metronidazole Gentamicin Penicillin 69 j aviana No --Yes Penicillin Gentamicin 70 j ava No --Yes Penicillin Gentamicin 71 miami Undetermined Yes Penicillin Metronidazole Amikacin 72 saint paul Undetermined Yes Metronidazole Trimethoprim Sulfamethoxazole Enrofloxacin 73 hartford Undetermined Yes Metronidazole 74 muenchen Undetermined Yes Metronidazole

PAGE 197

171 Table D-3. Continued Case ID Serovar Antibiotic Therapy Prior to Admission? Pre-Hospitalization Antibiotic(s) Antibiotic Therapy During Hospitalization? Hospitalization Antibiotics 75 j ava Undetermined Yes Penicillin Metronidazole Amikacin 76 infantis Undetermined Yes Enrofloxacin Metronidazole Penicillin Gentamicin 77 miami No --Yes Penicillin Gentamicin Metronidazole 78 j ava No --Yes Penicillin Gentamicin Enrofloxacin Metronidazole 79 j aviana Undetermined Yes Penicillin Gentamicin 80 muenchen No --Yes Cefazolin Gentamicin 81 multiple serovars (NVSL sample) Undetermined Yes Penicillin Gentamicin Metronidazole 82 saint paul Undetermined Yes Penicillin Gentamicin 83 reading 84 anatum No --No --85 j ava Undetermined Yes 86 j ava Yes Gentamicin Penicillin Ceftiofur No 87 newport No --Yes Metronidazole Enrofloxacin 88 j ava No --Yes Enrofloxacin

PAGE 198

172 Table D-3. Continued Case ID Serovar Antibiotic Therapy Prior to Admission? Pre-Hospitalization Antibiotic(s) Antibiotic Therapy During Hospitalization? Hospitalization Antibiotics 89 typhimurium (copenhagen) 90 tallahassee Yes Penicillin G Gentamicin Yes Metronidazole 91 newport No --No --92 j ava 93 miami Yes Gentamicin Penicillin Metronidazole 94 newington 95 newport No --Yes Metronidazole Enrofloxacin 96 j ava 97 j ava Undetermined Yes Gentamicin Penicillin Metronidazole 98 j ava 99 newport 100 anatum 101 saint paul 102 newport 103 newington 104 mbandaka 105 N/A 106 thompson 107 j aviana 108 N/A 109 N/A 110 N/A

PAGE 199

173 Table D-3. Continued Case ID Serovar Antibiotic Therapy Prior to Admission? Pre-Hospitalization Antibiotic(s) Antibiotic Therapy During Hospitalization? Hospitalization Antibiotics 111 j aviana 112 typhimurium (copenhagen) Yes Metronidazole Yes Cefazolin Metronidazole Gentamicin Enrofloxacin 113 typhimurium No --No --114 typhimurium No --No --115 typhimurium No --No --116 newington 117 newport

PAGE 200

174 APPENDIX E SALMONELLA ISOLATE ANTIMICRO BIAL SUSCEPTIBILITY DATA Tables E1 through E3 detail all of the in vitro susceptibility data for 108 salmonella isolates obtained from horses.

PAGE 201

175Table E-1. Salmonella isolate MIC antibiotic sensitivity profiles. Blank cells indicate missing data. Legend: AMI = amikacin AMOX = amoxicillin-clavulanic acid, AMP = ampicillin, CEFA = cefazolin, CEFZ = ceftazidime, NAX = ceftiofur, CHLP = chloramphenicol. Case ID MIC AMI AMI MIC AMOX AMOX MIC AMP AMP MIC CEFA CEFA MIC CEFZ CEFZ MIC NAX NAX MIC CHLP CHLP 1 <=2 S 8 S 2 S 4 S 0.5000 S <=0.5000 S 8 S 2 <=2 S <=2 S 2 S <=2 S 0.5000 S <=0.5000 S 8 S 3 <=2 S <=2 S 1 S <=2 S <=0.2500 S 1 S 8 S 4 <=2 S <=2 S 2 S <=2 S 0.5000 S 1 S 8 S 5 <=2 S <=2 S 2 S <=2 S 0.5000 S 1 S 8 S 6 <=2 S <=2 S 4 S <=2 S 0.5000 S 1 S 16 I 7 <=2 S <=2 S 2 S <=2 S 0.5000 S 1 S 8 S 8 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.2500 S <=4 S 13 <=2 S >16 R >16 R >16 R 32 R >4 R <=4 S 14 <=2 S >16 R >16 R >16 R 32 R >4 R >32 R 15 <=2 S >16 R >16 R >16 R 16 I >4 R <=4 S 16 <=2 S >16 R >16 R >16 R 32 R >4 R <=4 S 17 <=2 S >16 R >16 R >16 R 32 R >4 R <=4 S 19 4 S >16 R >16 R >16 R 32 R >4 R >32 R 20 <=2 S >16 R >16 R >16 R 32 R >4 R <=4 S 21 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S 8 S 22 <=2 S >16 R >16 R >16 R 32 R >4 R >32 R 23 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S 8 S 24 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S 8 S 25 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 26 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 27 4 S >16 R >16 R >16 R 32 R >4 R <=4 S 28 <=2 S >16 R >4 R 29 <=2 S >16 R >16 R >16 R 16 I >4 R >32 R 30 <=2 S >16 R >16 R >16 R 16 I >4 R <=4 S 31 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S 8 S 32 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S 8 S 33 <=2 S >16 R >16 R >16 R 32 R >4 R <=4 S 34 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 35 <=2 S <=2 S 1 S <=2 S 0.5000 S <=0.5000 S 8 S

PAGE 202

176 Table E-1. Continued Case ID MIC AMI AMI MIC AMOX AMOX MIC AMP AMP MIC CEFA CEFA MIC CEFZ CEFZ MIC NAX NAX MIC CHLP CHLP 36 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S 8 S 37 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 38 <=2 S <=2 S 1 S <=2 S 4 S <=0.5000 S <=4 S 39 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 40 <=2 S >16 R >16 R >16 R 32 R >4 R >32 R 41 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 42 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S <=4 S 43 <=2 S >16 R >16 R >16 R 32 R >4 R >32 R 44 <=2 S >16 R >16 R >16 R 32 R >4 R >32 R 45 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S <=4 S 46 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 47 <=2 S <=2 S 2 S <=2 S <=0.2500 S 1 S 8 S 48 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 49 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 50 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 51 <=2 S <=2 S 1 S <=2 S 0.5000 S <=0.5000 S 8 S 52 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 53 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S <=4 S 54 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S <=4 S 55 4 S >16 R >16 R >16 R 32 R >4 R 8 S 56 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S <=4 S 57 >16 R >16 R >16 R >16 R 32 R >4 R <=4 S 58 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 59 <=2 S >16 R >16 R >16 R 16 I >4 R <=4 S 60 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S 8 S 61 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 62 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S 8 S 63 4 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 64 <=2 S <=2 S 2 S <=2 S 0.5000 S <=0.5000 S <=4 S 65 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S <=4 S 66 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 67 <=2 S <=2 S 2 S <=2 S 0.5000 S <=0.5000 S <=4 S 68 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S <=4 S 69 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S

PAGE 203

177 Table E-1. Continued Case ID MIC AMI AMI MIC AMOX AMOX MIC AMP AMP MIC CEFA CEFA MIC CEFZ CEFZ MIC NAX NAX MIC CHLP CHLP 70 <=2 S >16 R >16 R >16 R 16 I >4 R <=4 S 71 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S 8 S 72 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S 8 S 73 <=2 S <=2 S 0.5000 S <=2 S <=0.2500 S <=0.5000 S <=4 S 74 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 75 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 76 <=2 S <=2 S 2 S <=2 S 0.5000 S <=0.5000 S 8 S 77 <=2 S <=2 S 2 S <=2 S <=0.2500 S <=0.5000 S <=4 S 78 <=2 S >16 R >16 R >16 R 32 R >4 R 32 R 79 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 81 <=2 S <=2 S 1 S <=2 S 0.5000 S <=0.5000 S 8 S 83 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S 8 S 84 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 84 <=2 S <=2 S 1 S <=2 S 0.5000 S <=0.5000 S 8 S 85 <=2 S >16 R >16 R >16 R 16 I >4 R <=4 S 86 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5 S 8 S 87 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5 S <=4 S 89 <=2 S <=2 S 1 S <=2 S <=0.2500 S 1 S <=4 S 90 <=2 S <=2 S 2 S <=2 S <=0.2500 S 1 S 8 S 91 4 S <=2 S 1 S <=2 S <=0.2500 S <=0.500 S <=4 S 92 <=2 S >16 R >16 R >16 R 16 I >4 R <=4 S 93 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.500 S <=4 S 94 <=2 S <=2 S 1 S <=2 S 0.5000 S 1 S <=4 S 95 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5 S <=4 S 96 <=2 S >16 R >16 R >16 R 16 I >4 R <=4 S 97 <=2 S >16 R >16 R >16 R 16 I >4 R <=4 S 98 <=2 S >16 R >16 R >16 R 16 I >4 R <=4 S 99 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5 S <=4 S 100 <=2 S <=1 S <=2 S <=2 S S <=8 S 101 <=2 S 2 S <=2 S <=2 S <=8 S 102 8 <=1 S <=2 S <=2 S S <=8 S 103 8 2 S <=2 S <=2 S S <=8 S 104 8 <=1 S <=2 S <=1 S S <=8 S 105 >16 R >16 R >16 R 8 S 8 S 1 S >32 R

PAGE 204

178 Table E-1. Continued Case ID MIC AMI AMI MIC AMOX AMOX MIC AMP AMP MIC CEFA CEFA MIC CEFZ CEFZ MIC NAX NAX MIC CHLP CHLP 106 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 107 >16 R >16 R >16 R 16 I 32 R >4 R >32 R 108 >16 R >16 R >16 R >16 R >32 R >4 R >32 R 109 >16 R >16 R >16 R 16 I 32 R >4 R >32 R 110 >16 R >16 R >16 R >16 R 32 R >4 R >32 R 111 >16 R >16 R >16 R >16 R >32 R >4 R >32 R 115 <=2 S <=2 S 1 S <=2 S <=0.2500 S <=0.5000 S <=4 S 117 <=2 S <=2 S 1 S <=2 S <=0.25 S <=0.5 S <=4 S 118 <=2 S <=2 S 0.5000 S <=2 S <=0.2500 S <=0.5000 S <=4 S 119 <=2 S >16 R >16 R >16 R 16 I >4 R <=4 S

PAGE 205

179Table E-2. Salmonella isolate MIC antibiotic sensitivity profiles. Bla nk cells indicate missing data. Legend: CLIN = clindamycin, DOX = doxycycline, ENRO = enrofloxacin, ERYT = eryt hromycin, GENT = gentamicin, IMIP = imipenem. Case ID MIC CLIN CLIN MIC DOX DOX MIC ENRO ENRO MIC ERYT ERYT MIC GENT GENT MIC IMIP IMIP 1 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 2 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 3 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 4 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 5 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 6 >2 R >4 R <=0.2500 S >4 R <=1 S <=1 S 7 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 8 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 13 >2 R >4 R <=0.2500 S >4 R 4 S <=1 S 14 >2 R >4 R <=0.2500 S >4 R <=1 S <=1 S 15 >2 R >4 R <=0.2500 S >4 R 8 I <=1 S 16 >2 R >4 R <=0.2500 S >4 R 8 I <=1 S 17 >2 R >4 R <=0.2500 S >4 R 4 S <=1 S 19 >2 R >4 R <=0.2500 S >4 R >8 R <=1 S 20 >2 R >4 R <=0.2500 S >4 R 8 I <=1 S 21 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 22 >2 R >4 R <=0.2500 S >4 R >8 R <=1 S 23 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 24 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 25 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 26 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 27 >2 R >4 R <=0.2500 S >4 R >8 R <=1 S 28 <=0.2500 S 4 S 29 >2 R >4 R <=0.2500 S >4 R >8 R <=1 S 30 >2 R >4 R <=0.2500 S >4 R 4 S <=1 S 31 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 32 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 33 >2 R >4 R <=0.2500 S >4 R 8 I <=1 S 34 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 35 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 36 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S

PAGE 206

180 Table E-2. Continued Case ID MIC CLIN CLIN MIC DOX DOX MIC ENRO ENRO MIC ERYT ERYT MIC GENT GENT MIC IMIP IMIP 37 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 38 <=0.2500 S >4 R <=0.2500 S <=0.2500 S <=1 S <=1 S 39 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 40 >2 R >4 R <=0.2500 S >4 R >8 R <=1 S 41 >2 R >4 R <=0.2500 S >4 R 8 I <=1 S 42 >2 R 4 S <=0.2500 S >4 R 4 S <=1 S 43 >2 R >4 R <=0.2500 S >4 R >8 R <=1 S 44 >2 R >4 R <=0.2500 S >4 R >8 R <=1 S 45 >2 R >4 R <=0.2500 S >4 R <=1 S <=1 S 46 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 47 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 48 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 49 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 50 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 51 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 52 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 53 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 54 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 55 >4 R >4 R <=0.2500 S >4 R 8 I <=1 S 56 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 57 >2 R >4 R 0.5000 S >4 R >8 R 2 S 58 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 59 >2 R >4 R <=0.2500 S >4 R 4 S <=1 S 60 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 61 >2 R >4 R <=0.2500 S >4 R 8 I <=1 S 62 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 63 >2 R >4 R <=0.2500 S >4 R <=1 S <=1 S 64 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 65 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 66 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 67 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 68 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 69 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 70 >2 R >4 R <=0.2500 S >4 R 8 I <=1 S

PAGE 207

181 Table E-2. Continued Case ID MIC CLIN CLIN MIC DOX DOX MIC ENRO ENRO MIC ERYT ERYT MIC GENT GENT MIC IMIP IMIP 71 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 72 >2 R >4 R <=0.2500 S >4 R 4 S <=1 S 73 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 74 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 75 >2 R >4 R <=0.2500 S >4 R 2 S <=1 S 76 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 77 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 78 >2 R >4 R 1 I >4 R >8 R <=1 S 79 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 81 >2 R >4 R <=0.2500 S >4 R <=1 S <=1 S 83 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 84 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 84 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 85 >2 R >4 S <=0.2500 S >4 R 8 I <=1 S 86 >2 R >4 R <=0.2500 S >4 R 4 S <=1 S 87 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 89 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 90 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 91 >2 R >4 R <=0.2500 S >4 R <=1 S <=1 S 92 >2 R >4 R <=0.2500 S >4 R 4 S <=1 S 93 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 94 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 95 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 96 >2 R >4 R <=0.2500 S >4 R 8 I <=1 S 97 >2 R >4 R <=0.2500 S >4 R 4 S <=1 S 98 >2 R >4 R <=0.2500 S >4 R 8 I <=1 S 99 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 100 S <=1 S <=1 S 101 S <=1 S <=1 S 102 S <=1 S <=1 S 103 S <=1 S <=1 S 104 S <=1 S <=1 S 105 >2 R >4 R <=0.2500 S >4 R >8 R <=1 S 106 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S

PAGE 208

182 Table E-2. Continued Case ID MIC CLIN CLIN MIC DOX DOX MIC ENRO ENRO MIC ERYT ERYT MIC GENT GENT MIC IMIP IMIP 107 >2 R >4 R <=0.2500 S >4 R >8 R <=1 S 108 >2 R >4 R <=0.2500 S >4 R >8 R <=1 S 109 >2 R >4 R <=0.2500 S >4 R >8 R <=1 S 110 >2 R >4 R <=0.2500 S >4 R >8 R <=1 S 111 >2 R >4 R <=0.2500 S >4 R >8 R <=1 S 115 >2 R 4 S <=0.2500 S >4 R <=1 S <=1 S 117 >2 R 2 S <=0.2500 S >4 R <=1 S <=1 S 118 >2 R 1 S <=0.2500 S >4 R <=1 S <=1 S 119 >2 R >4 R <=0.2500 S >4 R 4 S <=1 S

PAGE 209

183Table E-3. Salmonella isolate MIC antibiotic sensitivity profiles. Blank cells indicate missing da ta. Legend: NITR = nitrofu rantoin, OX = oxacillin, PEN = penicillin, RIF = rifampin, TET = tetracycline, TMP = trimethoprim-sulfamethoxazole. Case ID MIC NITR NITR MIC OX OX MIC PEN PEN MIC RIF RIF MIC TET TET MIC TMP TMP 1 <=32 S >4 R >16 R >4 R <=2 S <=0.2500 S 2 <=32 S >4 R 8 R 4 R <=2 S <=0.2500 S 3 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 4 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 5 <=32 S >4 R 16 R >4 R <=2 S <=0.2500 S 6 <=32 S >4 R 16 R >4 R 4 S <=0.2500 S 7 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 8 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 13 <=32 S >4 R >16 R >4 R >16 R >4 R 14 <=32 S >4 R >16 R >4 R >16 R >4 R 15 <=32 S >4 R >16 R >4 R >16 R >4 R 16 <=32 S >4 R >16 R >4 R >16 R >4 R 17 <=32 S >4 R >16 R >4 R >16 R >4 R 19 <=32 S >4 R >16 R >4 R >16 R >4 R 20 <=32 S >4 R >16 R >4 R >16 R >4 R 21 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 22 <=32 S >4 R >16 R >4 R >16 R >4 R 23 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 24 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 25 <=32 S >4 R 8 R >4 R <=2 S <=0.25 S 26 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 27 <=32 S >4 R >16 R >4 R >16 R >4 R 28 >4 R 29 <=32 S >4 R >16 R >4 R >16 R >4 R 30 <=32 S >4 R >16 R >4 R >16 R >4 R 31 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 32 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 33 <=32 S >4 R >16 R >4 R >16 R >4 R 34 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 35 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 36 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 37 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S

PAGE 210

184 Table E-3. Continued Case ID MIC NITR NITR MIC OX OX MIC PEN PEN MIC RIF RIF MIC TET TET MIC TMP TMP 38 <=32 S <=2 S 8 R <=0.5000 S >16 R >4 R 39 <=32 S >4 R 8 R >4 R <=2 S <=0.25 S 40 <=32 S >4 R >16 R >4 R >16 R >4 R 41 <=32 S >4 R 4 R >4 R >16 R >4 R 42 <=32 S >4 R 8 R >4 R <=2 S >4 R 43 <=32 S >4 R >16 R >4 R >16 R >4 R 44 <=32 S >4 R >16 R >4 R >16 R >4 R 45 <=32 S >4 R 8 R >4 R >16 R <=0.2500 S 46 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 47 <=32 S >4 R 16 R >4 R <=2 S <=0.2500 S 48 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 49 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 50 <=32 S >4 R 8 R 4 R <=2 S <=0.2500 S 51 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 52 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 53 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 54 <=32 S >4 R 16 R >4 R <=2 S <=0.2500 S 55 <=32 S >4 R >16 R >4 R >16 R >4 R 56 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 57 <=32 S >4 R >16 R >4 R >16 R >4 R 58 <=32 S >4 R 8 R 4 R <=2 S <=0.2500 S 59 <=32 S >4 R >16 R >4 R >16 R >4 R 60 <=32 S >4 R 16 R >4 R <=2 S <=0.25 S 61 <=32 S >4 R 16 R >4 R >16 R >4 R 62 <=32 S >4 R 16 R >4 R <=2 S <=0.2500 S 63 <=32 S >4 R 8 R >4 R >16 R <=0.2500 S 64 <=32 S >4 R 16 R >4 R <=2 S <=0.25 S 65 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 66 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 67 <=32 S >4 R 16 R >4 R <=2 S <=0.2500 S 68 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 69 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 70 <=32 S >4 R >16 R >4 R >16 R >4 R 71 <=32 S >4 R 16 R >4 R <=2 S <=0.2500 S

PAGE 211

185 Table E-3. Continued Case ID MIC NITR NITR MIC OX OX MIC PEN PEN MIC RIF RIF MIC TET TET MIC TMP TMP 72 <=32 S >4 R 16 R >4 R >16 R >4 R 73 <=32 S >4 R 4 R >4 R <=2 S <=0.2500 S 74 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 75 <=32 S >4 R 8 R >4 R >16 R >4 R 76 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 77 <=32 S >4 R 16 R >4 R <=2 S <=0.2500 S 78 <=32 S >4 R >16 R >4 R >16 R >4 R 79 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 81 <=32 S >4 R 8 R >4 R >16 R <=0.25 S 83 <=32 S >4 R 8 R >4 R <=2 S 0.5 S 84 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 84 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 85 <=32 S >4 R >16 R 4 R >16 R >4 R 86 <=32 S >4 R 16 R >4 R >16 R >4 R 87 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 89 <=32 S >4 R 16 R >4 R <=2 S <=0.2500 S 90 <=32 S >4 R >16 R >4 R <=2 S <=0.2500 S 91 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 92 <=32 S >4 R >16 R >4 R >16 R >4 R 93 <=32 S >4 R 16 R >4 R <=2 S <=0.2500 S 94 <=32 S >4 R 16 R >4 R <=2 S <=0.2500 S 95 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 96 <=32 S >4 R >16 R >4 R >16 R >4 R 97 <=32 S >4 R >16 R >4 R >16 R >4 R 98 <=32 S >4 R >16 R >4 R >16 R >4 R 99 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 100 <=32 S <=4 S <=0.0530 S 101 <=32 S >8 R >0.0530 R 102 <=32 S <=4 S <=0.0625 S 103 <=32 S <=4 S <=0.0530 S 104 <=32 S <=4 S <=0.0530 S 105 <=32 S >4 R >16 R >4 R >16 R >4 R 106 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 107 <=32 S >4 R >16 R >4 R >16 R >4 R

PAGE 212

186 Table E-3. Continued Case ID MIC NITR NITR MIC OX OX MIC PEN PEN MIC RIF RIF MIC TET TET MIC TMP TMP 108 <=32 S >4 R >16 >4 R >16 R >4 R 109 <=32 S >4 R >16 R >4 R >16 R >4 R 110 <=32 S >4 R >16 R >4 R >16 R >4 R 111 <=32 S >4 R >16 R >4 R >16 R >4 R 115 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 117 <=32 S >4 R 8 R >4 R <=2 S <=0.2500 S 118 <=32 S >4 R 2 R >4 R <=2 S <=0.2500 S 119 <=32 S >4 R >16 R >4 R >16 R >4 R

PAGE 213

187 APPENDIX F DESCRIPTIVE STATISTICS Table F-1. Descriptive statis tics for individual cecal SC FA measurements before and after four days of control (no) antibiotic treatment in 5 horses NMeanSD MinimumMaximum PRE Acetate 54.0320.773803.20 5.04 PRE Propionate 53.62802.644551.35 6.67 PRE Isobutyrate 55.87002.406774.18 9.94 PRE Butyrate 57.07005.237513.41 15.86 PRE Isovalerate 52.8380.294062.53 3.22 PRE Valerate 55.40204.043632.27 11.27 PRE Ethyl butyrate 58.55602.205255.96 11.64 POST Acetate 54.97001.000454.04 6.30 POST Propionate 54.28401.606902.70 5.97 POST Isobutyrate 56.90801.362805.57 9.03 POST Butyrate 58.36403.256755.46 13.72 POST Isovalerate 53.5980.505493.24 4.46 POST Valerate 56.31802.938073.86 10.15 POST Ethyl butyrate 59.6140.824948.61 10.78 Table F-2. Descriptive statis tics for individual cecal SC FA measurements in 5 horses treated with intramuscular ceftiofur sodi um at 2 mg/kg twice daily, before and after 4 days of treatment NMean SD Minimum Maximum PRE Acetate 55.0220.939774.38 6.67 PRE Propionate 55.29601.634763.01 7.52 PRE Isobutyrate 57.2480.745306.40 8.39 PRE Butyrate 510.83603.373046.04 15.07 PRE Isovalerate 53.9200.711862.99 4.96 PRE Valerate 57.51403.047364.00 11.41 PRE Ethyl butyrate 510.30003.925197.02 17.09 POST Acetate 54.9300.694874.34 6.11 POST Propionate 55.39201.140184.39 7.15 POST Isobutyrate 57.1960.994055.94 8.31 POST Butyrate 510.93202.564566.91 13.88 POST Isovalerate 53.9920.764053.14 5.04 POST Valerate 56.98601.797234.08 8.35 POST Ethyl butyrate 58.3860.899857.12 9.47

PAGE 214

188 Table F-3. Descriptive statis tics for individual cecal SC FA measurements in 5 horses treated with intravenous oxytetracyclin e at 10 mg/kg once daily, before and after 4 days of treatment NMean SD Minimum Maximum PRE Acetate 54.6000.74310 4.09 5.86 PRE Propionate 54.56402.43428 1.86 8.40 PRE Isobutyrate 56.71801.06528 5.38 8.10 PRE Butyrate 59.49804.82499 3.51 16.41 PRE Isovalerate 53.4680.27216 3.14 3.78 PRE Valerate 56.43203.00215 3.10 10.38 PRE Ethyl butyrate 58.1620.68460 7.71 9.27 POST Acetate 54.6080.36128 4.16 5.12 POST Propionate 55.37601.90697 3.68 8.65 POST Isobutyrate 56.93401.21829 5.09 8.14 POST Butyrate 510.91203.55481 8.40 17.02 POST Isovalerate 53.9820.99117 2.70 5.32 POST Valerate 57.26201.89179 5.73 10.26 POST Ethyl butyrate 57.2400.83979 6.56 8.67 Table F-4. Descriptive statis tics for individual cecal SC FA measurements in 5 horses treated with oral trimethoprim-sulfa methoxazole at 30 mg/kg twice daily, before and after 4 days of treatment NMeanSD MinimumMaximum PRE Acetate 54.3060.814883.41 5.49 PRE Propionate 54.4360.779573.64 5.50 PRE Isobutyrate 56.60601.642464.74 9.00 PRE Butyrate 59.48601.977446.90 12.40 PRE Isovalerate 54.06201.510982.55 6.50 PRE Valerate 56.23801.869635.04 9.56 PRE Ethyl butyrate 57.63801.627375.28 9.85 POST Acetate 54.4400.219664.18 4.76 POST Propionate 53.78601.487022.46 6.09 POST Isobutyrate 56.39201.246384.83 7.86 POST Butyrate 57.44202.939524.93 12.26 POST Isovalerate 53.8760.689372.84 4.58 POST Valerate 55.63602.862013.80 10.67 POST Ethyl butyrate 59.28601.936606.99 11.00

PAGE 215

189 Table F-5. Descriptive statistics for cecal protozoal counts in 5 horses treated with control (no treatment), ceftiofur sodi um, oxytetracycline, or trimethoprimsulfamethoxazole, before and after 4 days of treatment Time and Treatment NMean SD Minimum Maximum PRE Control 5 1533.2000 410.20385 966.00 2066.00 PRE Ceftiofur 5 6050.0000 4463.93392 1534.00 11950.00 PRE Oxytetracycline 5 4626.8000 4525.97715 716.00 10150.00 PRE TMPS 5 4320.8000 2842.31195 616.00 8566.00 POST Control 5 2946.8000 3049.84905 516.00 8100.00 POST Ceftiofur 5 5736.4000 2110.61005 4534.00 9500.00 POST Oxytetracycline 5 5960.0000 3484.33006 1966.00 10934.00 POST TMPS 5 5343.6000 4860.91317 1900.00 13650.00 Table F-6. Descriptive statistics for cecal pH in 5 horses treated with control (no treatment), ceftiofur sodium, oxyt etracycline, or trimethoprimsulfamethoxazole, before and after 4 days of treatment N RangeMinimumMaximumMean SE SD Var PRE Control 5 .837 6.103 6.940 6.33780.15278 .341637.117 POST Control 5 .747 5.743 6.490 6.10200.11833 .264592.070 PRE Ceftiofur 5 .573 6.560 7.133 6.74240.10717 .239628.057 POST Ceftiofur 5 .545 6.605 7.150 6.83500.10305 .230434.053 PRE TET 5 .530 6.327 6.857 6.61420.10345 .231322.054 POST TET 5 .753 6.397 7.150 6.92000.13413 .299932.090 PRE TMPS 5 1.1606.310 7.470 6.72600.20743 .463821.215 POST TMPS 5 .850 5.973 6.823 6.39320.15318 .342529.117 Table F-7. Descriptive statistics for salm onella growth in M9 supplemented with 10% sterile filtered cecal cont ents from 5 individual horses treated with oxytetracycline (TET), or trimet hoprim-sulfamethoxazole (TMPS) TET Treatment Mean Standard Deviation 95%CI N time 0 3.17E+03 2.37E+03 1067.069035 19 time 2 1.16E+04 4.53E+03 2805.479046 10 time 4 1.10E+05 9.11E+04 53827.87927 11 time 6 2.25E+05 1.15E+05 71131.04936 10 time 8 1.92E+06 1.02E+06 629226.4077 10 time 10 4.24E+06 1.99E+06 1231037.633 10 TMPS Treatment Mean Standard Deviation 95%CI N time 0 2.59E+03 1.44E+03 663.0500097 18 time 2 1.61E+04 5.40E+03 3349.823031 10 time 4 1.16E+06 3.61E+05 223516.9306 10 time 6 4.30E+06 2.00E+06 1241653.571 10 time 8 8.07E+06 1.28E+06 796435.4601 10 time 10 8.41E+06 1.89E+06 1169955.329 10

PAGE 216

190 Table F-7. Continued M9 ONLY Mean Standard Deviation 95%CI N time 0 2.50E+03 1605.739705 1573.595034 4 time 2 1.35E+04 707.1067812 979.9813937 2 time 4 8.85E+04 38890.87297 53898.97665 2 time 6 2.23E+05 94752.30868 131317.5068 2 time 8 1.55E+06 636396.1031 881983.2543 2 time 10 3.35E+06 2192031.022 3037942.32 2 Table F-8. Descriptive statistics for salm onella growth in M9 supplemented with 10% sterile filtered cecal c ontents from 5 individual hors es treated with control (no treatment) or ceftiofur sodium (NAX) CONTROL Treatment Mean Standard Deviation 95%CI N time 0 1.15E+06 581038.3426 360124.4584 10 time 2 3.09E+06 1104515.982 684573.1009 10 time 4 8.23E+06 1778295.064 1102177.773 10 time 6 8.44E+06 1557383.846 965258.1817 10 time 8 8.65E+06 1667499.792 1033507.456 10 time 10 8.74E+06 1889855.962 1171322.621 10 NAX Treatment Mean Standard Deviation 95%CI N time 0 9.51E+05 520265.9576 322458.0591 10 time 2 2.56E+06 718363.1084 445237.614 10 time 4 5.39E+06 1023555.893 634394.4705 10 time 6 5.88E+06 1054672.145 653680.1569 10 time 8 6.37E+06 1233828.729 764720.4501 10 time 10 5.84E+06 1456937.427 903002.0282 10 M9 ONLY Mean Standard Deviation 95%CI N time 0 7.25E+05 360624.4584 499790.5108 2 time 2 1.14E+06 502045.8146 695786.7895 2 time 4 3.70E+06 1131370.85 1567970.23 2 time 6 4.13E+06 1308147.545 1812965.578 2 time 8 4.55E+06 1484924.24 2057960.927 2 time 10 5.10E+06 2545584.412 3527933.017 2

PAGE 217

191 Table F-9. Descriptive statistics for all salmonella growth in LB broth supplemented with 10% sterile filtered cecal conten ts pooled from 5 horses treated with no treatment, ceftiofur (NAX), oxytetra cycline (TET), or trimethoprimsulfamethoxazole (TMPS). Units=CFU/m l, N=number of dilutions counted. Mean Standard Deviation 95%CI N CONTROL LB 7.10E+06 3535533.906 4899906.969 2 10% NO AB 6.68E+06 3066485.502 3469989.128 3 10% TMPS 1.10E+07 6668520.576 6535026.088 4 10% TET 6.75E+06 3118225.991 3528537.892 3 TIME 0 10% NAX 7.88E+06 3365016.097 3807801.885 3 CONTROL LB 2.92E+07 13510582.27 13240119.24 4 10% NO AB 2.25E+07 15719494.27 15404811.9 4 10% TMPS 1.06E+07 2157158.625 2441008.435 3 10% TET 2.16E+07 9220403.1 9035823.481 4 TIME 2 h 10% NAX 2.72E+07 18014901.24 17654268.02 4 CONTROL LB 9.95E+07 29881474.19 29283288.73 4 10% NO AB 9.48E+07 30645119.24 30031646.66 4 10% TMPS 4.74E+07 25166843.27 24663038.14 4 10% TET 5.08E+07 13998095.11 13717872.75 4 TIME 4 h 10% NAX 1.07E+08 28447187.91 27877714.86 4 CONTROL LB 1.36E+08 26171294.71 25647381.86 4 10% NO AB 1.45E+08 68819976.51 67442296.49 4 10% TMPS 8.95E+07 20327157.53 19920236.17 4 10% TET 1.66E+08 68409861.62 67040391.53 4 TIME 6 h 10% NAX 1.68E+08 93185728.52 91320280.11 4 CONTROL LB 1.83E+08 68024848.4 66663085.74 4 10% NO AB 1.80E+08 74925385.11 73425483.32 4 10% TMPS 1.54E+08 50956811.78 49936727.43 4 10% TET 1.60E+08 41579351.85 40746991.17 4 TIME 8 h 10% NAX 1.38E+08 48780964.53 47804437.6 4 CONTROL LB 1.68E+08 68454364.36 67084003.39 4 10% NO AB 1.70E+08 36278460.09 35552215.88 4 10% TMPS 2.73E+08 122904149.6 139076497.5 3 10% TET 2.11E+08 112076424.8 155327050.9 2 TIME 10 h 10% NAX 1.31E+08 22815984.89 22359240.67 4 CONTROL LB 3.38E+08 362794076.6 317996991.5 5 10% NO AB 2.62E+08 97085597.97 109860610.5 3 10% TMPS 2.16E+08 150693297.8 170522281.9 3 10% TET 2.40E+08 1381097480 1353449834 4 TIME 14 h 10% NAX 2.07E+08 50866844.15 49848560.82 4

PAGE 218

192 Table F-10. Descriptive statistics for all salmonella growth in M9 minimal medium supplemented with 10% sterile filtere d cecal contents pooled from 5 horses treated with control (no treatment) or ceftiofur sodium (NAX), oxytetracycline (TET), or trimethoprim-sulfamethoxazole (TMPS). Units for mean measurement are CFU/ml, and N=number of dilutions counted. Mean Standard Deviation 95%CI N CONTROL M9 2.52E+06 2667011.436 2337694.212 5 10% NO AB 2.11E+06 2317270.439 1716625.372 7 10% TMPS 1.56E+06 918257.4854 680241.7495 7 10% TET 1.79E+06 1343328.205 1074866.021 6 TIME 0 10% NAX 1.74E+06 1001527.809 741928.0974 7 CONTROL M9 2.29E+06 1246025.682 1092168.929 5 10% NO AB 2.22E+06 1084804.13 868007.6054 6 10% TMPS 2.89E+06 1513861.508 1211318.492 6 10% TET 2.48E+06 1295382.826 1036502.456 6 TIME 2 h 10% NAX 3.87E+06 2294268.947 2010977.218 5 CONTROL M9 7.38E+06 4891063.279 4793151.009 4 10% NO AB 1.51E+07 3879003.308 3801351.068 4 10% TMPS 9.15E+06 2143983.831 2101064.263 4 10% TET 7.70E+06 3340658.618 3273783.288 4 TIME 4 h 10% NAX 1.17E+07 1354929.272 1327805.477 4 CONTROL M9 1.37E+07 1248999.6 1223996.368 4 10% NO AB 1.93E+07 5068448.152 4966984.884 4 10% TMPS 1.21E+07 822597.512 806130.2562 4 10% TET 1.50E+07 9447927.462 9258793.122 4 TIME 6 h 10% NAX 1.95E+07 1734694.978 1699968.802 4 CONTROL M9 1.28E+07 1326649.916 1300092.234 4 10% NO AB 2.29E+07 1623524.972 1591024.264 4 10% TMPS 1.59E+07 5483611.948 5373837.68 4 10% TET 1.86E+07 2059935.274 2018698.241 4 TIME 8 h 10% NAX 1.88E+07 2609597.67 2557357.162 4 CONTROL M9 1.49E+07 3424787.098 3356227.633 4 10% NO AB 2.19E+07 5198317.035 5094253.973 4 10% TMPS 1.78E+07 1159022.577 1135820.56 4 10% TET 1.94E+07 1567109.866 1535738.511 4 TIME 10 h 10% NAX 2.09E+07 1422146.265 1393676.879 4 CONTROL M9 1.33E+07 2578597.81 2526977.876 4 10% NO AB 2.48E+07 2292015.125 2246132.177 4 TIME 12 h 10% TMPS 1.73E+07 3933933.57 3855181.703 4

PAGE 219

193 Table F-10. Continued Mean Standard Deviation 95%CI N 10% TET 1.79E+07 5102940.329 5000786.575 4 12 h 10% NAX 2.23E+07 3910136.4 3831860.919 4 CONTROL M9 1.50E+07 1838477.631 2547951.624 2 10% NO AB 2.60E+07 3394112.55 4703910.69 2 10% TMPS 1.91E+07 565685.4249 783985.115 2 10% TET 2.56E+07 919238.8155 1273975.812 2 TIME 24 h 10% NAX 2.55E+07 3252691.193 4507914.411 2 Table F-11. Descriptive statistics for sa lmonella growth in M9 minimal medium supplemented with sodium chloride, acetat e, butyrate, or propionate at 30 or 100mM Mean Standard Deviation 95%CI N SEM CONTROL 30mM NaCl 1.32E+05 2.36E+05 1.75E+05 7 89237.7 CONTROL 100mM NaCl 1.09E+05 1.52E+05 1.21E+05 6 61906.0 ACETATE 30mM 9.44E+04 1.31E+05 1.04E+05 6 53313.3 ACETATE 100mM 1.01E+05 1.41E+05 1.13E+05 6 57502.2 BUTYRATE 30mM 1.51E+05 2.54E+05 1.88E+05 7 95846.8 BUTYRATE 100mM 1.16E+05 1.95E+05 1.45E+05 7 73829.1 PROPRIONATE 30mM 1.29E+05 1.83E+05 1.46E+05 6 74610.5 TIME 0 PROPRIONATE 100mM 1.00E+05 1.38E+05 1.10E+05 6 56194.3 CONTROL 30mM NaCl 2.42E+05 2.74E+05 2.68E+05 4 136884.6 CONTROL 100mM NaCl 2.50E+05 3.15E+05 3.09E+05 4 157556.5 ACETATE 30mM 3.43E+05 5.73E+05 5.61E+05 4 286262.8 ACETATE 100mM 1.30E+05 1.80E+05 1.44E+05 6 73486.8 BUTYRATE 30mM 2.15E+05 2.70E+05 2.00E+05 7 102175.0 BUTYRATE 100mM 2.06E+05 3.18E+05 2.54E+05 6 129775.0 PROPRIONATE 30mM 1.70E+05 2.46E+05 1.96E+05 6 100236.5 TIME 2 h PROPRIONATE 100mM 1.09E+05 1.53E+05 1.22E+05 6 62433.4 4 CONTROL 30mM NaCl 4.34E+06 2.31E+06 2.27E+06 4 1156289.9

PAGE 220

194 Table F-11. Continued Mean Standard Deviation 95%CI N SEM CONTROL 100mM NaCl 4.45E+06 9.51E+05 9.32E+05 4 475576.4 ACETATE 30mM 2.22E+06 2.86E+06 2.80E+06 4 1427786.0 ACETATE 100mM 2.17E+05 2.01E+05 1.49E+05 7 75828.4 BUTYRATE 30mM 9.81E+05 1.12E+06 9.86E+05 5 502887.1 BUTYRATE 100mM 1.87E+05 2.24E+05 1.47E+05 9 83928.0 PROPRIONATE 30mM 2.41E+05 2.57E+05 1.90E+05 7 97025.2 PROPRIONATE 100mM 1.00E+05 1.61E+05 1.05E+05 9 59758.1 CONTROL 30mM NaCl 1.41E+08 1.10E+08 1.08E+08 4 54909584.8 CONTROL 100mM NaCl 1.12E+08 1.01E+08 9.90E+07 4 50504744.0 ACETATE 30mM 2.27E+07 1.83E+07 1.47E+07 6 7489537.1 ACETATE 100mM 1.22E+06 1.37E+06 1.20E+06 5 610938.6 BUTYRATE 30mM 1.24E+07 1.74E+07 1.71E+07 4 8719381.7 BUTYRATE 100mM 2.36E+05 3.75E+05 2.78E+05 7 141662.3 PROPRIONATE 30mM 1.24E+06 1.34E+06 9.95E+05 7 507786.2 TIME 6 h PROPRIONATE 100mM 1.43E+05 2.74E+05 2.03E+05 7 103549.2 CONTROL 30mM NaCl 2.18E+08 1.05E+08 8.39E+07 6 42805503.3 CONTROL 100mM NaCl 2.10E+08 9.21E+07 7.37E+07 6 37604742.6 ACETATE 30mM 9.80E+07 5.49E+07 5.38E+07 4 27452990.1 ACETATE 100mM 4.48E+06 4.82E+06 3.85E+06 6 1966734.5 BUTYRATE 30mM 4.14E+07 2.91E+07 2.85E+07 4 14551138.3 BUTYRATE 100mM 7.18E+05 9.06E+05 6.71E+05 7 342319.6 PROPRIONATE 30mM 6.65E+06 6.95E+06 5.56E+06 6 2837475.9 TIME 8 h PROPRIONATE 100mM 2.22E+05 4.61E+05 3.42E+05 7 163793.6 CONTROL 30mM NaCl 2.11E+08 2.87E+07 2.81E+07 4 14355602.9 TIME 10h CONTROL 100mM NaCl 2.10E+08 3.80E+07 3.72E+07 4 18993419.9

PAGE 221

195 Table F-11. Continued Mean Standard Deviation 95%CI N SEM ACETATE 30mM 2.11E+08 4.08E+07 4.00E+07 4 20422108.1 ACETATE 100mM 1.26E+07 1.45E+07 1.27E+07 5 6469961.4 BUTYRATE 30mM 1.08E+08 3.23E+07 3.17E+07 4 16152270.2 BUTYRATE 100mM 1.75E+06 2.30E+06 1.70E+06 7 869133.1 PROPRIONATE 30mM 1.64E+07 1.57E+07 1.37E+07 5 7012232.2 PROPRIONATE 100mM 2.82E+05 5.34E+05 3.95E+05 7 201735.8

PAGE 222

196 LIST OF REFERENCES 1. Grimont PAD, Grimont F, Bouve t P (2000), Taxonomy of the Genus Salmonella In: Wray C and Wray A eds. Salmonella in Domestic Animals. ; 1-16. 2. Popoff MY, Bockemuhl J, Gheesling LL (2003), Supplement 2001 (no. 45) to the Kauffmann-White scheme, Res Microbiol 154: 173-174. 3. Volk WA, Benjamin DC, Kadner RJ et al. (1991), Essentials of Medical Microbiology Philadelphia, PA: J.B. Lippincott Company; 400-403. 4. Santos RL, Zhang S, Tsolis RM, Kingsley RA, Adams LG, Baumler AJ (2001), Animal models of Salmonella infecti ons: enteritis versus typhoid fever, Microbes Infect 3: 1335-1344. 5. Kingsley RA, Baumler AJ (2000), Ho st adaptation and the emergence of infectious disease: the Salmonella paradigm, Mol Microbiol 36: 10061014. 6. Tsolis RM, Kingsley RA, Townsend SM, Ficht TA, Adams LG, Baumler AJ (1999), Of mice, calves, and men. Comparison of the mouse typhoid model with other Salmonella infections, Adv Exp Med Biol 473: 261-274. 7. Baumler AJ, Tsolis RM, Ficht TA, Adams LG (1998), Evolution of host adaptation in Salmonella enterica, Infect Immun 66: 4579-4587. 8. Zhang S, Kingsley RA, Santos RL Andrews-Polymenis H, Raffatellu M, Figueiredo J, Nunes J, Tsolis RM, Adams LG, Baumler AJ (2003), Molecular pathogenesis of Salmonella enterica serotype typhimuriuminduced diarrhea, Infect Immun 71: 1-12. 9. Watson PR, Paulin SM, Bland AP, Jone s PW, Wallis TS (1995), Characterization of intestinal invasion by Salmonella typhimurium and Salmonella dublin and effect of a mutation in the invH gene, Infect Immun 63: 2743-2754. 10. Libby SJ, Adams LG, Ficht TA, Allen C, Whitford HA, Buchmeier NA, Bossie S, Guiney DG (1997), The spv genes on the Salmonella dublin virulence plasmid are required for severe enteri tis and systemic infection in the natural host, Infect Immun 65: 1786-1792.

PAGE 223

197 11. Turner JL, Dritz SS, Higgins JJ, He rkelman KL, Minton JE (2002), Effects of a Quillaja saponaria extract on growth performance and immune function of weanling pigs challenged w ith Salmonella typhimurium, J Anim Sci 80: 1939-1946. 12. Trebichavsky I, Schulze J, Dlabac V, Cukrowska B, Tlaskalova-Hogenova H, Rehakova Z (1998), Salmonellosis: less ons drawn from a germ-free pig model, Folia Microbiol (Praha) 43: 697-701. 13. Barthel M, Hapfelmeier S, Quintanill a-Martinez L, Kremer M, Rohde M, Hogardt M, Pfeffer K, Russmann H, Hardt WD (2003), Pretreatment of mice with streptomycin provides a Salmonella enterica serovar Typhimurium colitis model that allows analysis of both pathogen and host, Infect Immun 71: 2839-2858. 14. Alberts B, Bray D, Lewis J (1994), Molecular Biology of the Cell New York, NY: Garland Publishing, Inc.; 286. 15. Fierer J, Guiney DG ( 2001), Diverse virulence traits underlying different clinical outcomes of Salmonella infection, J Clin Invest 107: 775-780. 16. Baumler AJ, Tsolis RM, Heffron F (2000), Virulence Mechanisms of Salmonella and their Genetic Basis. In: Wray A and Wray C eds. Salmonella in Domestic Animals. ; 57-67. 17. Sheehy RJ, Allison DP, Curtiss R, III (1973), Cryptic plasmids in a minicellproducing strain of Salmonella typhimurium, J Bacteriol 114: 439-442. 18. Jones GW, Rabert DK, Svinarich DM Whitfield HJ (1982), Association of adhesive, invasive, and virulent phenotypes of Salmonella typhimurium with autonomous 60-megadalton plasmids, Infect Immun 38: 476-486. 19. Hackett J, Kotlarski I, Mathan V, Francki K, Rowley D (1986), The colonization of Peyer's patches by a strain of Salmonella typhimurium cured of the cryptic plasmid, J Infect Dis 153: 1119-1125. 20. Gulig PA, Curtiss R, III (1987), Plasmid-associated virulence of Salmonella typhimurium, Infect Immun 55: 2891-2901. 21. Gulig PA, Doyle TJ (1993), The Sa lmonella typhimurium virulence plasmid increases the growth rate of salmonellae in mice, Infect Immun 61: 504511. 22. Fierer J, Krause M, Tauxe R, Guiney D (1992), Salmonella typhimurium bacteremia: association with the virulence plasmid, J Infect Dis 166: 639642.

PAGE 224

198 23. Montenegro MA, Morelli G, Helm uth R (1991), Heteroduplex analysis of Salmonella virulence plasmids and their prevalence in isolates of defined sources, Microb Pathog 11: 391-397. 24. Chiu CH, Chu C, Ou JT (2000), Lack of evidence of an association between the carriage of virulence plasmid and the bacteremia of Salmonella typhimurium in humans, Microbiol Immunol 44: 741-748. 25. Chiu CH, Lin TY, Ou JT (1999), Pr evalence of the virulence plasmids of nontyphoid Salmonella in the serovars isolated from humans and their association with bacteremia, Microbiol Immunol 43: 899-903. 26. Wallis TS, Paulin SM, Plested JS, Watson PR, Jones PW (1995), The Salmonella dublin virulence plasmid mediates sy stemic but not enteric phases of salmonellosis in cattle, Infect Immun 63: 2755-2761. 27. Gulig PA, Caldwell AL, Chiodo VA (1992) Identification, gene tic analysis and DNA sequence of a 7.8-kb virulence region of the Salmonella typhimurium virulence plasmid, Mol Microbiol 6: 1395-1411. 28. Gulig PA, Danbara H, Guiney DG, Lax AJ, Norel F, Rhen M (1993), Molecular analysis of spv virulence genes of the Salmonella virulence plasmids, Mol Microbiol 7: 825-830. 29. Wilson JA, Doyle TJ, Gulig PA (1997) Exponential-phase expression of spvA of the Salmonella typhimurium virulence plasmid: induction in intracellular salts medium and intracellularly in mice and cultured mammalian cells, Microbiology 143 ( Pt 12): 3827-3839. 30. Krause M, Roudier C, Fierer J, Harwood J, Guiney DG (1991), Molecular Analysis of the Virulence Locus of the Salmonella dublin Plasmid pSDL2, Molecular Microbiology 5: 307-316. 31. Matsui H, Bacot CM, Garlington WA Doyle TJ, Roberts S, Gulig PA (2001), Virulence plasmid-borne spvB and spvC genes can replace the 90-kilobase plasmid in conferring virulence to Salmonella enterica serovar Typhimurium in subcutaneously inoculated mice, J Bacteriol 183: 46524658. 32. Rhen M, Riikonen P, Taira S (1993), Tr anscriptional regulation of Salmonella enterica virulence plasmid genes in cultured macrophages, Mol Microbiol 10: 45-56. 33. Guiney DG, Fang FC, Krause M, Libby S, Buchmeier NA, Fierer J (1995), Biology and clinical significance of virulence plasmids in Salmonella serovars, Clin Infect Dis 21 Suppl 2: S146-S151.

PAGE 225

199 34. Heffernan EJ, Fierer J, Chikami G, Guiney D (1987), Natu ral history of oral Salmonella dublin infection in BALB/c mice: effect of an 80-kilobase-pair plasmid on virulence, J Infect Dis 155: 1254-1259. 35. Libby SJ, Lesnick M, Hasegawa P, Weidenhammer E, Guiney DG (2000), The Salmonella virulence plasmid spv gene s are required for cytopathology in human monocyte-derived macrophages, Cell Microbiol 2: 49-58. 36. Beninger PR, Chikami G, Tanabe K, Roudier C, Fierer J, Guiney DG (1988), Physical and genetic mapping of th e Salmonella dublin virulence plasmid pSDL2. Relationship to plasmids from other Salmonella strains, J Clin Invest 81: 1341-1347. 37. Roudier C, Krause M, Fierer J, Guiney D (1990), Correlation between the presence of sequences homologous to the vir region of Salmonella dublin plasmid pSDL2 and the virulence of twenty-two Salmonella serotypes in mice, Infect Immun 58: 1180-1185. 38. Tsolis RM, Adams LG, Ficht TA, Baum ler AJ (1999), Contribution of Salmonella typhimurium virulence factors to diarrheal disease in calves, Infect Immun 67: 4879-4885. 39. Surette MG, Bassler BL (1998), Quor um sensing in Escherichia coli and Salmonella typhimurium, Proc Natl Acad Sci U S A 95: 7046-7050. 40. Leyer GJ, Johnson EA (1993), Acid ad aptation induces cross-protection against environmental stresses in Salmonella typhimurium Appl Environ Microbiol 59: 1842-1847. 41. Elsden SR, Hitchcock MW, Marshall RA Phillipson AT (1946), Volatile acid in the digesta of ruminant s and other animals, J Exp Biol 22: 191-202. 42. Bugaut M (1987), Occurrence, absorp tion and metabolism of short chain fatty acids in the digestive tract of mammals, Comp Biochem Physiol B 86: 439-472. 43. Durant JA, Corrier DE, Byrd JA, Sta nker LH, Ricke SC (1999), Feed deprivation affects crop environment and modulates Salmonella enteriditis colonization and invasion of Leghorn hens., Appl Environ Microbiol 65: 1919-1923. 44. Hird DW, Casebolt DB, Carter JD, Pappaioanou M, Hjerpe CA (1986), Risk factors for salmonellosis in hospitalized horses, J Am Vet Med Assoc 188: 173-177. 45. Hird DW, Pappaioanou M, Smith BP (1984), Case-control study of risk factors associated with isolation of Salmone lla saintpaul in hospitalized horses, Am J Epidemiol 120: 852-864.

PAGE 226

200 46. Kim LM, Morley PS, Traub-Dargatz JL, Salman MD, Gentry-Weeks CR (2001), Factors associated with Salmonella shedding among equine colic patients at a veterinary teaching hospital, J Am Vet Med Assoc 218: 740-748. 47. House JK, Mainar-Jaime RC, Smith BP, House AM, Kamiya DY (1999), Risk factors for nosocomial Salmonella in fection among hospitalized horses, J Am Vet Med Assoc 214: 1511-1516. 48. Traub-Dargatz JL, Salman MD, Jones RL (1990), Epidemiologic study of salmonellae shedding in the feces of hor ses and potential risk factors for development of the infecti on in hospitalized horses, J Am Vet Med Assoc 196: 1617-1622. 49. Gustafsson A, Baverud V, Franklin A, Gunnarsson A, Ogren G, Ingvast-Larsson C (1999), Repeated administration of trimethoprim/sulfadiazine in the horse--pharmacokinetics, plasma prot ein binding and influence on the intestinal microflora, J Vet Pharmacol Ther 22: 20-26. 50. Larsen J (1997), Acute colitis in adult horses. A review with emphasis on aetiology and pathogenesis, Vet Q 19: 72-80. 51. White G, Prior SD (1982), Comparat ive effects of oral administration of trimethoprim/sulphadiazine or oxytetr acycline on the faecal flora of horses, Vet Rec 111: 316-318. 52. Andersson G, Ekman L, Mansson I, Persson S, Rubarth S, Tufvesson G (1971), Lethal complications following admini stration of oxytetracycline in the horse, Nord Vet Med 23: 9-22. 53. Baker JR, Leyland A (1973), Diarrhoea in the horse associated with stress and tetracycline therapy, Vet Rec 93: 583-584. 54. Owen R (1975), Post stre ss diarrhoea in the horse, Vet Rec 96: 267-270. 55. Keir AA, Stampfli HR, Crawford J ( 1999), Outbreak of acute colitis on a horse farm associated with tetracycl ine-contaminated sweet feed, Can Vet J 40: 718-720. 56. Murray MJ (2002), Diseases of the Alimentary Tract. In: Smith BP ed. Large Animal Internal Medicine: Diseases of Horses, Cattle, Sheep, and Goats. 3rd ed.; 653-661. 57. Larsen J, Dolvik NI, Teige J, Jr. (19 96), Acute post-treatment enterocolitis in 13 horses treated in a Norw egian surgical ward, Acta Vet Scand 37: 203-211. 58. Raisbeck MF, Holt GR, Osweiler GD (1981), Lincomycin-associated colitis in horses, J Am Vet Med Assoc 179: 362-363.

PAGE 227

201 59. Staempfli HR, Prescott JF, Brash ML (1992), Lincomycin-induced severe colitis in ponies: association with Clostridium cadaveris, Can J Vet Res 56: 168169. 60. Prescott JF, Staempfli HR, Barker IK Bettoni R, Delaney K (1988), A method for reproducing fatal idiopathic colitis (colit is X) in ponies and isolation of a clostridium as a possible agent, Equine Vet J 20: 417-420. 61. Ensink JM, Klein WR, Barneveld A, va n Miert AS, Vulto AG (1996), Side effects of oral antimicrobial agents in the horse: a comparison of pivampicillin and trimethoprim/sulphadiazine, Vet Rec 138: 253-256. 62. Wilson DA, MacFadden KE, Green EM, Crabill M, Frankeny RL, Thorne JG (1996), Case control and historical cohort study of diarrhea associated with administration of trimethoprimpotentiated sulphonamides to horses and ponies, J Vet Intern Med 10: 258-264. 63. Stratton-Phelps M, Wilson WD, Gardner IA (2000), Risk of adverse effects in pneumonic foals treated with erythrom ycin versus other antibiotics: 143 cases (1986-1996), J Am Vet Med Assoc 217: 68-73. 64. McCain CS, Powell KC (1990), Asymptomatic salmonellosis in healthy adult horses, J Vet Diagn Invest 2: 236-237. 65. Begg AP, Johnston KG, Hutchins DR, Edwards DJ (1988), Some aspects of the epidemiology of equine salmonellosis, Aust Vet J 65: 221-223. 66. Hogenauer C, Hammer HF, Krejs GJ Reisinger EC (1998), Mechanisms and management of antibiotic-associated diarrhea, Clin Infect Dis 27: 702-710. 67. Rantala M, Nurmi E (1973), Prevention of the growth of Salmonella infantis in chicks by the flora of the a limentary tract of chickens, Br Poult Sci 14: 627-630. 68. van den Bogaard AE, Weidema WF, van Boven CP, van der WD (1986), Recolonization and coloni zation resistance of the large bowel after three methods of preoperative preparation of the gastrointestinal tract for elective colorectal surgery, J Hyg (Lond) 97: 49-59. 69. Bourlioux P, Koletzko B, Guarner F, Braesco V (2003), The intestine and its microflora are partners for the protection of the host: report on the Danone Symposium "The Intellig ent Intestine," held in Paris, June 14, 2002, Am J Clin Nutr 78: 675-683. 70. Falk PG, Hooper LV, Midtvedt T, Gor don JI (1998), Creating and maintaining the gastrointestinal ecosystem: what we know and need to know from gnotobiology, Microbiol Mol Biol Rev 62: 1157-1170.

PAGE 228

202 71. Miller TL, Wolin MJ (1979), Fermentati ons by saccharolytic in testinal bacteria, Am J Clin Nutr 32: 164-172. 72. Mortensen PB, Holtug K, Rasmu ssen HS (1988), Short-chain fatty acid production from monoand disaccharide s in a fecal incubation system: implications for colonic fermenta tion of dietary fibre in humans, J Nutr 118: 321-325. 73. Mortensen PB, Clausen MR (1996), Shor t-chain fatty acids in the human colon: relation to gastrointestinal health and disease, Scand J Gastroenterol Suppl 216: 132-148. 74. Van Soest PJ (1994), Nutritional Ecology of the Ruminant Ithaca, NY: Cornell University Press. 75. von Engelhardt W, Bartels J, Kirschbe rger S, Meyer zu Duttingdorf HD, Busche R (1998), Role of short-chain fatty acids in the hind gut, Vet Q 20 Suppl 3: S52-S59. 76. Bergman EN (1990), Energy contributi ons of volatile fatty acids from the gastrointestinal tract in various species, Physiol Rev 70: 567-590. 77. Clausen MR, Bonnen H, Tvede M, Mort ensen PB (1991), Colonic fermentation to short-chain fatty acids is decreased in antibiotic-associated diarrhea, Gastro 101: 1497-1504. 78. Clausen MR (1998), Production and oxida tion of short-chain fatty acids in the human colon: implications for antib iotic-associated diarrhea, ulcerative colitis, colonic cancer, a nd hepatic encephalopathy, Dan Med Bull 45: 5175. 79. Bergogne-Berezin E (2000), Treatment and prevention of antibiotic associated diarrhea, Int J Antimicrob Agents 16: 521-526. 80. Que JU, Casey SW, Hentges DJ (1986), Factors responsible for increased susceptibility of mice to intestinal colonization after treatment with streptomycin, Infect Immun 53: 116-123. 81. Tsukahara T, Ushida K (2001), Organi c acid profiles in f eces of pigs with pathogenic or non-pathogenic diarrhea, J Vet Med Sci 63: 1351-1354. 82. Hoverstad T, Carlstedt-Duke B, Lingaas E, Norin E, Saxerholt H, Steinbakk M, Midtvedt T (1986), Influen ce of oral intake of se ven different antibiotics on faecal short-chain fatty acid ex cretion in healthy subjects, Scand J Gastroenterol 21: 997-1003.

PAGE 229

203 83. Bender A, Breves G, Stein J, Leonhard -Marek S, Schroder B, Winckler C (2001), Colonic fermentation as affected by an tibiotics and acidic pH: Application of an in vitro model, Z Gastroenterol 39: 911-918. 84. Oh H, Nord CE, Barkholt L, Hedberg M, Edlund C (2000), Ecological disturbances in intestinal microflo ra caused by clinafloxacin, an extendedspectrum quinolone, Infection 28: 272-277. 85. Meijer-Severs GJ, van Santen E, Me ijer BC (1990), Short-chain fatty acid and organic acid concentrations in feces of healthy human volunteers and their correlations with anaero be cultural counts during systemic ceftriaxone administration, Scand J Gastroenterol 25: 698-704. 86. Collinder E, Berge GN, Gronvold B, Lindholm A, Midtved T, Norin E (2000), Influence of bacitracin on microbial f unctions in the gastrointestinal tract of horses, Equine Vet J 32: 345-350. 87. Perrin-Guyomard A, Cottin S, Corpet DE Boisseau J, Poul JM (2001), Evaluation of residual and therapeutic doses of tetracycline in the human-floraassociated (HFA) mice model, Regul Toxicol Pharmacol 34: 125-136. 88. Edlund C, Beyer G, Hiemer-Bau M, Ziege S, Lode H, Nord CE (2000), Comparative effects of moxifloxaci n and clarithromycin on the normal intestinal microflora, Scand J Infect Dis 32: 81-85. 89. Que JU, Hentges DJ (1985), Eff ect of streptomycin administration on colonization resistance to Salmonella typhimurium in mice, Infect Immun 48: 169-174. 90. McHan F, Shotts EB (1992), Effect of feeding selected short-chain fatty acids on the in vivo attachment of Salmonella typhimurium in chick ceca, Avian Dis 36: 139-142. 91. Hove H, Tvede M, Mortensen PB (1996), Antibiotic-associated diarrhoea, Clostridium difficile, and short-chain fatty acids, Scand J Gastroenterol 31: 688-693. 92. Breves G, Faul K, Schroder B, Hols t H, Caspary WF, Stein J (2000), Application of the colon-simulation techni que for studying the effects of Saccharomyces boulardii on basic para meters of porcine cecal microbial metabolism disturbed by clindamycin, Digestion 61: 193-200. 93. Baverud V (2002), Clostridium difficile infections in animals with special reference to the horse. A review, Vet Q 24: 203-219. 94. Smith BP, Reina-Guerra M, Hardy AJ Habasha F (1979), Equine salmonellosis: experimental production of four syndromes, Am J Vet Res 40: 1072-1077.

PAGE 230

204 95. Traub-Dargatz JL, Garber LP, Fedorka -Cray PJ, Ladely S, Ferris KE (2000), Fecal shedding of Salmonella spp by horses in the United States during 1998 and 1999 and detection of Salmonella spp in grain and concentrate sources on equine operations, J Am Vet Med Assoc 217: 226-230. 96. Owen RA, Fullerton J, Barnum DA (1983) Effects of transp ortation, surgery, and antibiotic therapy in ponies infected with Salmonella, Am J Vet Res 44: 46-50. 97. Morse EV, Duncan MA, Page EA, Fessler JF (1976), Salmonellosis in Equidae: a study of 23 cases, Cornell Vet 66: 198-213. 98. Parraga ME, Spier SJ, Thurmond M, Hirs h D (1997), A clinical trial of probiotic administration for prevention of Salm onella shedding in the postoperative period in horses with colic, J Vet Intern Med 11: 36-41. 99. Carter JD, Hird DW, Farver TB, Hjer pe CA (1986), Salmonellosis in hospitalized horses: seasonality and case fatality rates, J Am Vet Med Assoc 188: 163167. 100. Gibbons DF (1980), Equine salmonellosis: a review, Vet Rec 106: 356-359. 101. Spier SJ (1993), Salmonellosis, Vet Clin North Am Equine Pract 9: 385-397. 102. Roberts MC, O'Boyle DA (1982), Experi mental Salmonella anatum infection in horses, Aust Vet J 58: 232-240. 103. Ebner PD, Mathew AG (2001), Three molecular methods to identify Salmonella enterica serotype Typhimurium DT 104: PCR fingerprinting, multiplex PCR and rapid PFGE, FEMS Microbiol Lett 205: 25-29. 104. Olsen JE (2000), Molecular Typing of Salmonella. In: Wray C and Wray A eds. Salmonella in Domestic Animals. ; 429-444. 105. Kurowski PB, Traub-Dargatz JL, Morley PS, Gentry-Weeks CR (2002), Detection of Salmonella spp. in f ecal specimens by use of real-time polymerase chain reaction assay, Am J Vet Res 63: 1265-1268. 106. Amavisit P, Browning GF, Lightfoot D, Church S, Anderson GA, Whithear KG, Markham PF (2001), Rapid PCR detecti on of Salmonella in horse faecal samples, Vet Microbiol 79: 63-74. 107. House JK, Smith BP (2000), Salmonella in Horses. In: Wray C and Wray A eds. Salmonella in Domestic Animals. ; 219-229. 108. Palmer JE, Benson CE, Whitlock RH (1985), Salmonella shed by horses with colic, J Am Vet Med Assoc 187: 256-257.

PAGE 231

205 109. Amavisit P, Markham PF, Lightfoot D, Whithear KG, Browning GF (2001), Molecular epidemiology of Salmonella Heidelberg in an equine hospital, Vet Microbiol 80: 85-98. 110. Ikeda JS, Hirsh DC, Jang SS, Biberstein EL (1986), Characteri stics of Salmonella isolated from animals at a vete rinary medical teaching hospital, Am J Vet Res 47: 232-235. 111. Bryans JT, Fallon EH, Shephard BP (1961), Equine salmonellosis, Cornell Vet 51: 467-477. 112. Madic J, Hajsig D, Sostaric B, Curic S, Seol B, Naglic T, Cvetnic Z (1997), An outbreak of abortion in mares associ ated with Salmonella abortusequi infection, Equine Vet J 29: 230-233. 113. Plym-Forshell L, Ekesbo I (1996), Surv ival of salmonellas in urine and dry faeces from cattle--an experimental study, Acta Vet Scand 37: 127-131. 114. Bornet FR, Brouns F, Tashiro Y, Duv illier V (2002), Nutriti onal aspects of shortchain fructooligosaccharides: natu ral occurrence, chemistry, physiology and health implications, Dig Liver Dis 34 Suppl 2: S111-S120. 115. Gustafsson A, Berstad A, Lund-Tonnese n S, Midtvedt T, Norin E (1999), The effect of faecal enema on five micr oflora-associated characteristics in patients with antibiotic-associated diarrhoea, Scand J Gastroenterol 34: 580-586. 116. Weese JS, Anderson ME, Lowe A, Mont eith GJ (2003), Preliminary investigation of the probiotic potentia l of Lactobacillus rhamnosus strain GG in horses: fecal recovery following oral administration and safety, Can Vet J 44: 299-302. 117. Sheoran AS, Timoney JF, Tinge SA, Sunda ram P, Curtiss R, I II (2001), Intranasal immunogenicity of a Delta cya Delt a crp-pabA mutant of Salmonella enterica serotype Typhimurium for the horse, Vaccine 19: 3787-3795. 118. Nelson JD, Kusmiesz H, Jackson LH (1980), Treatment of Salmonella gastroenteritis with ampicil lin, amoxicillin or placebo, Pediatrics 65: 1125-1130. 119. Barbara G, Stanghellini V, Berti-Ceroni C, De Giorgio R, Salvioli B, Corradi F, Cremon C, Corinaldesi R (2000), Role of antibiotic therapy on long-term germ excretion in faeces and digestive symptoms after Salmonella infection, Aliment Pharmacol Ther 14: 1127-1131. 120. van Duijkeren E, Houwers DJ (2000), A critical assessment of antimicrobial treatment in umcomplicated Salmonella enteritis, Vet Microbiol 73: 61-73.

PAGE 232

206 121. Baker JR (1969), An outbreak of sa lmonellosis involving ve terinary hospital patients, Vet Rec 85: 8-10. 122. Ewart SL, Schott HC, Robison RL, Dwy er RM, Eberhart SW, Walker RD (2001), Identification of sources of Salmonella organisms in a veterinary teaching hospital and evaluation of the effect s of disinfectants on detection of Salmonella organisms on surface materials, J Am Vet Med Assoc 218: 1145-1151. 123. Hartmann FA, Callan RJ, McGuirk SM, West SE (1996), Control of an outbreak of salmonellosis caused by drug-resistant Salmonella anatum in horses at a veterinary hospital and measures to prevent future infections, J Am Vet Med Assoc 209: 629-631. 124. Koterba A, Torchia J, Silverthorne C, Ramphal R, Merritt AM, Manucy J (1986), Nosocomial infections and bacterial antibiotic resistance in a university equine hospital, J Am Vet Med Assoc 189: 185-191. 125. Pare J, Carpenter TE, Thurmond MC (1996), Analysis of spatial and temporal clustering of horses with Salmonella kref eld in an intensiv e care unit of a veterinary hospital, J Am Vet Med Assoc 209: 626-628. 126. Schott HC, Ewart SL, Walker RD, Dwyer RM, Dietrich S, Eber hart SW, Kusey J, Stick JA, Derksen FJ (2001), An outbreak of salmonellosis among horses at a veterinary teaching hospital, J Am Vet Med Assoc 218: 1152-9, 1100. 127. Tillotson K, Savage CJ, Salman MD, Gentry-Weeks CR, Rice D, Fedorka-Cray PJ, Hendrickson DA, Jones RL, Nels on W, Traub-Dargatz JL (1997), Outbreak of Salmonella infantis infection in a large animal veterinary teaching hospital, J Am Vet Med Assoc 211: 1554-1557. 128. van Duijkeren E, Sloet van Oldrui tenborgh-Oosterbaan MM, Houwers DJ, van Leeuwen WJ, Kalsbeek HC (1994), Equine salmonellosis in a Dutch veterinary teaching hospital, Vet Rec 135: 248-250. 129. Lightfoot NF, Ahmad F, Cowden J ( 1990), Management of institutional outbreaks of Salmonella gastroenteritis, J Antimicrob Chemother 26: 37-46. 130. Rotger R, Casadesus J (1999), The virulence plasmids of Salmonella, Int Microbiol 2: 177-184. 131. Lesnick ML, Okamoto S, Fierer J, Guiney D (2000), The spvD gene of Salmonella dublin is important for mouse virule nce and is secreted in an invA dependent manner. ASM Abstract Database, American Society for Microbiology. 5-21-2000.

PAGE 233

207 132. Valone SE, Chikami GK, Miller VL (1993), Stress induction of the virulence proteins (SpvA, -B, and -C) from native plasmid pSDL2 of Salmonella dublin, Infect Immun 61: 705-713. 133. McClelland M, Sanderson KE, Spieth J, Clifton SW, Latreille P, Courtney L, Porwollik S, Ali J, Dante M, Du F, Hou S, Layman D, Leonard S, Nguyen C, Scott K, Holmes A, Grewal N, Mulvaney E, Ryan E, Sun H, Florea L, Miller W, Stoneking T, Nhan M, Waterston R, Wilson RK (2001), Complete genome sequence of Salm onella enterica serovar Typhimurium LT2, Nature 413: 852-856. 134. Popoff MY (2001), Antigenic Formulas of the Salmonella Serovars Paris, France: WHO Collaborating Centre for Reference and Research on Salmonella 135. Birnboim HC, Doly J (1979), A rapid al kaline extraction procedure for screening recombinant plasmid DNA, Nucleic Acids Res 7: 1513-1523. 136. House JK, Smith BP, Wildman TR, Carri gan MJ, Kamiya DY (1999), Isolation of Salmonella organisms from the mesenteric lymph nodes of horses at necropsy, J Am Vet Med Assoc 215: 507-510. 137. Olsen SJ, Bishop R, Brenner FW, Ro els TH, Bean N, Tauxe RV, Slutsker L (2001), The changing epidemiology of Salmonella : trends in serotypes isolated from humans in the United States, 1987-1997, J Infect Dis 183: 753-761. 138. Mainar-Jaime RC, House JK, Smith BP, Hird DW, House AM, Kamiya DY (1998), Influence of fecal shedding of Salmonella organisms on mortality in hospitalized horses, J Am Vet Med Assoc 213: 1162-1166. 139. Fierer J (2001), Extraintestinal Salmonella infecti ons: the significance of spv genes, Clin Infect Dis 32: 519-520. 140. Williamson CM, Baird GD, Manning EJ (1988), A common virulence region on plasmids from eleven serotypes of Salmonella, J Gen Microbiol 134 ( Pt 4): 975-982. 141. Boyd EF, Hartl DL (1998), Salmonella virulence plasmid. Modular acquisition of the spv virulence region by an Fplasmid in Salmonella enterica subspecies I and insertion into the ch romosome of subspecies II, IIIa, IV and VII isolates, Genetics 149: 1183-1190. 142. Aabo S, Brown DJ, Olsen JE (2000), Vi rulence characterizati on of a strain of Salmonella enterica subspecies houten (subspecies IV) with chromosomal integrated Salmonella plasmid virulence (spv) genes, Res Microbiol 151: 183-189.

PAGE 234

208 143. Cohen ND, Woods AM (1999), Character istics and risk fact ors for failure of horses with acute diarrhea to survive: 122 cases (1990-1996), J Am Vet Med Assoc 214: 382-390. 144. Zansky S, Wallace B, Schoonmaker-Bopp D, Smith P, Ramsey F, Painter J, Gupta A, Noviello S (2002), Ou tbreak of multidrug-resistant Salmonella Newport--United States, January-April 2002. 51, 545-548. Centers for Disease Control and Prevention. Morbid ity and Mortality Weekly Report. 145. Allen KJ, Poppe C (2002), Phenotypic and genotypic char acterization of food animal isolates of Salmonella with reduced sensitivity to ciprofloxacin, Microb Drug Resist 8: 375-383. 146. Olsen SJ, DeBess EE, McGivern TE, Marano N, Eby T, Mauvais S, Balan VK, Zirnstein G, Cieslak PR, Angulo FJ (2001), A nosocomial outbreak of fluoroquinolone-resistant salmonella infection, N Engl J Med 344: 15721579. 147. Bager F, Helmuth R (2001), Epidem iology of resistance to quinolones in Salmonella, Vet Res 32: 285-290. 148. Patterson JE (2000), Extended-spectrum beta-lactamases, Semin Respir Infect 15: 299-307. 149. Amyes SG (1989), The success of plasmi d-encoded resistance genes in clinical bacteria. An examination of pl asmid-mediated ampicillin and trimethoprim resistance genes and their resistance mechanisms, J Med Microbiol 28: 73-83. 150. Buu-Hoi A, Goldstein FW, Acar JF (1986), R-factors in gram-positive and gramnegative aerobic bacteria select ed by antimicrobial therapy, Scand J Infect Dis Suppl 49: 46-55. 151. Rodriguez-Martinez JM, Pascual A, Garcia I, Martinez -Martinez L (2003), Detection of the plasmid-mediated quinolone resistance determinant qnr among clinical isolates of Klebsi ella pneumoniae producing AmpC-type beta-lactamase, J Antimicrob Chemother 52: 703-706. 152. Skold O (2001), Resistance to trimethoprim and sulfonamides, Vet Res 32: 261273. 153. Prescott JF, Baggot JD, Walker RD (2000), Antimicrobial Therapy in Veterinary Medicine Ames, IA: Iowa State University Press. 154. Kern DL, Slyter LL, Leffel EC, Weav er JM, Oltjen RR (1974), Ponies vs. steers: microbial and chemical characteri stics of intestinal ingesta, J An Sci 38: 559-564.

PAGE 235

209 155. Glinsky MJ, Smith RM, Spires HR, Davis CL (1976), Measurement of volatile fatty acid production rates in the cecum of the pony, J An Sci 42: 14651470. 156. Reeves MJ, Salman MD, Smith G (1996), Risk factor for equine acute abdominal disease (colic): results from a multi-centered case-control study, Prev Vet Med 26: 285. 157. Goncalves S, Julliand V, Leblond A (2002) Risk factors associ ated with colic in horses, Vet Res 33: 641-652. 158. Argenzio RA, Lowe JE, Pickard DW Stevens CE (1974), Digesta passage and water exchange in the equine large intestine, Am J Physiol 226: 10351042. 159. Hintz HF, Argenzio RA, Schryver HF (1971), Digestion coefficients, blood glucose levels and molar percentage of volatile acids in intestinal fluid of ponies fed varying forage-grain ratios, J An Sci 33: 992. 160. Holtug K, Clausen MR, Hove H, Christ iansen J, Mortensen PB (1989), The colon in carbohydrate malabsorption: shortchain fatty acids, pH, and osmotic diarrhoea, Scand J Gastroenterol 27: 545-552. 161. Garner HE, Moore JN, Johnson JH, Clark L, Amend JF, Tritschler LG, Coffman JR (1978), Changes in the caecal flora associated with the onset of laminitis, Equine Vet J 10: 249-252. 162. Goodson J, Tyznik WJ, Cline JH, Dehorit y BA (1988), Effects of an abrupt diet change from hay to concentrate on microbial numbers and physical environment in the cecum of the pony, Appl Environ Microbiol 54: 19461950. 163. Maczulak AE, Dawson KA, Baker JP ( 1985), Nitrogen utiliza tion in bacterial isolates from the equine cecum, Appl Environ Microbiol 50: 1439-1443. 164. Kern DL, Slyter LL, Weaver JM, Leffel EC, Samuelson G (1973), Pony cecum vs. steer rumen: the effect of oats and hay on the microbial ecosystem, J An Sci 37: 463-469. 165. McCreery S, Fulghum RS, Baker JP ( 1971), Microflora of the equine cecum, J An Sci 33: 234. 166. Mackie RI, Wilkins CA (1988), Enumera tion of anaerobic bact erial microflora of the equine gastrointestinal tract, Appl Environ Microbiol 54: 2155-2160. 167. Julliand V, de Vaux A, Millet L, Fo nty G (1999), Identification of Ruminococcus flavefaciens as the predominant celluloly tic bacterial species of the equine cecum, Appl Environ Microbiol 65: 3738-3741.

PAGE 236

210 168. Medina B, Girard ID, Jacotot E, Jul liand V (2002), Effect of a preparation of Saccharomyces cerevisiae on microbial profiles a nd fermentation patterns in the large intestine of horses fe d a hig fiber or high starch diet, J An Sci 80: 2600-2609. 169. de Fombelle A, Julliand V, Drogoul C, Jacotot E (2001), Feeding and microbial disorders in horses: 1--Effects of an ab rupt incorporation of two levels of barley in a hay diet on micr obial profile and activities, J Equine Vet Sci 21: 439-445. 170. Horspool LJI (1992), Antimicrobial disposition and in teractions with gastrointestinal microflora in Equidae PhD Dissertation, University of Glasgow. 171. Horspool LJ, Taylor DJ, McKellar QA (1994), Plasma disposition of amikacin and interactions with gastrointestinal microflora in Equidae following intravenous and oral administration, J Vet Pharmacol Ther 17: 291-298. 172. Edlund C, Nord CE (2000), Effect on the human normal microflora of oral antibiotics for treatment of urinary tract infections, J Antimicrob Chemother 46 Suppl A: 41-48. 173. Brumbaugh GW (1987), Rational selec tion of antimicrobial drugs for the treatment of infections in horses, Vet Clin North Am Equine Pract 3: 191220. 174. Horspool LJ, McKellar QA (1991), Determ ination of short-chain fatty acids in equine caecal liquor by ion ex change high performance liquid chromatography after solid phase extraction, Biomed Chromatogr 5: 202206. 175. Russell JB, Diez-Gonzalez F (1998), The effects of fermentation acids on bacterial growth, Adv Microb Physiol 39: 205-234. 176. Cherrington CA, Hinton M, Mead GC, Chopra I (1991), Organic acids: chemistry, antibacterial activity and practical applications, Adv Microb Physiol 32: 87-108. 177. Baik HS, Bearson S, Dunbar S, Foster JW (1996), The acid tolerance response of Salmonella typhimurium provides protection against organic acids, Microbiology 142: 3195-3200. 178. Kwon YM, Ricke SC (1998), Inducti on of acid resistance of Salmonella typhimurium by exposure to short-chain fatty acids, Appl Environ Microbiol 64: 3458-3463.

PAGE 237

211 179. Lee IS, Lin J, Hall HK, Bearson B, Fo ster JW (1995), The stationary-phase sigma factor s (RpoS) is required for a sust ained acid tolerance response in virulent Salmonella typhimurium Mol Microbiol 17: 155-167. 180. Kwon YM, Park SY, Birkhold SG, Rick e SC (2000), Induction of resistance of Salmonella typhimurium to environm ental stresses by exposure to shortchain fatty acids, J Food Sci 65: 1037-1040. 181. Diez-Gonzalez F, Callaway TR, Kizou lis MG, Russell JB (1998), Grain feeding and the dissemination of acid-resistant Escherichia coli from cattle, Science 281: 1666-1668. 182. Rankin JD, Taylor RJ (1966), The estimation of the doses of Salmonella typhimurium for experimental production of disease in calves, Vet Rec 78: 706-707. 183. Nicpon J, Czerw P, Zawadzki W (2000), Research of pathomechanism and therapy of colitis X, Electronic Journal of Polish Agricultural Universities 3:2. http://www.ejpau.media.pl/series/ volume3/issue2/veterinary/hs~art01.html. 184. Van Immerseel F, De Buck J, Pasmans F, Velge P, Bottreau E, Fievez V, Haesebrouck F, Ducatelle R (2003), I nvasion of Salmonella enteritidis in avian intestinal epithelial cells in vi tro is influenced by short-chain fatty acids, Int J Food Microbiol 85: 237-248. 185. Durant JA, Lowry VK, Nisbet DJ, Stanker LH, Corrier DE, Ricke SC (1999), Short-chain fatty acids affect cell-as sociation and invasion of HEp-2 cells by Salmonella typhimurium, J Environ Sci Health B 34: 1083-1099. 186. Axe DD, Bailey JE (1995), Transport of lactate and acetate through the energized cytoplasmic membrane of Escherichia coli Biotech Bioeng 47: 8-19. 187. Caldwell AL, Gulig PA (1991), The Sa lmonella typhimurium virulence plasmid encodes a positive regulator of a plasmid-encoded virulence gene, J Bacteriol 173: 7176-7185. 188. Fang FC, Krause M, Roudier C, Fier er J, Guiney DG (1991), Growth regulation of a Salmonella plasmid gene essential for virulence, J Bacteriol 173: 6783-6789. 189. El Gedaily A, Paesold G, Chen CY, Guiney DG, Krause M (1997), Plasmid virulence gene expression induced by short-chain fatty acids in Salmonella dublin: identification of rpoS-dep endent and rpo-S-independent mechanisms, J Bacteriol 179: 1409-1412.

PAGE 238

212 190. Corrier DE, HintonJr A, Ziprin RL, Beier R, DeLoach JR (1990), Effect of dietary lactose on cecal pH, bacteriostatic volatile fatty acids, and Salmonella typhimurium colonization of broiler chicks, Avian Dis 34: 617-625. 191. Holt PS (2003), Molting and Salmonella en terica serovar enteri tidis infection: the problem and some solutions, Poult Sci 82: 1008-1010. 192. McHan F, Shotts EB (1993), Effect of short-chain fatty acids on the growth of Salmonella typhimurium in an in vitro system, Avian Dis 37: 396-398. 193. Annison G, Illman RJ, Topping DL (2003), Acetylated, propionylated or butyrylated starches raise large bowel short-chain Fatty acids preferentially when fed to rats, J Nutr 133: 3523-3528. 194. Prohaszka L, Jayarao BM, Fabian A, Kovacs S (1990), The role of intestinal volatile fatty acids in the Salmonella shedding of pigs, Zentralbl Veterinarmed B 37: 570-574. 195. Smith BP (2002), Diseases of the Alimentary Tract. In: Smith BP ed. Large Animal Internal Medicine: Diseases of Horses, Cattle, Sheep, and Goats. 3rd ed.; 775-779. 196. Chambers PG, Lysons RJ (1979), The i nhibitory effect of bovine rumen fluid on Salmonella typhimurium, Res Vet Sci 26: 273-276. 197. Mattila T, Frost AJ, O'Boyle D (1988) The growth of Salmonella in rumen fluid from cattle at slaughter, Epidemiol Infect 101: 337-345. 198. Reis RB, Combs DK (2000), Effects of increasing levels of grain supplementation on rumen environment and lactation pe rformance of dairy cows grazing grass-legume pasture, J Dairy Sci 83: 2888-2898. 199. Schoonmaker JP, Cecava VM, Faulkner DB, Fluharty FL, Zerby HN, Loerch SC (2003), Effect of source of energy an d rate of growth on performance, carcass characteristics, ruminal fe rmentation, and serum glucose and insulin of early-weaned steers, J Anim Sci 81: 843-855. 200. Plumb DC (2002), Veterinary Drug Handbook Ames, IA: Iowa State Press. 201. Adam KMG (1951), The quantity and dist ribution of the ciliate protozoa in the large intestine of the horse, Parasitology 43: 301. 202. Argenzio RA, Southworth M, Stevens CE (1974), Sites of organic acid production and absorption in the equine gastrointestinal tract, Am J Physiol 226: 1043-1050.

PAGE 239

213 203. Dunbar SA, Vander Zee CA, Oliver KG, Karem KL, Jacobson JW (2003), Quantitative, multiplexed detection of bacterial pathogens: DNA and protein applications of the Luminex LabMAP system, J Microbiol Methods 53: 245-252. 204. Hsiung T (1930), A monogra ph on the protozoa of the la rge intestine of the horse, Iowa St Coll J Sci 4: 359-424. 205. Cook WR (1973), Diarrhoea in the horse associated with stre ss and tetracycline therapy, The Veterinary Record 93: 15-16. 206. Ensink JM, van Klingeren B, Houwer s DJ, Klein WR, Vulto AG (1993), In-vitro susceptibility to antimicrobial drugs of bacterial isolates from horses in The Netherlands, Equine Vet J 25: 309-313. 207. Sjovall J, Huitfeldt B, Magni L, Nord CE (1986), Effect of beta-lactam prodrugs on human intestinal microflora, Scand J Infect Dis Suppl 49: 73-84. 208. Divers T, Kirker J (1998), Pharmaco logy and Adverse Drug Reactions. In: Orsini J and Divers T eds. Manual of Equine Emergencies: Treatment and Procedures. ; 565-598. 209. Genz AK, Busche R, von Engelhardt W (1997), Surface pH of the distal colonic epithelium of guinea pigs, Comp Biochem Physiol 118A: 407-408. 210. Moore BE, Dehority BA (1993), Effect s of diet and hindgut defaunation on diet digestibility and microbial concentra tions in the cecum and colon of the horse, J An Sci 71: 3350-3358. 211. Williams AG, Coleman GS (1988), Th e Rumen Protozoa. In: Hobson PN ed. The Rumen Microbial Ecosystem. ; 77-129. 212. Wright GD, Berghuis AM, Mobasher y S (1998), Aminoglycoside Antibiotics: Structures, Functions, and Resistance. In: Rosen BP and Mobashery S eds. Resolving the Antibiotic Par adox: Progress in Drug Design and Resistance. Adv.Exp.Med.Biol.; 27-69. 213. Clarke LL, Roberts MC, Argenzio RA (1990), Feeding and digestive problems in horses. Physiologic responses to a concentrated meal, Vet Clin North Am Equine Pract 6: 433-450. 214. Stevens CE, Hume ID (1998), Cont ributions of microbes in vertebrate gastrointestinal tract to producti on and conservation of nutrients, Physiol Rev 78: 393-427. 215. Hohmann EL (2001), Nontyphoidal salmonellosis, Clin Infect Dis 32: 263-269.

PAGE 240

214 216. Matsuki T, Watanabe K, Tanaka R (2003), Genusand species-specific PCR primers for the detection and id entification of bifidobacteria, Curr Issues Intest Microbiol 4: 61-69. 217. Huijsdens XW, Linskens RK, Mak M, Meuwissen SG, Vandenbroucke-Grauls CM, Savelkoul PH (2002), Quantifica tion of bacteria adherent to gastrointestinal mucosa by real-time PCR, J Clin Microbiol 40: 44234427.

PAGE 241

215 BIOGRAPHICAL SKETCH Tamara Shea Vetro Widenhouse was born April 15, 1970, in Fort Lauderdale, Florida, the first and only child of Ronald and Vicky Vetro. She attended the University of Florida in Gainesville, Florida, and receiv ed a Bachelor of Science degree in animal science from the College of Agriculture in 1992. She continued on toward fulfillment of a lifelong aspiration to become a veterinari an, by accepting a place in the College of Veterinary Medicine’s Cla ss of 1999, pausing briefly along the way to marry Christopher W. Widenhouse in 1995. She graduated with high honors from the University of Florida’s College of Veterinary Medicine in 1999. Tamara was also accepted into graduate school and commenced work on her PhD the same year as starting veterinary school, and worked simultaneously toward both degrees. After graduation from veterinary school, Tamara was awarded the inaugural Deedie Wrigley-Hancock Fellowship in Equine Colic Research in th e Department of Large Animal Clinical Sciences at the University of Florida, unde r the guidance of Dr. G uy D. Lester and Dr. Alfred M. Merritt—directors of the internationally renowne d Island Whirl Equine Colic Research Laboratory. She also became a mo ther to two wonderful children during her graduate career, Alexis Mack enna in 2000 and Carissa Mack enzie in 2002, and is now expecting twins in September 2004. She move d to Pembroke Pines, Florida, in 2001 to finish work on this document, be a fulltime mom, and manage Veterinary Medical Solutions Incorporated with her husband, a small biotech research and manufacturing

PAGE 242

216 company focusing on veterinary product deve lopment. Her future plans include a residency in large animal internal medicine.


Permanent Link: http://ufdc.ufl.edu/UFE0003062/00001

Material Information

Title: Equine Salmonellosis--Molecular Epidemiology of Clinical Isolates and the Effect of Antibiotics on the Cecal Microenvironment with Particular Reference to Short-Chain Fatty Acids and the Salmonella Plasmid Virulence (spv) Genes
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0003062:00001

Permanent Link: http://ufdc.ufl.edu/UFE0003062/00001

Material Information

Title: Equine Salmonellosis--Molecular Epidemiology of Clinical Isolates and the Effect of Antibiotics on the Cecal Microenvironment with Particular Reference to Short-Chain Fatty Acids and the Salmonella Plasmid Virulence (spv) Genes
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0003062:00001


This item has the following downloads:


Full Text












EQUINE SALMONELLOSIS-MOLECULAR EPIDEMIOLOGY OF CLINICAL
ISOLATES AND THE EFFECT OF ANTIBIOTICS ON THE CECAL
MICROENVIRONMENT WITH PARTICULAR REFERENCE TO SHORT-CHAIN
FATTY ACIDS AND THE SALMONELLA PLASMID VIRULENCE (spv) GENES

















By

TAMARA SHEA VETRO WIDENHOUSE


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2004

































Copyright 2004

by

Tamara Shea Vetro Widenhouse




























I would like to dedicate this work to my family, Mom, Dad,
Christopher, Alexis and Carissa-you define me. Without you, I
am nothing, and cannot imagine my life in your absence. You
have made me a better daughter, student, teacher, scientist, wife,
mother, friend, and human being. You are my universe, and this
work is just as much yours as it is mine.


I also dedicate this dissertation to every animal that has made the
ultimate sacrifice in the name of research. Though it was not by
choice, your gifts were never taken for granted, nor will they
ever be forgotten.















ACKNOWLEDGMENTS

I would like to thank my parents, for making sure that I grew up safe, loved, and

turned into the kind of person you would not cross the street to avoid. They will always

be my role models. I would also like to thank my husband Christopher, for standing

beside me all these years. Even though a long distance relationship, vet school, a Punky,

two doctoral degrees, billions in student loans, a Peach, in-laws, and writing two

dissertations might have been enough to vaporize any marriage, we have only gotten

stronger. May we have many more "adventures" together. I would like to thank my

mentor and colleague, Dr. Guy Lester, for the eternal open door (even though I had to

travel across several continents to walk through it!), and the incredible amount of respect,

direction, and free-rein given to me throughout my veterinary school and graduate tenure.

A good mentor is difficult to find, and a great one is only dreamt of-he is one of the

best. I would also like to thank the few friends who have managed to stick around long

after their statute of limitations ran out. I thank Dr. Lori Wendland and Dr. Chris

Sanchez, both of whom provided lodging, food, babysitting services, good coffee,

libations, a sympathetic ear, and a helping hand whenever it was needed-which was

often. I cannot thank them enough, and am forever grateful. I would especially like to

thank Misdee Wrigley-Milligan for her financial support, and the endowment of the

Deedie Wrigley-Hancock Fellowship for Equine Colic Research. Her tireless dedication

to the University of Florida, College of Veterinary Medicine, and to the study of equine

colic is to be commended. I would like to thank the members of my committee, Dr.









Alfred M. Merritt, Dr. Paul A. Gulig, Dr. Saundra TenBroeck, Dr. Steeve Giguere, and

Dr. Maureen T. Long. I know at times it seemed like I had dropped off the face of the

earth and that this process might drag on forever, but I thank them for their patience and

guidance; it is appreciated more than they will ever realize. I would also like to thank An

Nguyen, who went above and beyond the call of duty to provide the isolates and

sensitivity data. Most importantly, without Hilken V. Kuck, I would probably still be

stuck in the dungeon of a laboratory somewhere or on the Florida Turnpike. He has

saved me infinite amounts of time in the lab and on the road that was better spent with

my family, and have also put up with my "less-than-charming" stress-induced personality

for many years-I hope he can someday forgive me. I sincerely thank him.

Last, but certainly not least, I would like to thank those horses and ponies who were

intimately involved in my journey, especially Tony, Joni, Cody, Oreo, Hide, Seek, Easy,

Scott, Bill, Ted, Fly, Rapture, and Willie. Please know that I am forever grateful for their

sacrifice and am a better person for only having known them.
















TABLE OF CONTENTS
Page



A C K N O W L E D G M E N T S ................................................................................................. iv

LIST O F TA B LE S .......... ... .. ..... ....... .... ........... .......... .... .............

LIST OF FIGURES ............. .. ..... ...... ........ ........ ....................... xv

A B B R E V IA TIO N S ................................................. ................... .... .. ....xx

A B ST R A C T ..................................................................................................................... xxv

CHAPTER

1 IN TR OD U CTION ............................................... .. ......................... ..

The Genus Salm onella................................................. ..... .... .............. .. 1
T he B bacteria ...................................................................................... .......1
A nim al M odels of D disease ......................................................... ............. 2
The Salm onella V irulence Plasm id ............................................ .....................3
Salm onella Plasm id Virulence (spy) Genes ........................................ ...............4
Short-Chain or Volatile Fatty Acids and Salmonella .......................................6
Antibiotic-Associated Diarrhea (AAD) in the Horse ................................................7
The Gastrointestinal Microenvironment.....................................................9
The N orm al Flora ............................ ................. ....... .. ......... .. .9
Short-Chain Fatty Acids-Production and Intestinal Function......................... 10
Effects of Antimicrobial Therapy: Dysbacteriosis................... ...............14
S p ecific A im s..................................................... ................ 15
H y p oth eses............................. ........................................................... ............... 16

2 SALMONELLA IN HORSES-DISEASE DEFINITION AND GENERAL AND
MOLECULAR EPIDEMIOLOGY ... .. ................... .................18

Background ......................... ..............................18
D disease O v erv iew ........... ...... .................. .................................. .. .... .. .. .. ... 18
Prevalence .................... .............................. ........ ........ 19
Reported Risk Factors for Salmonella Infection ...........................................19
Salmonella Serovars Associated with Equine Infection......................... 20
Role of Microbial Virulence Factors in Equine Salmonella Infection ..............21









Disease Prevention-Diet, Probiotics, Immunity ............................................21
D disease Treatm ent ............................. .. .................. .................. ......... 24
The Salm onella Virulence Plasm id .......................................... ............... 24
Salm onella Plasmid Virulence (spv) Genes ................................ ............... 25
Function of the spy genes .................................... .......................... ......... 25
Significance of the spy genes ........... ................................. ...............26
S p e c ific A im s ........................................................................................................ 2 6
M materials and M methods ....................................................................... ..................27
C a se S e le ctio n ............................................................................................... 2 7
M icrobiological Techniques................................................... ............... ... 27
F ie ld sa m p le s .......................................................................................... 2 7
Clinical and reference isolates....................... ... ...............28
Salmonella Identification and Antibiotic Resistance Profile.............................30
Salm onella Isolate Storage ............................................................................ 30
R reference Strains .................................................................. ............. 31
Plasmid Profiling of Salmonella Isolates ..................................... ............... 31
Polymerase Chain Reaction (PCR) Identification of spy Genes ....................34
Salmonella Plasmid Transformations into Susceptible Bacteria-Effects on
A ntibiotic R resistance ..................................... ................... ..............35
Statistical M ethods ...................................... .............................37
Results ................. ........................ ..... ... .........................37
A sym ptom atic Population ............................................................................37
C lin ic al C a se s .................................. ........ ....................................... ............... 3 7
Relationship Between Gender or Age and Outcome................ ..................38
C a se S easo n ality ............ ................... .................................... .... .... ...... 3 9
Group and Serovar D distribution ........................................ ....... ............... 40
Outcom e by Group or Serovar ........................................ ........................ 42
Plasm id Profiling ......... .. ................................. ..... ......................... 44
sp G ene A analysis ................................................................................. 47
Outcome by Presence of the Virulence Plasmid or spy Genes............................54
Effect of Clinical and Laboratory Parameters on Outcome .............................56
Relationship Between Proportion of Positive Fecal Salmonella Cultures and
Outcome ................ ......... .... ....... ................... ... ....... .....58
A ntibiotic R resistance Profiles ........................................ ......... ............... 58
Antibiotic Resistance Transformation.......................................................59
Site of Salm onella Isolation........................................... .......................... 62
M ulti-Serovar Salm onella Infections ...................................... ............... 65
D isc u ssio n ............................................... .. ................... ................ 6 7

3 E X PE R IM E N T S ..................... .. ........................................................... 75

Background ............... .................... ....... ............... 75
The Horse: Classic Large Intestine Fermenter ............................... .................. 75
Equine Cecal Anaerobic Flora and SCFAs in the Normal Animal ...................76
Antimicrobial Effects on Normal Anaerobic Flora................ ......... .......77
Antimicrobial Effects on SCFAs.................................................................... 81
Current Theory on the Pathogenesis of Antibiotic-Associated Diarrhea (AAD)82









Effects of SCFA on Anaerobic Growth of Bacteria................................ ....84
Acid Tolerance Response of Salmonella and Other Enterobacteriaceae ............84
SCFA Effects on Salmonella Growth and Invasion................ .............. ....86
SCFA Effects on Expression of spy Genes in vitro...........................................88
SCFAs and Salmonella Colonization and Infection of Avian Species................88
SCFA s and Salm onella in Sw ine......................................... ........... .................89
SCFAs and Salmonella Colonization and Infection of Bovine Species..............90
Specific A im s........................................................................................ 90
M materials and M methods ....................................................................... ..................9 1
IA CU C A pproval.............. .... ........................................ .... .. ........ ......91
Subject Coding for Experiments and Data Analysis ..................................91
Surgical Placement of Cecal Cannula in the Horse ...................................91
Antibiotic Treatment of Horses ............... .......... ............................... 94
Equine Cecal Sampling Procedure ............................. .................................. 95
Physical Effects on the Horse....................... .......... .................... 96
Effects on Fecal C onsistency......................................... .......................... 97
Effects on Cecal Content Character ....................................... ............... 97
Equine Cecal Anaerobe Quantification............... ................... .................97
pH Analysis of Equine Cecal Contents .......................................... .................. 98
Short-Chain Fatty Acid Analysis of Equine Cecal Contents...............................98
Protozoal Quantification of Cecal Contents from Horses Treated with
A n tib io tic s ............................. ..... ... ................................... ............... 9 9
In vitro Short-Chain Fatty Acid Growth Comparison..................................99
In vitro Effects of Cecal Liquor from Antibiotic-treated Horses on Anaerobic
G row th of Salm onella .......................................................................100
Statistical M ethods .............................................. ............ .. .............. 102
R e su lts .................. ........... ................................................................. ............. 1 0 2
E effects on the H orse ...................... ................ .................. ........ 102
Effects on Cecal pH.................... .. ........................................ 103
Effects on Cecal Protozoal Counts .............................................................. 104
Effects on Cecal SCFA Quantities and Proportions.................................106
Effects on Cecal Anaerobic Bacteria ................................ ............... 111
In vitro Effects of SCFAs on Anaerobic Growth of Salmonella ..................13
E effect of A cetate ............................ ...... .. ................................. ................ ... 116
Effect of the Plasmid and spy Genes on Acetate Response.............................1116
Effect of Butyrate ........................................................... ......... .................. 117
Effect of the Plasmid and spy Genes on Butyrate Response.............................17
Effect of Propionate ............ .... ......... ... ......... .......... .............. 118
Effect of the Virulence Plasmid and spy Genes on Propionate Response......... 118
In vitro Effects of Cecal Liquor from Antibiotic-treated Horses on Anaerobic
G row th of Salm onella ................................................... ........ .................... ... 119
Effect of the Virulence Plasmid and spy Genes on Anaerobic Growth of
Salmonella Exposed to Cecal Liquor from Antibiotic-Treated Horses.........122
D iscu ssion ................................................................................................ ..... 122









4 SUMMARY, CONCLUSIONS, AND FUTURE DIRECTIONS............................133

APPENDIX

A SALMONELLA EPIDEMIOLOGY DATA COLLECTION SHEET ....................139

B INDEX OF SUPPLIERS AND CONTACT INFORMATION.............................141

C SALM ONELLA ISOLATE INDEX .......................................................................143

D SALMONELLA DATABASE CASE DESCRIPTIVE INFORMATION.............146

E SALMONELLA ISOLATE ANTIMICROBIAL SUSCEPTIBILITY DATA........ 174

F D E SC R IPTIV E STA TISTIC S....................................................... .....................187

LIST OF REFERENCES ....... ........................................................ ............... 196

B IO G R A PH IC A L SK E T C H ........................................ ............................................215
















LIST OF TABLES


Table page

1-1. Short-chain fatty acid chemical formulas and common names..............................10

2-1. sp gene characteristics ................................................. ................................ 26

2-2. Composition of bacterial culture m edia................................... ....... ............... 29

2-3. Salmonella serovar Typhimurium reference strains used in this study ...................31

2-4. Composition of buffers and solutions used in plasmid extraction protocols............33

2-5. Times and temperatures for PCR reactions ........................... ..... ...........34

2-6. Primers utilized in PCR reactions................... ............ ............ .... 35

2-7. Breed distribution of 84 equine salmonella cases 1999-2002.................................38

2-8. Effect of gender on mortality in 96 cases of equine salmonellosis ...........................38

2-9. Effect of age on mortality in 85 cases of equine salmonellosis..............................39

2-10. Average minimum temperatures in Gainesville, Florida, USA (1961-1990)..........40

2-11. Salmonella serovars isolated from 98 equine cases 1999-2002 ...........................41

2-12. Salmonella isolates of environmental and species other than equids collected
1 9 9 9 -2 0 0 2 ..................................................... ................ 4 1

2-13. Effect of salmonella group on mortality in 88 cases of equine salmonellosis ........44

2-14. Plasmid-positive salmonella isolates by serovar 1999-2002...............................44

2-15. Summary outcome as determined by presence of the virulence plasmid and spy
genes in 98 equine salm onella cases ............................................. ............... 54

2-16. Effect of spy gene presence on mortality in 86 cases of equine salmonellosis where
outcome e was known .................. ............................ ........ ................. 55

2-17. Logistic regression model with variables predictive of outcome............................57









2-18. Antibiotic susceptibilities for 101 equine salmonella isolates. The reported %
susceptible, % intermediate, and % resistant, are only for those isolates with data
for that antibiotic. ................................................... ................. 58

2-19. Antibiotic susceptibility report for Case 78, with intermediate resistance to
en ro flox acin ...................................... ............................. ................ 5 9

2-20. Clinical salmonella isolates from 105 equine cases by location of cultured
sp e c im e n ............................ ........... ...... ....................... ................ 6 2

2-21. Systemic sites of salmonella infection in horses by serovar .................................63

2-22. Relationship of the virulence plasmid and spy genes to isolate location in 98 cases
of equine salm onellosis ............................ ................ ......................... 64

2-23. Details of multiserovar salmonella infections in six horses 1999-2002..................66

3-1. Summary of literature reports quantifying equine cecal anaerobic bacteria ............76

3-2. Literature reports quantifying normal equine cecal SCFA concentrations ...............77

3-3. Literature summary of antibiotic effects on fecal bacteria and short-chain
fatty acids .............. ...... ............. ......... .. ......................... 79

3-4. Summary of single dose antibiotic effects on the equine cecal microenvironment ..82

3-5. Coding legend for experimental animals........................................ ............... 91

3-6. Antibiotic treatments of horses..................... ......... .. ............ ........... .... 94

3-7. Cecal liquor pH of cannulated horses before and after 4 days of control (no)
antibiotic treatm ent .................. ............................... .. ...... .. ........ .... 104

3-8. Cecal liquor pH of cannulated horses treated with intramuscular ceftiofur sodium at
2 mg/kg twice daily, before and after 4 days of treatment...................................104

3-9. Cecal liquor pH of cannulated horses treated with intravenous oxytetracycline at 10
mg/kg once daily, before and after 4 days of treatment .............. ................104

3-10. Cecal liquor pH of cannulated horses treated with oral trimethoprim-
sulfamethoxazole at 30 mg/kg twice daily, before and after 4 days of treatment.. 104

3-11. Total protozoal counts per ml of cecal contents from cannulated horses before and
after 4 days of control (no) antibiotic treatment ....................................................105

3-12. Total protozoal counts per ml of cecal contents from cannulated horses treated with
intramuscular ceftiofur sodium at 2 mg/kg twice daily, before and after 4 days of
treatm ent ............................................................... .... ..... ......... 105









3-13. Total protozoal counts per ml of cecal contents from cannulated horses treated with
intravenous oxytetracycline at 10 mg/kg once daily, before and after 4 days of
treatm ent ............... ........... ......................... ...........................105

3-14. Total protozoal counts per ml of cecal contents from cannulated horses treated with
oral trimethoprim-sulfamethoxazole at 30 mg/kg twice daily, before and after 4
days of treatm ent ................. ......... .... .......... .......... ........ 105

3-15. Cecal liquor concentrations of individual and total SCFAs from cannulated horses,
before and after 4 days of control (no) treatment............................107

3-16. Cecal liquor concentrations of individual and total SCFAs from cannulated horses
treated with intramuscular ceftiofur sodium at 2 mg/kg twice daily, before and after
4 days of treatment ................ ............. .................. ......... 108

3-17. Cecal liquor concentrations of individual and total SCFAs from cannulated horses
treated with intravenous oxytetracycline at 10 mg/kg once daily, before and after 4
days of treatm ent ................................. ........ ..... .... .. ........ .... 109

3-18. Cecal liquor concentrations of individual and total SCFAs from cannulated horses
treated with oral trimethoprim-sulfamethoxazole at 30 mg/kg twice daily, before
and after 4 days of treatm ent ....................................................................... .... 110

3-19. Mean counts of culturable anaerobic bacteria from serial dilutions of raw equine
cecal liquor, from 5 cannulated horses, before and after 4 days of control (no)
treatment. The dilution shaded in green was chosen for comparison. .................11

3-20. Mean counts of culturable anaerobic bacteria from serial dilutions of raw equine
cecal liquor, from 5 cannulated horses treated with intramuscular ceftiofur sodium
at 2 mg/kg twice daily, before and after 4 days of treatment. The dilution shaded in
green was chosen for comparison. ....................................................................... 112

3-21. Mean counts of culturable anaerobic bacteria from serial dilutions of raw equine
cecal liquor, from 5 cannulated horses treated with intravenous oxytetracycline at
10 mg/kg once daily, before and after 4 days of treatment. The dilution shaded in
green was chosen for comparison. ....................................................................... 112

3-22. Mean counts of culturable anaerobic bacteria from serial dilutions of raw equine
cecal liquor, from cannulated horses treated with oral trimethoprim-
sulfamethoxazole at 30 mg/kg twice daily, before and after 4 days of treatment.
The dilution shaded in green was chosen for comparison. ................................. 112

C-1. Index of salmonella isolates by group and serovar, with plasmid and spy
g e n e status s ..............................................................................................................14 3

D-1. Salmonella case descriptive information: breed, age, sex, presenting complaint, risk
factors for salmonellosis, specimen origin, and salmonella groups) and serovar(s).
Blank cells indicate missing or unavailable records. ........................................... 147









D-2. Salmonella case descriptive information: serovar, date sample taken, presence of
diarrhea, total hospitalization cost, case outcome, hospitalization days, number of
positive cultures, hematologic indices at time of positive culture, and total protein
changes during hospitalization. Blank cells indicate missing or unavailable
records. ............................................................................. 159

D-3. Salmonella case descriptive information: serovar, antibiotic therapy prior to
admission and types, antibiotic therapy during hospitalization and types. Drugs in
boldface type were used specifically to treat the salmonella infection. Blank cells
indicate missing or unavailable records, and three dashes indicates that nothing
could be determined from the record. ....................................... ............... 165

E-1. Salmonella isolate MIC antibiotic sensitivity profiles. Blank cells indicate missing
data. Legend: AMI = amikacin, AMOX = amoxicillin-clavulanic acid, AMP =
ampicillin, CEFA = cefazolin, CEFZ = ceftazidime, NAX = ceftiofur, CHLP =
chloram phenicol. ............................................................. .. ............ 175

E-2. Salmonella isolate MIC antibiotic sensitivity profiles. Blank cells indicate missing
data. Legend: CLIN = clindamycin, DOX = doxycycline, ENRO = enrofloxacin,
ERYT = erythromycin, GENT = gentamicin, IMIP = imipenem ........................179

E-3. Salmonella isolate MIC antibiotic sensitivity profiles. Blank cells indicate missing
data. Legend: NITR = nitrofurantoin, OX = oxacillin, PEN = penicillin, RIF =
rifampin, TET = tetracycline, TMP = trimethoprim-sulfamethoxazole...............183

F-1. Descriptive statistics for individual cecal SCFA measurements before and after four
days of control (no) antibiotic treatment in 5 horses.............................................187

F-2. Descriptive statistics for individual cecal SCFA measurements in 5 horses treated
with intramuscular ceftiofur sodium at 2 mg/kg twice daily, before and after 4 days
o f treatm en t ...................................................... ................ 18 7

F-3. Descriptive statistics for individual cecal SCFA measurements in 5 horses treated
with intravenous oxytetracycline at 10 mg/kg once daily, before and after 4 days of
treatm ent ........... ............ ............................. ........ .. . .. ............. 188

F-4. Descriptive statistics for individual cecal SCFA measurements in 5 horses treated
with oral trimethoprim-sulfamethoxazole at 30 mg/kg twice daily, before and after
4 day s of treatm ent ................................................................ ........ ...... 188

F-5. Descriptive statistics for cecal protozoal counts in 5 horses treated with control (no
treatment), ceftiofur sodium, oxytetracycline, or trimethoprim-sulfamethoxazole,
before and after 4 days of treatment........ .......................................... 189

F-6. Descriptive statistics for cecal pH in 5 horses treated with control (no treatment),
ceftiofur sodium, oxytetracycline, or trimethoprim-sulfamethoxazole, before and
after 4 days of treatment ............. ... ................................. 189









F-7. Descriptive statistics for salmonella growth in M9 supplemented with 10% sterile
filtered cecal contents from 5 individual horses treated with oxytetracycline (TET),
or trim ethoprim -sulfam ethoxazole (TM PS)..........................................................189

F-8. Descriptive statistics for salmonella growth in M9 supplemented with 10% sterile
filtered cecal contents from 5 individual horses treated with control (no treatment)
or ceftiofur sodium (N AX) ........................................................ ............. 190

F-9. Descriptive statistics for all salmonella growth in LB broth supplemented with 10%
sterile filtered cecal contents pooled from 5 horses treated with no treatment,
ceftiofur (NAX), oxytetracycline (TET), or trimethoprim-sulfamethoxazole
(TMPS). Units=CFU/ml, N=number of dilutions counted. ................................. 191

F-10. Descriptive statistics for all salmonella growth in M9 minimal medium
supplemented with 10% sterile filtered cecal contents pooled from 5 horses treated
with control (no treatment) or ceftiofur sodium (NAX), oxytetracycline (TET), or
trimethoprim-sulfamethoxazole (TMPS). Units for mean measurement are
CFU/ml, and N=number of dilutions counted.................... .................. ................ 192

F- 1. Descriptive statistics for salmonella growth in M9 minimal medium supplemented
with sodium chloride, acetate, butyrate, or propionate at 30 or 100mM .............193















LIST OF FIGURES


Figure page

1-1. spy gene expression regulation is dependent on growth phase and cellular location
in Salm onella ................................................... ........... ..... ......... ..... 5

1-2. Polysaccharide metabolism and SCFA production pathways in the rumen .............11

1-3. Diagram of major gastrointestinal microbial digestive and energy functions,
nitrogen and carbon recycling, and SCFA production...........................................12

1-4. Summary of potential enterotrophic effects of SCFA.............................................14

2-1. API20E rapid identification strip showing typical reaction results for Salmonella
sp e cie s .............................................................................. 2 8

2-2. Salmonella group C2 isolate as provided on Hektoen-Enteric agar......................28

2-3. Age distribution of 98 equine salmonella cases 1999-2002............................. 39

2-4. Seasonal distribution of salmonella cases from horses 1999-2002.........................40

2-5. Mortality distribution, within serovar, of non-surviving equine salmonella cases
1999-2002................................... ................................. ........... 43

2-6. Mortality by salmonella group in 88 cases with known outcomes..........................43

2-7. Plasmid profiles of 9 clinical salmonella isolates. Refer to Appendix C for specific
isolate information. Lanes: 1) Previously extracted 100-kb plasmid of 3306, 2)
Case 8, 3) Bovine isolate of S. Typhimurium var. Copenhagen, 4) Case 11, 5)
Bovine isolate of S. Typhimurium var. Copenhagen, 6) Case 12, 7) Bovine isolate
of S. Typhimurium var. Copenhagen, 8) Case 6, 9) Case 10, 10) Case 7..............47

2-8. Plasmid profiles of 4 clinical salmonella isolates. Refer to Appendix C for specific
isolate information. Lanes: 1) supercoiled marker DNA, 2) 100-kb plasmid of
x3306, 3) Case 78, 4) Case 71, 5) Case 66, 6) Case 77, 7) 100-kb plasmid of X3306,
8 ) b lan k ...................................... ............................. ..... ........ ...... 4 8









2-9. Plasmid profiles of 5 clinical salmonella isolates. Refer to Appendix C for specific
isolate information. Lanes: 1) supercoiled marker DNA, 2) 100-kb plasmid of
x3306, 3) Case 89, 4) Case 92, 5) Case 85, 6) Case 93, 7) Case 96, 8) 100-kb
plasm id of X3306. .................................................. .................. 49

2-10. Plasmid profiles of 4 clinical salmonella isolates. Refer to Appendix C for specific
isolate information. Lanes: 1) blank, 2) 100-kb plasmid of 3306, 3) Aged
monthnt) plasmid extract of X3306, 4) Case 46, 5) Case 44, 6) Case 43, 7) Case
53, 8) blank ................................................................................ ...... ... 49

2-11. Plasmid profiles of 4 clinical salmonella isolates. Refer to Appendix C for specific
isolate information. Lanes: 1) supercoiled marker DNA, 2) 100-kb plasmid of
x3306, 3) Aged monthnt) plasmid extract of 3306, 4) Case 41, 5) Case 63, 6)
Case 64, 7) Case 65, 8) supercoiled marker DNA. ............................................. 50

2-12. Plasmid profiles of 3 clinical salmonella isolates. Refer to Appendix C for specific
isolate information. Lanes: 1) supercoiled marker DNA, 2) 100-kb plasmid of
x3306, 3) 100-kb plasmid of X3306, 4) Case 82, 5) Case 83, 6) Case 40, 7) 100-kb
plasmid of 3306, 8) blank ................................... ......... .................... 50

2-13. Plasmid profiles of 7 clinical salmonella isolates. Refer to Appendix C for specific
isolate information. Lanes: 1) 100-kb plasmid of 3306, 2) Case 32, 3) Case 36, 4)
Case 37, 5) Case 91, 6) Case 90, 7) Case 87, 8) Case 86............... .....................51

2-14. PCR product results for spvA and spvC genes in 9 clinical salmonella isolates, with
positive and negative controls. Refer to Appendix C for specific isolate
information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) x3337 spy negative
control, 3) x3306 spy positive control, 4) Case 86, 5) Case 87, 6) Case 88, 7) Case
117, 8) Case 100, 9) Case 101, 10) Case 102, 11) Case 103, 12) Case 104, 13) 1-kb
ladder DNA marker (Promega), 14) blank ........................................ ................51

2-15. PCR product for asd gene in 9 clinical salmonella isolates (same isolates and
orientation as Figure 2-14). Refer to Appendix C for specific isolate information.
Lanes: 1) 1-kb ladder DNA marker (Promega), 2) x3337 spy negative control, 3)
x3306 spy positive control, 4) Case 86, 5) Case 87, 6) Case 88, 7) Case 117, 8)
Case 100, 9) Case 101, 10) Case 102, 11) Case 103, 12) Case 104, 13) 1-kb ladder
DNA marker (Promega), 14) blank ....................................................................... 52

2-16. PCR product results for spvA genes in 11 clinical salmonella isolates, with positive
and negative controls. Refer to Appendix C for specific isolate information. Lanes:
1) 1-kb ladder DNA marker (Promega), 2) x3306 spy positive control, 3) x3337 spy
negative control, 4) lost isolate, 5) Case 8, 6) Case 7, 7) Case 12, 8) Case 13, 9)
Case 10, 10) Case 9, 11) Case 5, 12) Case 3, 13) Case 11, 14) Case 4....................53









2-17. PCR product results for spvC genes in 11 clinical salmonella isolates, with positive
and negative controls. Refer to Appendix C for specific isolate information. Lanes:
1) 1-kb ladder DNA marker (Promega), 2) x3306 spy positive control, 3) x3337 spy
negative control, 4) Case 21, 5) Case 19, 6) Case 16, 7) Case 22, 8) Case 27, 9)
Case 26, 10) Case 25, 11) Case 24, 12) Case 23, 13) Case 15, 14) Case 14............53

2-18. PCR product results for the asd gene in 11 clinical salmonella isolates (same
isolates and orientation as Figure 2-17). Refer to Appendix C for specific isolate
information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) x3306 spy positive
control, 3) x3337 spy negative control, 4) Case 21, 5) Case 19, 6) Case 16, 7) Case
22, 8) Case 27, 9) Case 26, 10) Case 25, 11) Case 24, 12) Case 23, 13) Case 15, 14)
C ase 14 ..............................................................................54

2-19. Outcome in equine salmonella cases, as influenced by presence of the spy gene
locu s. ...............................................................................55

2-20. Outcome in equine salmonella cases, as influenced by absence of the spy gene
locu s. ...............................................................................56

2-21. Plasmid profiles of 3 clinical salmonella isolates and E. coli transformed with
plasmid DNA from those isolates. Refer to Appendix C for specific isolate
information and Appendix E for antimicrobial susceptibilities. Lanes: 1) 100-kb
plasmid of X3306, 2) Untransformed E. coli DH5a, 3) E. coli DH5a transformed
with Case 97, grown in CEF, 4) E. coli DH5a transformed with Case 92, grown in
CEF, 5) E. coli DH5a transformed with Case 98, grown in CEF, 6) Transforming
plasmid DNA from Case 97, 7) Transforming plasmid DNA from Case 92, 8)
Transforming plasmid DNA from Case 98. ........................................ .................60

2-22. Plasmid profiles of 2 clinical salmonella isolates and E. coli transformed with
plasmid DNA from those isolates. Refer to Appendix C for specific isolate
information and Appendix E for antimicrobial susceptibilities. Lanes: 1)
Untransformed E. coli DH5a, 2) 100-kb plasmid of X3306, 3)E. coli DH5a
transformed with Case 98, grown in AMP, 4) E. coli DH5a transformed with Case
98, grown in NAX, 5) E. coli DH5a transformed with Case 98, grown in CEF, 6) E.
coli DH5a transformed with Case 92, grown in AMP, 7) E. coli DH5a transformed
with Case 92, grown in CEF, 8) blank. ........................................ ............... 60

2-23. Plasmid profiles of 2 clinical salmonella isolates and E. coli transformed with
plasmid DNA from those isolates. Refer to Appendix C for specific isolate
information and Appendix E for antimicrobial susceptibilities. Lanes: 1)
Untransformed E. coli DH5a, 2) 100-kb plasmid of X3306, 3)E. coli DH5a
transformed with Case 97, grown in AMP, 4) E. coli DH5a transformed with Case
92, grown in NAX, 5) E. coli DH5a transformed with Case 97, grown in NAX, 6)
E. coli DH5a transformed with Case 97, grown in CEF, 7) blank, 8) blank...........61









2-24. Plasmid profile of Case 97-lane 5. The red box delineates 3 large plasmid bands
that are visible in the upper part of the lane. This isolate transferred ceftiofur,
cefazolin, and ampicillin resistance via two different plasmids (the lower two).....62

2-25. Systemic equine salmonella isolates compared to gastrointestinal isolates
b y g ro u p ...........................................................................6 4

3-1. Pathogenesis of antibiotic-associated diarrhea ...................................................... 83

3-2. Components of indwelling cecal cannula placed into experimental horses.
Clockwise from the top: side view of sliding flange placed on the lateral serosal
aspect of cecal wall, cannula with fixed interior flange, silicone filled, thick walled
tubing used to plug cannula, front view of sliding flange, hose clamp to secure plug
w within cannula. ..................................................... ................. 92

3-3. Experimental horse E (2) with cecal cannula 3 years post-implantation. ................93

3-4. Close-up view of cannula in situ in experimental horse E (2). Note the formation of
a firm swelling intimately associated with the cannula insertion. This is internal
granulation tissue forming around the interior silicone flanges which will result in
the eventual expulsion of the device. ............................................ ............... 93

3-5. Collection of equine cecal contents from indwelling silicone cannula. Note the
rapid flow and liquid nature of the contents........................ ... .............. 96

3-6. Mean cecal anaerobic culture counts expressed as CFU / ml of liquor from five
horses before and after treatment with control (no treatment), ceftiofur,
oxytetracycline, or trimethoprim-sulfamethoxazole. ................. ......... .......... 113

3-7. The effect of LB broth with sodium chloride (control treatment) added at 30mM or
100mM on anaerobic growth of S. Typhimurium x3306 vs. x3337. All solutions
were pH 6.5. Error bars represent 95%CI for two replicates of the experiment... 114

3-8. The effect of LB broth with sodium acetate added at 30mM or 100mM on anaerobic
growth of S. Typhimurium x3306 vs. x3337. All solutions were pH 6.5. Error bars
represent 95%CI for two replicates of the experiment.....................................114

3-9. The effect of LB broth with sodium butyrate added at 30mM or 100mM on
anaerobic growth of S. Typhimurium x3306 vs. x3337. All solutions were pH 6.5.
Error bars represent 95%CI for two replicates of the experiment.........................115

3-10. The effect of LB broth with sodium propionate added at 30mM or 100mM on
anaerobic growth of S. Typhimurium x3306 vs. x3337. All solutions were pH 6.5.
Error bars represent 95%CI for two replicates of the experiment.........................115

3-11. The effect of LB broth with sodium acetate at 30mM or 100mM compared to NaCl
at 30mM or 100mM on anaerobic growth of S. Typhimurium. All solutions were
pH 6.5. Error bars represent 95%CI for four replicates of the experiment .........16


xviii









3-12. The effect of LB broth with sodium butyrate at 30mM or 100mM compared to
NaC1 at 30mM or 100mM on anaerobic growth of S. Typhimurium. All solutions
were pH 6.5. Error bars represent 95%CI for four replicates of the experiment. .117

3-13. The effect of LB broth with sodium propionate at 30mM or 100mM compared to
NaCl at 30mM or 100mM on anaerobic growth of S. Typhimurium. All solutions
were pH 6.5. Error bars represent 95%CI for four replicates of the experiment. .118

3-14. The effect of LB broth with 10% added filter-sterilized cecal contents pooled from
five horses by treatment on the anaerobic growth ofS. Typhimurium. The horses
were treated with control (no treatment), ceftiofur (NAX), oxytetracycline (TET),
or trim ethoprim -sulfam ethoxazole (TM PS).......................................................... 119

3-15. The effect of M9 minimal medium (+ glucose) with 10% added filter-sterilized
cecal contents pooled from five horses by treatment on the anaerobic growth of S.
Typhimurium. The horses were treated with control (no treatment), ceftiofur
(NAX), oxytetracycline (TET), or trimethoprim-sulfamethoxazole (TMPS)........120

3-16. The effect of M9 minimal medium (+ glucose) with 10% added filter-sterilized
cecal contents from antibiotic-treated horses on the anaerobic growth of S.
Typhimurium. Data points are the mean of 5 individual horses treated with control
(no treatment) or ceftiofur (NAX). Time 6h is a missing data point. .................121

3-17. The effect of M9 minimal medium (+ glucose) with 10% added filter-sterilized
cecal contents from antibiotic-treated horses on the anaerobic growth of S.
Typhimurium. Data points are the mean of 5 individual horses treated with
oxytetracycline (TET) or trimethoprim-sulfamethoxazole (TMPS). .....................122















ABBREVIATIONS


% percentage)
+ positive
negative
ApH pH gradient
almost equal to
oC degrees Centigrade (Celsius)
OF degrees Fahrenheit
95%CI 95% confidence interval
A acetate
AAD antibiotic-associated diarrhea
ADH test for arginine dihydrolase, red/orange = +
ADP adenosine diphosphate
AMI amikacin
AMOX amoxicillin-clavulanic acid
AMP ampicillin
AMY amygdalin fermentation/oxidation test, yellow = +
ARA arabinose fermentation/oxidation test, yellow = +
asd aspartate semialdehyde dehydrogenase
ASP acid shock protein
ATP adenosine triphosphate
ATR acid tolerance response
B butyrate
BHI brain heart infusion
bp base pairs
C. Clostridium
CaCl2 calcium chloride
CEC competent Escherichia coli
CEF or CEFA cefazolin
CEFZ ceftazidime
cfu colony forming units
CH4 methane
CHLP chloramphenicol
CIT test for citrate utilization, blue-green/blue = +
CLIN clindamycin
cm centimeter
CO2 carbon dioxide
COD cause of death









CON control
df degrees of freedom
DMC direct microscopic count
DNA deoxyribonucleic acid
dNTP DNA nucleotides (A,C,G,T)
DOA dead on arrival
DOX doxycycline
E. Escherichia
e.g. for example
EB ethyl butyrate
EDTA ethylene diamine tetra acetic acid
ENRO enrofloxacin
ERYT erythromycin
et al. and others
euth. euthanatized
ex vivo outside the living body
FOS fructo-oligosaccharides
FUO fever of unknown origin
g grams
g gravity, 1011 N.m/s2
GDUD gastro duodenal ulcer disease
GEL gelatinase production test, diffusion of black = +
GENT gentamicin
GLU glucose fermentation/oxidation test, yellow = +
gyr gyrase
h hours)
H2 hydrogen
H202 hydrogen peroxide
H2S test for hydrogen sulfide production, black = +
HCI hydrochloric acid
HE Hektoen-Enteric agar
i.e. that is
IACUC Institutional Animal Care and Use Committee
IB isobutyrate
ICH iodochlorhydroxyguin
IM intramuscular
IMIP imipenem
in vivo inside the living body
IND test for indole production, red = +
INO inositol fermentation/oxidation test, yellow = +
IV intravenous
IVA isovalerate
kb kilobase(s)
kg kilogram
kV kilovolts









1 or L liter
lb pound (weight)
LB Luria-Bertani
LBN Luria-Bertani (sodium)
LDC test for lysine decarboxylase, red/orange = +
LI large intestine
logo logarithm base 10
M molar
m meters
M molar
M9 minimal media
MAN mannitol fermentation/oxidation test, yellow = +
MDa megadaltons
MEL melibiose fermentation/oxidation test, yellow = +
mg milligram
mg/kg milligrams per kilogram bodyweight
MgCl2 magnesium chloride
MgSO4 magnesium sulfate
MIC minimum inhibitory concentration
min minutes
ml milliliter
mM millimolar
mm millimeters
mmol millimoles
MOPS morpholinepropanesulphonic acid (buffer solution)
MQMFK modified Qiagen Midi Filter Kit for plasmid analysis
mRNA messenger ribonucleic acid
N normal
N/A not applicable
NaCl sodium chloride
NAHMS National Animal Health Monitoring System
NAL nalidixic acid
Nalr nalidixic acid resistant
NaOH sodium hydroxide
NAX ceftiofur sodium
ND none determined
NG nasogastric
NITR nitrofurantoin
No. number
NVFA non-volatile fatty acids
ODC test for ornithine decarboxylase, red/orange = +
ONPG test for beta galactosidase, yellow = +
OX oxidase test, violet = +
P propionate
PBS phosphate buffered saline









PCR polymerase chain reaction
PEN penicillin
PF pelvic flexure
PFGE pulsed field gel electrophoresis
pg picograms
PGMAA pH-gradient mediated anion accumulation
pH negative logarithm of hydrogen ion concentration
pKa negative logarithm of the acid dissociation constant Ka
PMN polymorphonuclear leukocyte
PO per os (orally)
ppm parts per million
PRAS pre-reduced anaerobically sterilized
q every
QBT Equilibration Buffer (Qiagen Midi Filter Kit)
QC Wash Buffer (Qiagen Midi Filter Kit)
QF Elution Buffer (Qiagen Midi Filter Kit)
R plasmid or factor resistance plasmid or factor
RFLP restriction fragment length polymorphism
RHA rhamnose fermentation/oxidation test, yellow = +
RIF rifampin
rpm revolutions per minute
rpoS alternative sigma factor (referring to the gene)
rpoS alternative sigma factor (referring to the protein)
RT room temperature
s seconds)
S. Salmonella
SAAAD Salmonella-attributed antibiotic-associated diarrhea
SAC sucrose fermentation/oxidation test, yellow = +
SC small colon
SCFA short-chain fatty acid
SD standard deviation
SDS sodium dodecyl sulfate
SEM standard error of the mean
SI small intestine
SOR sorbitol fermentation/oxidation test, yellow = +
SPF specific pathogen free
spp. bacterial species
spy Salmonella plasmid virulence (referring to the gene)
spv Salmonella plasmid virulence (referring to the protein)
subsp. sub-species
TBE tris-borate EDTA
TDA test for deaminase, brown/red = +
TE tris-EDTA
TET (oxy)tetracycline
TMP or TMPS trimethoprim sulfamethoxazole


xxiii









TNTC too numerous to count
TSP or TP total serum protein or total protein
U units
URE test for urea hydrolysis, red/orange = +
V valerate
v/v volume per volume
var. variant (serovariant)
VFA volatile fatty acid
VMTH Veterinary Medical Teaching Hospital
VP Voges-Proskauer test for acetoin, pink/red = +
w/v weight per volume
wt. weight
_ _bacterial strain


xxiv















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

EQUINE SALMONELLOSIS-MOLECULAR EPIDEMIOLOGY OF CLINICAL
ISOLATES AND THE EFFECT OF ANTIBIOTICS ON THE CECAL
MICROENVIRONMENT WITH PARTICULAR REFERENCE TO SHORT-CHAIN
FATTY ACIDS AND THE SALMONELLA PLASMID VIRULENCE (spv) GENES

By

Tamara Shea Vetro Widenhouse

May 2004

Chair: Guy D. Lester
Major Department: Veterinary Medicine

Antibiotic-associated diarrhea (AAD) is a common and potentially fatal disorder in

horses and is often attributable to Salmonella spp. Disturbances in anaerobic microflora

are thought to cause altered intestinal levels of bacteriostatic short-chain fatty acids

(SCFA). Salmonella virulence plasmid (spv) genes are reported to increase ability of

Salmonella to grow in organs outside the gastrointestinal tract. Horses treated with

intravenous oxytetracycline (TET), oral trimethoprim-sulfamethoxazole (TMPS), and

intramuscular ceftiofur (NAX) had significant differences in concentrations of seven

individual cecal SCFA with TET having the most significant effects, followed by TMPS

and NAX. No differences were detected in cecal protozoal counts, total culturable cecal

anaerobes, or cecal pH compared to untreated horses.

Epidemiological techniques were used to investigate 106 cases of salmonella

infection in horses at a veterinary teaching hospital over 2 years. Total mortality was









36.5%. Plasmid profiles, spy gene analysis, serovar, and antibiotic sensitivity data were

recorded for all isolates. Odds ratios predicted an increased risk of a fatal outcome in

horses younger than 4 years of age (3.3 times), horses infected with group B salmonellae

(15.7 times relative to group D), and horses whose salmonella isolate possessed the spy

genes (12.3 times). Extra-intestinal salmonella isolates were 12.2 times more likely to

contain the spy genes. The majority of large plasmids in salmonella serovars isolated

from horses were not virulence plasmids, but likely antibiotic resistance plasmids (3/3

tested transferred multiple resistances). This information suggests that the spy genes may

play a similar role in horses as they do in humans, mice, and calves: to potentiate

systemic infection after gastrointestinal infection.

Sterile-filtered cecal liquor from horses treated with ceftiofur or trimethoprim-

sulfamethoxazole increased the in vitro anaerobic growth rates of Salmonella relative to

plain media, and slightly more than untreated control horses cecal liquor. Salmonella

grew equally as well (but much slower than NAX or TMPS) in TET treated horses cecal

liquor and plain M9 medium. The SCFAs acetate, butyrate, and propionate, added to M9

minimal medium at 30mM or 100mM, exhibited a dose-dependent inhibition of

anaerobic salmonella growth that was not attributable to the spy genes, with propionate

100mM > butyrate 100mM > acetate 100mM z propionate 30mM > butyrate 30mM >

acetate 30mM.


xxvi














CHAPTER 1
INTRODUCTION

The Genus Salmonella

The Bacteria

The first mention of the yet-to-be-named genus Salmonella was a report in 1880 on

a "typhoid bacillus" observed in the spleen and mesenteric lymph nodes of a fever

patient.' A second organism discovered at approximately the same time, which failed to

agglutinate in serum from typhoid patients, was designated "bacille paratyphique." The

first documented cases of salmonellosis in animals were described by Salman and Smith

in 1886 of swine affected with hog cholera. This bacterium was later designated S.

Choleraesuis, and the genus eventually named after the former.1

Salmonellae are gram-negative members of the family Enterobacteriaceae. As of

August 2002, the genus is represented by 2,523 distinctive serovariants (serovars)2 of

flagellated, facultatively anaerobic bacilli.3 Salmonellae are speciated and sub-

characterized by their O (LPS), H (flagellar), and Vi (capsular) antigens. O antigens are

located on the surface of the outer membrane and are determined by specific

polysaccharide sequences. H antigens are expressed on flagella, and they are composed

of the proteins called flagellin. H antigens are biphasic and occur in either or both of two

forms, phase 1 and phase 2. The bacteria are capable of switching from one phase to the

other depending on environmental pressures. Vi is a unique antigen in that it overlays

the O antigen and is present in a limited number of serovars, the most important being

Salmonella Typhi, a host-adapted serovar of humans.4 5 These bacteria are stable and






2


ubiquitous in the environment, and they are capable of colonizing and infecting nearly all

higher species, although some serovars are known to have host preferences as well as

syndrome phenotypes.

Animal Models of Disease

Non-typhoidal salmonellae are global enteric pathogens of humans and other

vertebrates, and decades of research have been devoted to the epidemiology,

pathogenesis, diagnosis, control, and effective treatment of salmonellosis. To date, the

most economic and thoroughly characterized animal model of salmonellosis has been the

mouse. Salmonella infection in the mouse typically produces a syndrome of fever and

bacteremia. Until recently, it was thought that the mouse species did not develop enteritis

secondary to orogastric inoculation with Salmonella,6 which is typical of the pathogenesis

in most mammals, including humans and horses.7 In humans, cows, and horses, the

inflammatory reaction of the gastrointestinal tract is predominantly neutrophilic, while in

the mouse the mononuclear cell is the principal inflammatory cell.8 The serovar- and

route-dependant clinical response to experimental infection in the mouse is different to

that of most other vertebrates. Watson et al. showed that the cellular route of intestinal

invasion is different between mice and calves, with M cells and Peyer's patches being the

preferred targets in the murine host in contrast to enterocytes in calves.9 This creates

difficulty in the extrapolation of experimental data from mice to larger mammals.

Several alternative models to human gastrointestinal salmonellosis have been developed,

using the calf10 or pig,11J12 but these are expensive, logistically difficult to maintain, and

carry significant animal welfare concerns.46 In 2003, a newly proposed mouse model of

enteric salmonellosis was described and successfully tested by Barthel et al. in Belgium.13

This model more closely approximates the neutrophilic inflammatory infiltrate seen in









response to the bacteria, yet the mice still do not become diarrheic. Despite this, the

model is still a step forward in terms of salmonella investigation on genetic,

immunologic, and environmental levels, as there are readily available genetic knockout

strains and immunohistologic media and protocols developed for the mouse species.

The Salmonella Virulence Plasmid

The term plasmid is used to describe autonomously replicating extrachromosomal

DNA. This DNA is not critical to cell survival in vitro but can confer specific

characteristics that allow the host cell to survive during adverse conditions or to cause

disease.14 Pathogenic salmonellae possess a collection of these attributes, called

virulence factors.15;16 These include factors that convey acid resistance, enhance the

ability to invade non-phagocytic cells, elicit inflammation, support resistance to

destruction by phagocytic cells, suppress the immune system of the host organism,

enhance intracellular replication, and encode antimicrobial resistance. Several of these

factors can be attributed to the presence of a large 50-100 kb plasmid, originally termed

the "cryptic plasmid" as its purpose was unclear, but now described as a virulence

plasmid.17;18 The virulence plasmid of Salmonella has been characterized extensively in

the mouse typhoid model and appears to be most important in the ability of the organism

to multiply in systemic tissues after dissemination from the gastrointestinal tract.19-21

Clinical significance of this virulence plasmid has been examined in several studies, and

there remains disagreement regarding contribution of the virulence plasmid to bacteremia

and replication in extra-intestinal tissues. Clinical isolate data from several human

studies also agree with the murine model: virulence plasmids are more likely to be

present in those isolates obtained from systemic sources such as blood, liver and spleen,

compared to unrelated isolates obtained from feces.22;23 Conflicting reports utilizing









comparable experimental methods have shown no causal relationship between bacteremia

and presence of the virulence plasmid in humans.24;25 Discordant results are also seen

within the model using calves orally infected with S. Dublin. The virulence plasmid

containing the spy genes was shown to be important in determining the severity of

diarrhea in calves,10 while other investigators demonstrated no influence of the virulence

plasmid (as compared to naturally occurring plasmid-free, or plasmid-cured isolates) on

enteropathogenesis either in vivo, or in ex vivo ligated ileal loop experiments.26 This

serovar-host-syndrome interrelationship is most certainly a confounding factor in

determining the pathophysiologic importance of the salmonella virulence plasmid.

Salmonella Plasmid Virulence (spy) Genes

The plasmids of several serovars contain a 7.8-kb salmonella plasmid virulence

(spv) region, which contains five genes (spvRABCD) that are highly conserved across the

serovars that possess them.10;20;27 Those serovars tend to be natural host-adapted

salmonellae, including S. Dublin, S. Choleraesuis, S. Abortusovis, and S. Gallinarum-

Pullorum, but have also been found in broad host range serovars such as S. Typhimurium,

and S. Enteriditis.10 The genes contained within that small region are sufficient to replace

the virulence phenotype of the entire plasmid in animal systemic infection models.27

spvR encodes a transcriptional activator of the LysR/Met R family of regulatory proteins

and is transcribed independently from the four effector genes (spvABCD). SpvR binds to

the spvR and spvA promoters and directs transcription of itself and spvABCD during

stationary phase growth.28 The full significance of spv genes on bacterial pathogenicity is

becoming more clear, and they have been associated with enhanced virulence in mouse

systemic infection models,20;29-30;31 as well as showing enhanced expression after

invasion of both phagocytic and epithelial cells. 32;33 The spy genes are not necessary for









the bacteria to colonize the mouse gastrointestinal tract or invade mucosal cells to initiate

a systemic infection.20 They are also not required to survive in mouse secondary organs

such as liver and spleen.34 They have been shown, however, to accelerate proliferation of

the organism in the reticuloendothelial system,21 are essential to cause cytopathology in

mononuclear cells,35 and are associated with increased mortality in the calf model of oral

infection.10 A simplistic diagram showing the current opinion of how the spy genes are

regulated in salmonella serovars is shown in Figure 1-1.

Log Phase .... ... """"".............

R A B C D


a w*. W* rpoS .6****



Stationary Phase mRNA transcripts
or Intracellular .... .


R A B C D


rpoS
44040"04.. rpoS ..*



Figure 1-1. spy gene expression regulation is dependent on growth phase and cellular
location in Salmonella

A significant proportion of salmonella serovars isolated from clinical cases of

human22 and bovine23 diarrhea do not contain virulence plasmids and, therefore, the spy

genes. It appears that the ability of the spy genes in Salmonella to cause or enhance

pathology depends on other bacterial factors (e.g., chromosomal) as well as host factors.









The S. Enteriditis virulence plasmid containing the spy genes was placed into a plasmid-

cured S. Dublin strain, and virulence was restored in a mouse model,36 while the

virulence plasmid from S. Dublin pSDL2 only variably transferred a virulent phenotype

to serovars that did not commonly carry a virulence plasmid.37 Although the spy genes

are present and conserved across several serovars, many different syndromes and

outcomes of infection have been clinically or experimentally observed within those

subgroups. The work described in this dissertation attempts to analyze the role of the

virulence plasmid and the spy genes in the pathogenesis and epidemiology of equine

salmonellosis. 1038

Short-Chain or Volatile Fatty Acids and Salmonella

Salmonella spp. are enteroenvironmentally transmitted pathogens of humans and

animals. It is therefore expected that during their life cycle they are exposed to extremes

in temperature, oxygen availability, pH, osmolarity, nutrient availability, organic acid

concentration, and presence of other bactericidal compounds such as reactive oxygen

species. Salmonella have shown remarkable ability to sample their environmental

conditions and use this information as a signal for growth, stasis, or expression of

virulence factors. This ability is known as "quorum sensing" and is present in several

opportunistic and/or pathogenic species of bacteria.39 It has also been shown in

Salmonella that the induction of a stress resistance response to one condition, (e.g., low

pH) confers resistance to multiple stress conditions.40 Ironically, the end-result of

carbohydrate feeding to horses (instead of a complete forage-based diet) may actually be

priming resident or transient salmonella organisms and extending their spectrum of

resistance to organic acids and other stressors. This may indirectly predispose horses to

development of salmonella-induced diarrheal disease by seeding their environment with









bacteria that are more virulent than their acid-susceptible or otherwise stress-naive

cohorts.

Short-chain fatty acids are normally found in relatively high concentrations in the

forestomachs of ruminants,41 the cecum and large intestine of all warm-blooded

vertebrates,42 and the crop, cecum, and large intestine of birds.43 Acetate, butyrate, and

propionate are typically found in the highest percentages, with smaller amounts of

isomeric and variable sized carbon-chain compounds.41 In a general sense, these acidic

end products of anaerobic fermentation reactions help to keep the endogenous population

of bacteria within the intestines at a stable level and discourage transient pathogens from

becoming established. They can also be absorbed and function as an energy source for

the host animal, or they can be directly utilized by colonocytes.

Antibiotic-Associated Diarrhea (AAD) in the Horse

Diarrhea is one of the most common and recognized side effects of antibiotic

therapy in all species, especially the horse. Symptomatically, it can range from mild loss

of fecal consistency to projectile liquid feces and/or intestinal pseudomembrane

formation. A long-standing hypothesis suggests that disruption of the normal chemical

and biological balance within the intestine is responsible for the development of colitis,

either during or after the cessation of antibiotic therapy. This relationship may or may

not be true in horses. In one case-control study, horses which had received parenteral or

oral antibiotics were 40 times more likely to develop diarrhea than horses which had

received no therapy.44 Also, in a documented outbreak of salmonella diarrhea in a large

hospital, horses that had received parenteral antimicrobial therapy were at 10.9 times

greater risk of having Salmonella isolated from their feces than were matched cohorts not

receiving antibiotics.45 However, three longitudinal studies have demonstrated no clear









association between antibiotic administration and salmonella infection in horses.46-48

Any antibiotic, given by any route, to any horse, for any length of time, has the potential

to cause diarrhea,49 though orally administered agents and those drugs having a biliary

component to their metabolic-cycle pose a greater risk.50 Oxytetracycline,5155 penicillin

V and G,56 aminopenicillins,57 metronidazole,56 lincosamides,58-60 trimethoprim-

potentiated sulphonamides,61;62 third generation cephalosporins,56 and macrolides63 all

have diarrhea as a reported side effect in the horse, though there are conflicting data for

specific antibiotics (e.g., trimethoprim-potentiated sulphonamides).49;51 The situation

becomes pivotal in the equine species due to several factors, most importantly 1) the

large capacity of the digestive tract, therefore the potential of enormous amplification and

dissemination of the infectious agent into the environment, and 2) the intensive

management of horse operations-with overly susceptible animals such as neonatal,

geriatric, pregnant, and immunocompromised individuals often kept in direct contact with

asymptomatic animals shedding Salmonella. From a therapeutic standpoint the horse

also presents more unique challenges. First, the potentially large volume of fluid

excreted per day is difficult to replace-oral and/or parenteral fluid therapy is the

cornerstone of therapy in treatment of horses with large colon disease. Second, the horse

is uniquely susceptible to many secondary complications of enterocolitis that in and of

themselves could be as life-threatening as the diarrhea itself. The large bio-burden of

gram-negative bacterial cell wall (endotoxin) contained within the adult equine

gastrointestinal tract is more than adequate to cause severe disease or mortality should it

gain access to the circulatory system. Third, it has been shown clinically as well as

experimentally that horses can asymptomatically harbor and shed virulent organisms for









unpredictable amounts of time, either following acute infections or without previous

illness64;65 and the ability to positively identify a carrier animal based on appearance

alone is impossible.

The Gastrointestinal Microenvironment

The Normal Flora

The terms "resistance to colonization" or "competitive exclusion" have been used

to describe the passive ability of the gastrointestinal tract to keep pathogenic organisms

from becoming established.66;67 In humans, the anaerobic component of the commensal

microflora has been determined to be primarily responsible for maintaining the

colonization resistance toward pathogens.68 Despite the multitude of potentially virulent

organisms ingested on a continual basis, the innate functions of the intestinal

microenvironment restrict a pathogens ability to attach, multiply, invade and cause

disease. Intestinal anatomy and motility, mucosal epithelial and immune cells, the enteric

nervous system, residential bacteria, protozoa and their by-products, and mucosal

immunoglobulin all combine with digesta to comprise this effective barrier to

pathogens.69

The predominant species and demographics of the bacterial population change with

respect to the section of the intestine being colonized. Host diet, oxygen tension, pH,

redox potential, and intestinal motility all determine the constitution of the normal

intestinal flora, and even this may change on an individual or daily basis. Generally

speaking, anaerobic bacteria significantly increase as a percentage of the total bacteria

progressing aborally through the gastrointestinal tract.70 These anaerobic bacteria are

responsible for the breakdown of otherwise indigestible saccharide bonds and the

production of SCFA and gases such as methane and carbon dioxide. Short-chain fatty









acids are also important food sources for the colonic mucosal cells and are used by the

host organism as an energy source.42

Short-Chain Fatty Acids-Production and Intestinal Function

Short-chain fatty acids are bacterial by-products of fermentation reactions that

occur in an anaerobic environment. Non-spore forming anaerobes are the principal

facilitators of this process through the Embden-Meyerhof-Pamas pathway.71 They have

been studied extensively with respect to production sites, rates of appearance, and

biological fate in many species.41 SCFAs are important for development and proper

function of the rumen, intestine, and mucosal epithelium. The SCFAs, methane, carbon

dioxide, and hydrogen are the main end-products of anaerobic bacterial fermentation of

carbohydrates, while the branched-chain SCFAs are breakdown products of proteins and

are produced independently of the others.72;73

Table 1-1. Short-chain fatty acid chemical formulas and common names
Chemical Formula Common Name
CH3-COOH Acetate
CH3-CH2-COOH Propionate
CH3-(CH2)2-COOH Butyrate
CH3-CH-COOH Isobutyrate

CH3
CH3-(CH2)3-COOH Valerate
CH3-CH-CH2-COOH Isovalerate

CH3
CH3-(CH2)2-CO2CH2-CH3 Ethyl butyrate (ethyl butanoate)

Herbivores (especially the ruminants) obtain significant amounts of energy (up to

70-80% of daily maintenance) from the absorption and metabolism of SCFAs, which are

produced via bacterial breakdown of dietary lignin, pectin, cellulose, and hemicellulose.









SCFA production and anaerobic respiration pathways in the ruminant with substrates and

intermediate compounds are shown in Figure 1-2, modified from Van Soest.74


Figure 1-2. Polysaccharide metabolism and SCFA production pathways in the rumen.
Modified from Van Soest.74 Tan boxes indicate substrate, red boxes indicate
SCFAs, green boxes indicate NVFAs, blue boxes indicate important
intermediate compounds, and purple boxes indicate accumulated end-
products.







Humans and other monogastric species such as the dog obtain much less energy (6-
9%) from the utilization of endogenously produced SCFAs.42 Additional sources of
substrate include sloughed intestinal epithelial cells, blood, mucins, digestive enzymes,
and miscellaneous resistant starches.73 Figure 1-3 depicts the interrelationship between
anaerobic microbial function and the products of fermentation.

Carbohydrate:
N 3 Lignin, Pectin,
DAam N 3 Cellulose, Hemlcellulose


- --- ---------
N 3+
Carbon Skeleton
I Amino Acids

Microbial Maintenance

Microbial Growth
I


VFA: acetate, bul


at


I
I
I
I
I

P '





e, propionate
Gi


Figure 1-3. Diagram of major gastrointestinal microbial digestive and energy functions,
nitrogen and carbon recycling, and SCFA production
In addition to local consumption, SCFAs are shuttled directly into the portal
circulation for peripheral and hepatic metabolism. Short-chain fatty acid contributions to
maintenance energy requirements of the host range from less than 10% in humans and









dogs up to more than 80% in the ruminant and large-intestine fermenters such as the

horse.75 Those SCFAs are utilized by the rumen or ceco-colonic mucosal epithelial cells

as an energy source and also influence intestinal blood flow and water and electrolyte

secretion and absorption.73 Short-chain fatty acids are intimately involved in the proper

function and regulation of the terminal digestive processes as shown in Figure 1-4.73;75;76

"Colonic starvation" or "nutritional colitis" are phrases used to describe a diarrhea seen

in patients fed either total parenteral (intravenous) nutrition or enteral tube formulas low

in fiber.7 The hypothesis involves decreased SCFA production in the colon, with the

colonocytes becoming malnourished, leading to abnormal water and sodium absorption.

It was also shown that deranged fermentation in the large intestine in response to

antibiotic administration did not necessarily predict the development of diarrhea, though

all patients that developed antibiotic-associated diarrhea had fermentation

abnormalities.78 This suggests that purely the absence or impairment of SCFA synthesis

is not enough to cause diarrhea but may be an essential predisposing condition.






























Figure 1-4. Summary of potential enterotrophic effects of SCFA

Effects of Antimicrobial Therapy: Dysbacteriosis61

The incidence of AAD is estimated to be between 5-25% of all humans receiving

antibiotics, though patient risk group, type of antibiotic, and route of administration will

affect true prevalence.79 Current hypotheses suggest that the gastrointestinal side-effects

of antimicrobials are manifested through disruption of autochthonous anaerobic flora,

particularly Bacteroides, Bifidobacterium, Lactobacillus and Streptococcal spp.

Anaerobic bacteria are critical for fermentation of carbohydrates and production of

SCFAs, and it is these acids that are believed to have natural and regulatory bactericidal

and bacteriostatic properties against enteric commensals as well as pathogens.80 Several

investigators have reported significant disruptions in anaerobic flora and SCFA

concentrations in animals, humans, xeno-transplanted flora models, and in vitro colon

simulation systems treated with antimicrobials.81-88 Intestinal colonization and increased

multiplication rates of S. Typhimurium in response to streptomycin treatment in mice









were associated with decreased concentrations of fecal SCFAs and increased luminal

pH.89 In vitro supplementation of SCFA to the cecal contents of treated animals inhibited

salmonella growth in this model.80 Further studies have shown that this protection may

be conferred by specific SCFAs as elevated concentrations of propionic or formic acid

added to feed conferred significant protection against S. Typhimurium cecal colonization

in chicken hatchlings.90

Another theory links the etiology of AAD to the reduction or disappearance of

SCFAs. These acids are regulators of sodium and water uptake in the colon, and their

absence causes an indirect accumulation of sodium and water in the intestinal lumen.91

Sodium is a potent cellular osmolyte which draws more water across membranes and into

the lumen, causing increases in fecal water content. This theory does not adequately

account for the magnitude of diarrhea seen in some AAD patients, but it could easily be

an initiator or contributor to pathogenesis.

Alternative popular assumptions of the pathogenesis of AAD include unchecked

overgrowth of Clostridium difficile (especially in human neonates) with production of

potent entero- and cyto-toxins or the vacating of attachment sites or toxin receptors

normally occupied by host commensal bacteria.92 C. difficile has been identified as a

pathogen in equine AAD.93

Specific Aims

The specific aims of the reported studies were to:

* Collect Salmonella spp. isolates from clinical cases of equine salmonellosis and
from normal horses.

* Examine case history and collect relevant host data (age, breed, gender, presenting
disease, risk factors, biochemical profiles, antimicrobial susceptibilities, treatments)
for all salmonella isolates.









* Determine if the salmonella isolates carried large plasmids.

* Determine if the salmonella isolates carried spy genes.

* Determine the cecal SCFA concentrations, luminal pH, total culturable anaerobic
bacterial counts, and protozoal counts of horses before and after treatment with
selected antibiotics.

* To examine spv+ and spv- salmonella isolates in terms of growth rate during
anaerobic culture in nutrient broth supplemented with sterile-filtered cecal contents
from antibiotic-treated versus non-treated horses.

* To examine spv+ and spv- salmonella isolates in terms of growth rate during
anaerobic culture in nutrient broth adjusted to the mean luminal cecal pH and
supplemented with individual SCFAs normally found in horse cecal liquor.

* To examine plasmid containing spv- isolates for antibiotic resistance determinants
located on the plasmids.

Hypotheses

* Large plasmids in salmonella isolates are directly correlated with presence and type
of disease.

o Isolates from normal horses will not have plasmids.
o Isolates from cases of diarrhea will variably contain plasmids.
o Isolates from systemic cases will always contain plasmids.

* Salmonella isolates with large plasmids will also contain spy genes on those
plasmids.

* The administration of repeated doses of commonly used antimicrobial agents to
healthy horses will reduce the total culturable anaerobic bacterial population of the
cecum, resulting in a reduction in the concentration or disruption of relative
proportion of SCFAs and an increase in luminal pH.

* The administration of antimicrobial agents to healthy horses will reduce the
numbers of cecal protozoa.

* Sterile-filtered cecal contents from horses that were not treated with antibiotics will
inhibit the growth of Salmonella compared to sterile-filtered cecal contents from
animals that received antibiotics in a spv-dependant manner.

* Nutrient broth containing individual SCFA will inhibit growth of Salmonella under
anaerobic conditions in a dose-dependant and spv-dependant manner.






17


Large plasmids in salmonella isolates that do not contain spy genes are likely
antibiotic resistance plasmids.














CHAPTER 2
SALMONELLA IN HORSES-DISEASE DEFINITION AND GENERAL AND
MOLECULAR EPIDEMIOLOGY

Background

Disease Overview

In spite of pharmacological and therapeutic advances, diarrhea in the adult horse

continues to be one of the most challenging and frustrating medical syndromes facing the

equine veterinarian. Salmonella spp. are one of the primary etiological agents of equine

diarrhea, although a large number of diarrhea cases will progress or resolve without a

definitive diagnosis. Salmonella infection of horses is not limited to the intestinal tract.

There is potential for bacteremia, particularly in foals, with seeding of synovial

structures, bone, lung, umbilical remnants, brain and meninges, liver, and kidneys.

Salmonellosis can quickly become a financial disaster for the intensively managed horse

farm or equine hospital given the potential copious nature of contaminating feces

produced by one diarrheic adult horse, along with the environmental persistence of the

organism. There are also serious human health issues regarding the zoonotic potential

from treating and handling these animals.

There are four recognized clinical syndromes of salmonella infection in horses: 1)

an asymptomatic carrier or latent state;64 2) a severe and sometimes fatal fibrinonecrotic

enterotyphlocolitis; 3) bacteremia-with or without secondary foci of infection; and 4)

pyrexia, depression, and leukopenia without diarrhea-similar to the "enteric fever"

syndrome seen in humans infected with S. typhi.94









Prevalence

Excretion of Salmonella into the environment commonly occurs in horses without

signs of enteric disease. This may be an animal that has recently recovered from

infection, an animal that has acquired bacterial organisms via ingestion of contaminated

feed, water, or bedding material and is simply a transient portal, or a chronically

colonized host that has adapted a traditionally pathogenic relationship into a commensal

one. It is the apparently healthy, but chronically colonized animal that represents the

greatest danger to the population. Estimates vary widely depending on the population

sampled and the diagnostic methodology used of the percentage of the horse population

that is shedding Salmonella. The recent National Animal Health Monitoring System

(NAHMS) survey reported that 0.8% of resident horses sampled in the US excreted

Salmonella in their feces.95 The majority of horses in this survey had normal fecal

appearance at the time of sampling, although 2.1% had loose or watery feces. The

prevalence of salmonella shedding was not higher in animals that had received antibiotics

within the past 30 days.

Reported Risk Factors for Salmonella Infection

It has been noted that horses have an increased risk of developing salmonella-

induced diarrhea after certain "stressors" have been placed on them, including but not

96 47 54
limited to transportation,6 hospitalization,47 nutritional excess or deficiency,54 dietary

change,97 colic-especially large colon impaction,47;98 nasogastric intubation,44;45
44;45;48
debilitating injury or illness, antibiotic therapy,444548 parturition, weaning, surgery,

anesthesia, or anthelmintic therapy.99;100 The challenge inoculum for these "at risk"

individuals can be up to 100-fold smaller than for non-stressed and immunocompetent

cohorts.94;101;102 It is for these reasons that horses admitted to veterinary hospitals, even









on an outpatient basis, are highly susceptible to infection. The populations at greatest

risk are those horses with gastrointestinal diseases admitted to referral hospitals for

medical or surgical therapy.44;45;101

Salmonella Serovars Associated with Equine Infection

Approximately 60% of known salmonella serovars belong to the S. enterica subsp.

Enterica group and within this group the O-antigen designations A, B, C1, C2, D, & E

account for 99% of all warm-blooded animal infections. All O-antigen groups have been

isolated from horses, but groups B, D, and E are the most common.101 Commonly,

phenotypic and molecular analyses are married to form the most accurate picture of an

isolate as possible. Analysis of antimicrobial susceptibility, serogroup, serovar, phage

type, plasmid profile, ribotype, or restriction endonuclease examination allows more

specific identification of salmonella organisms. Newer and more precise methods of

distinguishing salmonellae include polymerase chain reaction (PCR) fingerprinting,

multiplex PCR, pulsed-field gel electrophoresis (PFGE),103 restriction fragment length

polymorphism (RFLP), IS200 typing,104 and real-time PCR.105 PCR has demonstrated

itself to be one of the most sensitive and expedient methods of detecting Salmonella spp.

in equine fecal samples, though culture is still the most cost-effective and widely

available.106 This is most helpful from an epidemiologic and control standpoint or for

biologic surveillance programs. Though the treatment does not vary between serovars,

specific identification could help in cases of outbreak, treatment failure, or when more

than one strain of Salmonella is suspected.

Salmonella serovars frequently reported isolated from horses over the last 40 years

include Agona, Anatum, Arizonae, Enterica, Enteriditis, Heidelberg, Infantis, Krefeld,

London, Miami, Muenchen, Muenster, Newport, Oraneienburg, Rubislaw, Saintpaul,









Senftenberg, Thompson, Typhimurium, and Typhimurium var.

Copenhagen.46;47;95;97;99;107111 Almost all serovars of Salmonella infecting horses are

non-host adapted strains,101 with the exception of Abortusequi, which does not cause

gastrointestinal disease, but rather early abortion in mares and systemic sepsis in newborn

foals.112 Horses are also susceptible to some of the normally host-adapted serovars of

other species such as S. Dublin (bovine) and S. Choleraesuis (porcine).110

The herbivorous and gregarious nature of horses makes them efficient dispersal

agents as well as susceptible recipients for the entero-environmental cycling of

Salmonella. Compounding this issue, salmonellae are ubiquitous and environmentally

resistant and can remain infectious in fecal material for years under the appropriate

conditions.113

Role of Microbial Virulence Factors in Equine Salmonella Infection

Specific virulence factors that mediate systemic or gastrointestinal salmonella

infections in horses have not been extensively studied. Likely this is due to reluctance or

difficulty in using the horse as a model of disease. Retrospective studies examining

isolates obtained from clinical cases of salmonellosis have been published, but

investigators focused on more epidemiological than molecular techniques of comparison.

Disease Prevention-Diet, Probiotics, Immunity

Methods utilized by veterinarians to decrease the morbidity and mortality of

salmonella infection in horses have either limited scientific basis or are applied based on

results obtained from other species. Very little information is available on specific

preventative strategies or therapies once clinical signs become evident.

Fructo-oligosaccharides (FOS) have been utilized extensively as feed additives in

the poultry and companion animal industries for many years. They exert their effects by









increasing the amount of fermentable carbohydrate that reaches the large intestine, which

can be acted upon by the bacterial population. This in turn raises the concentrations of

organic acids and drops the pH, which presents an inhospitable environment to

pathogenic species.114

"Direct-fed microbial" and "competitive exclusion" are terms used frequently in the

poultry industry to describe a practice and physiologic phenomenon of directly feeding or

facilitating the establishment of a desirable microbial population in order to discourage

colonization by an undesirable one, typically Salmonella. Transfaunation via fecal slurry

or cecal or colonic contents from a recently euthanatized or cannulated horse are

techniques used in a hospital situation to re-establish commensal protozoa and bacterial

flora in horses with diarrhea. Enemas of slurried fecal material from normal individuals,

have been shown quite effective at treating or preventing antibiotic associated diarrhea in

humans, but are unlikely to be beneficial in horses due to anatomical differences.115 A

commercial probiotic preparation is available for use in horses (Probios Equine One

Gel, Chr. Hansen BioSystems), however clinical efficacy data of this type of product in

horses is limited. In a prospective study of hospitalized horses neither of two commercial

probiotic formulations had any effect on salmonella shedding, incidence of diarrhea, or

length of hospitalization following abdominal surgery.98 A recent prospective study

examining the probiotic potential ofLactobacillus rhamnosus strain GG in horses failed

to show efficient colonization of the adult gastrointestinal tract unless extremely large

doses were administered, though foals were more consistently and efficiently

colonized.116 These conflicting results should be further investigated, as human evidence

is strongly in favor of the use of direct-fed microbials in the prevention and management









of antibiotic-associated diarrhea or other diarrheas attributed to dysbacteriosis.

Significant benefit could be obtained from a small daily dose of orally administered

bacteria during periods of increased susceptibility to salmonellosis, such as during

extended travel or preceding and concurrent with antibiotic administration.

Immunity to Salmonella is dependent on a combination of cell-mediated

recognition and destruction by activated granulocytes, as well as an antibody driven

humoral response. Salmonella antibody-containing equine plasma products are

commercially available. These products are almost exclusively used for the treatment of

systemic salmonellosis in foals, or as preventative therapy in foals with failure of passive

transfer in areas with a history or high prevalence of disease. These products are usually

cost prohibitive for use in adult horses, and more importantly, are serovar specific, thus

providing no cross protection to the significant number of other serovars able to infect

horses. Mucosal immunization of horses with mutant strains of Salmonella rendered

non-pathogenic has also been examined. Sheoran et al. demonstrated strong production

of S. Typhimurium specific mucosal IgA in jejunal, nasal, and vaginal compartments

after intra-nasal vaccination of ponies with a Acya Acrp-pabA mutant of S.

Typhimurium.117 This strain is attenuated for virulence by deletion of the genes

necessary for adenylate cyclase production (cya) and the cyclic AMP receptor protein

(crp). This live vaccine did not cause any signs of disease, was not shed in the feces, nor

was it transferred to cohabitated non-vaccinates. Mucosal specific antibody is an

attractive first line of defense against enteric pathogens, and exploitation of the

gastrointestinal mucosal immune system in the horse is attractive in terms of prevention

and protection.









Disease Treatment

The treatment of salmonella infection is controversial and dependent on several

factors, including severity of disease, immune status, metabolic state, age, concurrent

malignancy, drug cost, drug availability, side-effects, and the presence of colonizable foci

(e.g., implanted materials, catheters). Conventional antibiotic therapy of uncomplicated

salmonella gastroenteritis in human beings is often not efficacious and may actually

prolong the convalescent phase and/or extend the length of time that Salmonella is shed

from the feces.118-120 Even antibiotics preferred for the directed therapy of Salmonella in

horses and humans (e.g., fluoroquinolones) have not had any scientifically reproducible

or predictive effects on fecal carriage post-infection. Post-convalescent shedding is an

important salmonella-related morbidity issue facing the equine practitioner.

Contamination of the environment with persistent, virulent, and potentially antibiotic

resistant bacteria is a cause for concern in a horse facility, especially a veterinary

hospital. Outbreaks of nosocomial salmonellosis have resulted in institutional shut-

downs world-wide.109;121-128 In these circumstances antibiotic therapy of clinically silent

or uncomplicated cases would be useful if the period of environmental contamination

could possibly be shortened, thereby limiting exposure of other animals while the facility

is depopulated and disinfected.129

The Salmonella Virulence Plasmid

Clinical isolate data from several human studies agrees with the murine model of

salmonellosis: virulence plasmids are more likely to be present in those isolates obtained

from systemic sources such as blood, liver, and spleen, compared to unrelated isolates

obtained from feces.22;23 Conflicting reports utilizing comparable experimental methods

have shown no causal relationship between bacteremia and presence of the virulence









plasmid in humans.24;25 Discordant results are also seen within the model using calves

orally infected with S. Dublin. The virulence plasmid containing the spy genes was

shown to be important in determining the severity of diarrhea in calves,10 while other

investigators demonstrated no influence of the virulence plasmid (as compared to

naturally occurring plasmid-free or plasmid-cured isolates) on pathogenesis either in vivo

or in ex vivo ligated ileal loop experiments.26 This serovar-host-syndrome

interrelationship is most certainly a confounding factor in determining the

pathophysiologic importance of the salmonella virulence plasmid.

Salmonella Plasmid Virulence (spv) Genes

Function of the spy genes

The function of the spy genes in Salmonella has been a focus of investigation for

many years. Highly conserved genomic elements should theoretically be important to the

survival and host-to-host transmission of pathogenic bacterial species. Of the entire spy

locus, it has been shown that only spvB and spvC are essential for full virulence in the

mouse model of subcutaneous infection,31 and more recently, that spvB was required for

cytotoxic pathology (progressive detachment of adherent cells, vacuolization) and

apoptosis after phagocytosis by human monocyte-derived macrophages.35 The apparent

accelerated growth of spy positive strains (as compared to spy negative) and their ability

to cause systemic disease may actually be an extension of their ability to survive and

travel within macrophages to these sites. A summary of the current understanding of the

molecular and functional information regarding the spy genes can be found in Table 2-1.









Table 2-1. spy gene characteristics
Gene Activity Protein Significance
Localization
spvR Transcriptional Cytoplasm Positive regulator (promoter)
activator of spvR of itself and the other spy
and spvABCD genes
spvA unknown Outer Unknown, mutations do not
membrane130 affect virulence in mouse
intraperitoneal infection model.
spvB ADP- Cytoplasmic and Essential for full virulence
ribosyltransferase transported out of (mouse model).31 Effector
cytoplasm, small protein causing
amounts in inner depolymerization of actin
membrane130 cytoskeleton within
macrophages. Inhibition of
phagolysosome fusion.
spvC unknown Cytoplasm Essential for full virulence
(mouse model)31
spvD unknown Exported outside Mutations attenuate virulence
_of cell (mouse model)131

Significance of the spy genes

The role of spy genes in equine salmonella infection has not been investigated.

There is conflicting evidence, as demonstrated in the mouse and calf models, that spy

genes play a primary role in the establishment and persistence of systemic infections and

do not contribute significantly to the enteric phase of the disease. Anaerobiasis was

shown to significantly retard the growth rate of Salmonella with a significantly reduced

cell density at stationary phase, and the spy genes were not expressed.132 This lends

further support to the hypothesis that the spy genes are not involved in the enteric phase

of infection, but this has not been examined in species other than the calf and mouse.

Specific Aims

The overall aim of this section of the study was to describe the general and

molecular characteristics of Salmonella spp. isolated from hospitalized symptomatic









animals in North Central Florida and contrast these isolates from those collected from

asymptomatic animals in the same geographic region. The specific aims were:

* To collect, describe, and store Salmonella spp. isolates from hospitalized horses.
* To collect, describe, and store Salmonella spp. isolates from asymptomatic horses
at pasture.
* To determine if the salmonella isolates carried plasmids and classify them based on
size.
* To determine if the salmonella isolates carried spy genes.
* To examine plasmid containing spy negative isolates for antibiotic resistance
determinants located on those plasmids.

Materials and Methods

Case Selection

Bacterial cultures were obtained from hospitalized foals and adult horses with

clinical signs consistent with salmonella infection. Material submitted to the clinical

microbiology laboratory included feces, gastric secretions, blood, synovial fluid, and

tissue samples from post-mortem examinations. Sequential fecal samples were also

collected from asymptomatic horses at several farms in the North Central Florida area,

over a period of 24 months. Individual records were kept for each animal and horses

were sampled at least three times on separate occasions.

Microbiological Techniques

Field samples

Freshly voided or rectal fecal samples were collected and placed into sterile,

labeled containers. Two to five grams of fecal material was placed into selenite broth and

incubated at 370C in a 5% CO2 environment for 12-18 h to maximize isolation of

Salmonella spp. A Hektoen-Enteric plate (Remel Inc., Lenexa, KS) was streaked for

isolation from the overnight culture broth. The plates were incubated 18-24 h at 370C in

a 5% CO2 environment. Non-lactose-fermenting and H2S-producing colonies were









selected and streaked onto urease slants (Remel Inc., Lenexa, KS) which were incubated

18-24 h at 370C in a 5% CO2 environment. Urease-negative organisms were further

characterized utilizing API 20E enteric test strips (bioMerieux USA, Durham, NC) for

positive identification of Salmonella spp. An incubated strip with reactions typical of

Salmonella is shown in Figure 2-1. Tests and interpretations from left to right include

ONPG -, ADH +, LDC +, ODC +, CIT +, H2S +, URE -, TDA -, IND -, VP -, GEL -,

GLU +, MAN +, INO -, SOR +, RHA +, SAC -, MEL +, AMY -, ARA +





OPO L i I NO jig L' HgACL- ME-A I AV

Figure 2-1. API20E rapid identification strip showing typical reaction results for
Salmonella spp.

Clinical and reference isolates

All clinical isolates were provided as pure cultures on Hektoen-Enteric (HE) agar

plates (Figure 2-2) by the clinical microbiology service at the University of Florida

College of Veterinary Medicine, Gainesville, Florida.


Figure 2-2. Salmonella group C2 isolate as provided on Hektoen-Enteric agar









Bacterial cultures of all salmonella reference strains and clinical isolates were

subsequently grown in Luria-Bertani broth (LB) or on LB agar without antibiotics at

35'C, in a 5% CO2 atmosphere unless otherwise indicated. One and a half percent (w/v)

agar was added to LB broth for plates. Composition of culture media is in Table 2-2.

Table 2-2. Composition of bacterial culture media
Media Ingredients per Liter Sterilization Storage
and/or Supplier with
Catalog Number
LB Broth 10 g tryptone Autoclave 15 RT
5 g yeast extract min at 1210C
5 g NaCl
Sigma-Aldrich
L3152
LB Agar 10 g tryptone Autoclave 15 2-60C
5 g yeast extract min at 1210C
5 g NaCl
15 g agar
LB-N Broth 10 g tryptone Autoclave 15 RT
5 g yeast extract min at 1210C
8.5 ml 5 M NaCl (0.85%
w/v)
Hektoen-Enteric Agar Sigma-Aldrich Autoclave 15 2-60C
H7532 min at 1210C
Minimal Medium (M9) 200 ml 5x M9 Salts Autoclave 15 RT or 2-
20 ml 1 M glucose min at 1210C 60C
2 ml 1 M MgSO4 *before*
0.1 ml 1 M CaC12 addition of filter
sterilized
glucose
Brain-Heart Infusion Sigma-Aldrich Autoclave 15 RT
(BHI) Broth B7403 min at 1210C
Selenite Broth w/ Remel Pre-sterilized 2-60C
Cystine 064506 commercial
product
Urea Agar Remel Pre-sterilized 2-60C
065210 commercial
product
PRAS Brucella blood- Anaerobe Systems Pre-sterilized, RT
agar with 75 AS-141G custom
micrograms per ml manufactured,
gentamicin commercial
product










Antibiotic supplementation to LB agar plates, when referenced, was made in the

following concentrations: ampicillin (AMP) 100 micrograms/ml; nalidixic acid (NAL) 50

micrograms/ml; ceftiofur sodium (NAX) 8 micrograms/ml; cefazolin sodium salt (CEF)

8-32 micrograms/ml; and tetracycline hydrochloride (TET) 25 micrograms/ml.

Salmonella Identification and Antibiotic Resistance Profile

Isolates identified as Salmonella were grouped using group specific antisera (Fisher

Scientific International, Hampton, NH) and serotyped through National Veterinary

Service Laboratories (NVSL), Ames, IA. Those isolates positively identified as

Salmonella were sub-cultured and frozen.

Antibiotic resistance profiles were determined for each isolate via an automated

minimum inhibitory concentration (MIC) system (Sensititre Microbiology Systems,

software version-SAMS V2.3 Release 1, Trek Diagnostics, Cleveland, OH, USA).

Salmonella Isolate Storage

Subcultures of salmonella strains were stored as pure cultures at -800C in LB or

Brain-Heart Infusion (BHI) and 35% (v/v) glycerol. A standing overnight culture was

prepared by selecting approximately ten to fifteen colonies from the plate provided by the

microbiology laboratory. Multiple colonies were sampled to avoid the selection of any

single genotype or abnormal colony. The isolate was inoculated into LB and incubated at

37C without agitation. The following morning, 1 ml of this culture was added to 30 ml

of either BHI or LB and incubated at 37C with agitation for 1 to 1.5 hours (approximate

OD600 = 0.5-0.6). The sample was centrifuged at 10,000 x g for 10 min to pellet the

cells. The supernatant was removed, and the cells were re-suspended into 2 ml BHI or

LB. Two ml of 70% glycerol was added to the cell suspension and mixed gently. The









isolates were transferred immediately to pre-labeled standard cryogenic storage vials and

flash frozen in a dry ice and ethanol bath.

Reference Strains

Reference Salmonella serovar Typhimurium strains x3306 and x3337 were kindly

provided by Dr. Paul Gulig in the Department of Molecular Genetics and Microbiology at

the University of Florida College of Medicine. Specific information regarding these

strains is detailed in Table 2-3. The sequenced size of the S. Typhimurium strain LT2

virulence plasmid pSLT is 93,939 bp,133 but the virulence plasmid sizes of similar strains

such as SR-11 may vary. The reported size for the virulence plasmid of the strain used

herein is approximately 100,000 bp.

Table 2-3. Salmonella serovar Typhimurium reference strains used in this study
Strain Genotype Source Phenotype
SR-11 73306 gyrA]816, pStSR100 Dr. Paul Gulig Nalr, virulent, spy
SR-11 X3337 gyrA1816, pStSR100l Dr. Paul Gulig Nalr, spv-, avirulent,
plasmid cured derivative of
73306

Plasmid Profiling of Salmonella Isolates

Plasmid extraction was achieved using a modification of a commercial kit for large

construct and very low copy number plasmid purification (Qiagen Filter Midi Kit,

Qiagen, Inc., Valencia, CA). The S. Typhimurium plasmid copy number has been

estimated to be between 2-3 per cell.133 Bacteria were grown in 50 ml LB for 12-16 h

(approximately A600 = 1 1.5). Cells were divided into two sterile 50 ml polypropylene

centrifuge tubes and pelleted by centrifugation at 7,000 rpm in JA-20 rotor for 15 min at

4C. The cells were re-suspended thoroughly by vortexing in 10 ml buffer P1-

Resuspension Buffer per tube. Ten ml of buffer P2-Lysis Buffer was added to each tube,

the cells were mixed by gentle rolling and inversion and incubated at RT for 5 min









exactly. Ten ml of chilled buffer P3-Neutralization Buffer was added per tube and the

samples were mixed immediately but gently by inversion. The tubes were incubated for

15 min at RT. Columns (Qiagen-tip 100, Qiagen Inc., Valencia, CA) were equilibrated

with 4 ml of buffer QBT and columns were allowed to empty by gravity flow during this

incubation to be ready when needed. Samples were poured into pre-labeled high-speed

centrifuge tubes (Oak Ridge Centrifuge Tubes, Fisher Scientific International, Hampton,

NH) and centrifuged at 15,000 rpm in a Beckman JA-20 rotor, for 10 min at 40C. The

supernatant was removed and applied to a vertically supported filtration syringe

(QIAfilter cartridge, Qiagen Inc., Valencia, CA), and the plunger was inserted. The

filtrate was dispensed onto the columns gently and slowly, over a period of

approximately 10-20 min, keeping visible sample in the reservoir of the column at all

times. The column was then washed with 2 volumes of 10 ml buffer QC-Wash Buffer at

RT and allowed to empty by gravity flow. The wash solutions were discarded and clean

40 ml high-speed centrifuge tubes were placed under the columns to collect the eluted

DNA. Plasmid DNA was eluted with 5 ml 56C buffer QF-Elution Buffer per column.

The plasmid DNA was precipitated with 0.7 volumes (4 ml per isolate) of RT

isopropanol. The samples were centrifuged at 16,000 rpm in a JA-20 rotor for 30 min at

4C and the supernatant was removed. The DNA pellet was washed with 70% ethanol,

dried, and re-suspended in 150 microliters of TE for agarose gel analysis and

transformation experiments. A summary of the solution ingredients for the modified

plasmid extraction procedures and storage is found in Table 2-4.

Agarose gel analysis was performed via common method. Equal volumes of the

plasmid extract and 10x sample loading buffer were combined as a droplet on paraffin









paper, loaded, and run on a 0.5% agarose in Tris-borate-EDTA (TBE) gel at 125 volts for

1.5 h. Size was estimated by comparison to the approx. 100-kb plasmid from S.

Typhimurium x3306 run on the same gel in addition to a super coiled DNA ladder with a

range from 16.2 to 2-kb pairs (Gibco BRL, Carlsbad, CA). Plasmid bands were

visualized by staining the gel with ethidium bromide (1 microgram/ml) and photographed

using a digital gel imaging and documenting system (Chemi System, UVP Biolmaging

Systems, Upland, CA).

If an isolate had at least one large plasmid it was considered plasmid-positive. A

plasmid was only considered to be a virulence plasmid if the spy gene primer sets

hybridized to the isolate. Otherwise, it was simply a large plasmid of unknown type.

Table 2-4. Composition of buffers and solutions used in plasmid extraction protocols
Reagent Composition Storage
P1 (Resuspension Buffer) 50mM Tris-HCl pH 8.0 2-80C
10mM EDTA
100 micrograms/ml RNase A
P2 (Lysis Buffer) 200mM NaOH RT
1% SDS (w/v)
P3 (Neutralization Buffer) 3.0M potassium acetate pH 5.5 RT or 2-80C
QBT (Equilibration Buffer) 750mM NaCl RT
50mM MOPS pH 7.0
15% isopropanol (v/v)
0.15% Triton X-100 (v/v)
QC (Wash Buffer) 1.OM NaCl RT
50mM MOPS pH 7.0
15% isopropanol (v/v)
QF (Elution Buffer) 1.25M NaCl RT
50mM Tris-HCl pH 8.5
15% isopropanol (v/v)
Tris-EDTA (TE) 10mM Tris-HCl pH 8.0 RT
ImM EDTA
10x Sample Loading Buffer 40% Sucrose RT
0.17% Xylene Cyanol
0.17% Bromophenol Blue









Polymerase Chain Reaction (PCR) Identification of spy Genes

PCR was performed to evaluate all clinical isolates for the presence of the spy

genes. Positive (x3306) and negative (x3337) isolates, as well as a series for a

chromosomal gene, aspartate semialdehyde dehydrogenase (asd), were run as controls in

each experiment. These controls were vital for two reasons: 1) to verify that the

reactions, reagents, and conditions were appropriate, and 2) to ensure that the isolates

were truly Salmonella spp. (which was most important in validating negative reactions).

A loop of pure culture was added to 200 microliters of sterile water in a 1.5 ml

microcentrifuge tube, and boiled for 10 min to be used as template DNA. The master

mix and all reactions were prepared and maintained on ice until the run was started.

Master mix was made fresh for each experimental run, and consisted of 24.75 microliters

deionized H20, 5 microliters 10x PCR buffer, 8 microliters 1.25mM dNTP mix, 0.25

microliters 5U/microliter Taq DNA polymerase, and 4 microliters 50mM MgC12

(GibcoBRL, Carlsbad, CA). Five microliters of template DNA was added to each tube

for a total reaction volume of 50 microliters per tube. The samples were placed in a

thermocycler (Programmable Thermal Controller PTC-100, MJ Research, Inc. Reno,

NV) for the cycle described in Table 2-5.

Table 2-5. Times and tem )eratures for PCR reactions
STEP 1 Melt 940C 180 seconds
STEP 2 Melt 940C 60 seconds
STEP 3 Anneal 450C or 500C 60 seconds
STEP 4 Extend 720C 120 seconds
REPEAT STEPS 2, 3, & 4 FOR 30 CYCLES
STEP 5 End 720C 180 seconds
STEP 6 Hold 40C indefinitely

Primer sets for the spy gene PCR were provided by Dr. Paul Gulig. These

consisted of 3' and 5' primers for asd, spvA, spvC, and spvR. All clinical isolates were









examined for presence of the asd gene of Salmonella and probed with at least two

different spy gene primer sets. An isolate was determined to be positive for the spy gene

locus if two conditions were met: 1) if the asd product was present, and 2) if two or more

of the spy gene products were present. These primer sets are situated to bracket the entire

open reading frame of the gene, so the PCR products are in essence whole spy genes. All

isolates were probed with no less than two spy gene primer sets each (usually spvA and

spvC, occasionally including spvR). These primers were extremely effective in their

ability to identify the genus Salmonella and the presence of spy genes. Sequence and

other important primer information are contained in Table 2-6.

Table 2-6. Primers utilized in PCR reactions
Primer Sequence Product
Size
spvA 5' 5'-CCCCCGGGATGAATATGAATCAGACCACCA-3' ---
spvA 3' 5'-GGGAATTCTGGTAGCGCGGGAAGC-3' z784 bp
asd 5' 5'-CAGCACATCTCTTAGCAGGAAAAAAACGC-3' ---
asd 3' 5'-GGGAAGCTTCTACGCCAACTGGCGCA-3' 1,100 bp
spvR 5' 5'-CCCCGGGATCCATGGATTTCTTGATTAATAAA-3' ---
spvR 3' 5'-CCCCGGGAATTCGCTGCATAAGGTCAGAAGG-3' z905 bp
spvC 5' 5'-CCCCCGGGATGCCCATAAATAGGCCTAATC-3' ---
spvC 3' 5'-GCCGGAATTCGTCAGTAAGGG-3' z875 bp

Salmonella Plasmid Transformations into Susceptible Bacteria-Effects on
Antibiotic Resistance

Based on a discovery that the minority of large plasmids in the clinical salmonella

isolates were virulence plasmids, transformations of extracted plasmid DNA into a select

antibiotic-sensitive strain of Escherichia coli were performed to investigate the

possibility that the large plasmids may be carrying antibiotic resistance (R) determinants.

Three clinical salmonella isolates were selected based on their antimicrobial

sensitivity profiles and the presence of a single large plasmid that did not contain the spy

genes. Successful transference was confirmed through plasmid extraction of the









transformed E. coli isolates and gel electrophoresis with untransformed E. coli as well as

the original plasmid extracts used for the transformation. Briefly, the procedure was

performed as follows: competent E. coli DH5a cells (Invitrogen Corporation, Carlsbad,

CA) were thawed on ice. A 40 microliter aliquot of E. coli was added to an ice-cold

electroporation cuvette. All solutions were maintained on ice throughout the procedure

unless otherwise specified. Two microliters of plasmid extract in TE was added to the E.

coli and mixed gently with a pipette. The mixture was electroporated at 1.25 kV, 25

microfarad capacitance 200 ohms resistance, on a Bio-Rad Gene-Pulser (Bio-Rad,

Hercules, CA). The time constant was as close to 4.9 as possible, and if below 4.5, the

procedure was repeated with less plasmid extract. Nine-hundred microliters LBN broth

was added to the cuvette and it was incubated in a water bath at 370C for 30 minutes.

The cell suspension was transferred to a 1.5 ml microcentrifuge tube and 100 microliters

was spread onto several different LB agar plates, each containing a relevant selective

antibiotic based on the antimicrobial resistance profile of the original salmonella isolate.

The plates were incubated overnight at 370C and observed for growth the next day.

Transformants (as evidenced by growth on selective plates) were grown in LB broth with

continued selective pressure and subjected to the same extraction procedure described

previously to verify the presence and size of newly acquired plasmid DNA.

Chemical transformations were performed by mixing 5 microliters of extracted

donor plasmid DNA in water with 40 microliters of recipient strain (same DH5a E. coli

as for electroporation transformations). The mixture was incubated on ice for 15 min and

then transferred to 42C water bath for 2 min to heat-shock the cells. Five-hundred

microliters of LBN broth was added, and the mixture was incubated at 37C without









shaking for 30 min. One-hundred microliters of the mixture was then plated on

appropriate selective plates as described for the electroporation transformation.

Statistical Methods

Clinical isolates, patient information, and resistance data were collected weekly as

they became available over a period of three years (December 1999 through September

2002). Survival was based on discharge from the hospital. Animals were initially

grouped with respect to age into the following categories: <0.5y; 0.5-4y; 5-8y; 9-12y; 13-

15y; and >15y. They were then further grouped as follows for statistical comparison: 0-

5y; 6-15y; and >15y. Cases with missing data were excluded from calculation of

descriptive percentages (e.g., survival, gender, age). The effects of gender, age, and

breed on outcome were investigated independently using the Pearson Chi-square test.

Stepwise logistic regression was used to form a model in which multiple clinical

variables could be used to predict outcome. A Kruskal-Wallis test was used to

investigate the percentage of submitted samples that were positive with respect to

outcome.

Results

Asymptomatic Population

Isolates were sought from asymptomatic animals in order to contrast bacterial

genotype with isolates from clinical cases. Salmonella spp. were not identified from any

of 381 cultures performed on 105 different asymptomatic horses over a period of two

years.

Clinical Cases

There were 106 hospitalized animals during the period of interest that had at least

one positive culture of Salmonella. Within this population there were more males (61%)









than females (39%), although this difference was not significant. The mean age of the

population was 4.9 years, with a range of 2 days to 38 years. The age distribution by

categorical group is shown in Figure 2-3. Fifty seven percent of the clinical cases

survived. Of those non-survivors 26% died spontaneously. Breed distribution for the

affected horses is shown in Table 2-7.

Table 2-7. Breed distribution of 84 equine salmonella cases 1999-2002
Breed No. of Cases % of Cases
Thoroughbred 32 38.09
Quarter Horse 15 17.86
Paso Fino 10 11.90
Miniature Horse 5 5.95
Arabian 4 4.76
Paint Horse 4 4.76
Warmblood 4 4.76
Standardbred 3 3.57
Pony 3 3.57
Mixed Breed 2 2.38
Draft 1 1.19
Appaloosa 1 1.19

Relationship Between Gender or Age and Outcome

There was no gender bias with respect to short-term survival (Table 2-8). There

was a statistically significant association between age and case outcome (Table 2-9).

Horses less than 5 years of age were 3.3 times more likely to die when infected with

Salmonella than older animals.

Table 2-8. Effect of gender on mortality in 96 cases of equine salmonellosis*
Gender Died Survived % Dead Odds Ratio Lower 95%CI Upper 95%CI
Female 13 19 40.6 1.25 0.54 2.88
Male 22 30 42.3 1.00 0.47 2.13
*Pearson Chi-square value of 0.277 with 1 degree of freedom, p=0.599






















Number of
Cases 15






<6mo 6m-4y 5y-8y 9y-12y 13y-15y >15y
Age


Figure 2-3. Age distribution of 98 equine salmonella cases 1999-2002

Table 2-9. Effect of age on mortality in 85 cases of equine salmonellosis*
Odds Lower Upper
Age Group Died Lived % Dead Ratio 95%CI 95%CI
0 4 yrs 28 28 50.00% 3.33 1.18 9.42
5 15 yrs 6 17 26.09% 1.00 0.28 3.63
> 15 yrs 2 4 33.33% 1.67 0.24 11.45
Pearson Chi-square value of 6.002 with 2 degrees of freedom, p = 0.05.

Case Seasonality

Cases were examined for the month of occurrence and the data are shown in Figure

2-4. The majority of cases in the present study occurred during the warmer months of the

year, with 68% between the months of April and September. Thirty-year average

minimum temperatures in Gainesville, Florida remained above 60.3oF (15.7C) during

the months of May through October (Table 2-10).







Table 2-10. Average minimum temperatures in Gainesville, Florida, USA (1961-1990)*
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year Avg
oC 6.1 6.8 10.2 13.1 17.1 20.6 21.7 21.8 20.7 15.7 10.9 7.4 14.3
OF 43.0 44.2 50.4 55.6 62.8 69.171.171.2 69.3 60.3 51.6 45.3 57.7
*Obtained from www.worldclimate.com


IIII


IIEEE


0


Figure 2-4. Seasonal distribution of salmonella cases from horses 1999-2002
Group and Serovar Distribution
All serovars in the present study have been previously identified in horses with the
exception of a group F Salmonella serotyped as Rubislaw. There were three cases
serotyped as Salmonella 4,5,12:i-monophasic, a type closely related to S. Typhimurium,
whose antigenic formula is 1,4,5,12:i-1,2 biphasic.134 Eight isolates (7.5%) were not
serotyped as the samples were either contaminated with other bacterial genera or
contained more than one group or serovar of Salmonella. The breakdown of salmonella
group, serovar, and prevalence in this study is summarized in Table 2-11.


,..111111111









Forty-eight additional isolates from environmental sampling and animal species

other than horses were collected and archived during the reporting period. These isolates

were serotyped, and antibiotic sensitivity profiles determined, but no other analyses were

performed. Serovar and species of isolation information regarding these isolates is

detailed in Table 2-12.

Table 2-11. Salmonella serovars isolated from 98 equine cases 1999-2002
Serovar Group Number of Cases % of Isolates
Java B 23 23.45
Newport C2 13 13.27
Typhimurium B 8 8.16
Typhimurium var. Copenhagen B 7 7.14
Javiana D 7 7.14
Miami D 7 7.14
Saintpaul B 6 6.12
Muenchen C2 5 5.10
Anatum E 4 4.08
4,5,12:i-monophasic B 3 3.06
Newington E 2 2.04
London E 2 2.04
Mbandaka Cl 2 2.04
Hartford Cl 1 1.02
Agona B 1 1.02
Braenderup Cl 1 1.02
Infantis Cl 1 1.02
Meleagridis E 1 1.02
Reading B 1 1.02
Rubislaw F 1 1.02
Tallahassee C2 1 1.02
Thompson Cl 1 1.02


Table 2-12. Salmonella isolates of environmental and species
1999-2002


other than equids collected


Isolate Origin Serovar Group No. of
Isolates
Avian Manila E 1
Avian Infantis Cl 1
Bovine Typhimurium B 4
Bovine Typhimurium var. Copenhagen B 4
Bovine Anatum E 2
Bovine Newport C2 1









Table 2-12. Continued
Isolate Origin Serovar Group No. of
Isolates
Bovine Mbandaka Cl 1
Canine Typhimurium var. Copenhagen B 2
Canine Adelaide Not A-E 2
Canine Miami D 1
Environmental Java B 15
Environmental Newport C2 3
Environmental Typhimurium var. Copenhagen B 2
Environmental Javiana D 1
Environmental Anatum E 1
Environmental Typhimurium B 1
Environmental Tallahassee C2 1
Non-human Primate Typhimurium B 1
Other Mammal Hartford Cl 1
Reptile Sub group 3 1
Rodent Typhimurium var. Copenhagen B 1

Outcome by Group or Serovar

The relationship between group or serovar and outcome was investigated. The

results are presented graphically in Figure 2-5. There was a significant difference

between isolates according to antigenic grouping in terms of mortality (p=0.033; Figure

2-6); survival was decreased with isolation of group B salmonella serovars (43%

survival). Other groups included Cl (60% survival), C2 (59% survival), D (92%

survival), E (67% survival), and F (100% survival). Odds ratio data were determined by

salmonella group; if the horse was infected with a group B Salmonella, it was 15.7 times

more likely to die (1.9 to 129.25, 95%CI) than if it were infected with a group D.

Statistical summary of data is shown in Table 2-13. Additional odds ratios, relative to

infection with group D, were: C1-6 times more likely to die; C2-7 times more likely to

die; and E-6 times more likely to die. Although these ratios appeared large they were

not significant as the respective confidence intervals included 1.0.










Colpell
sYOW
IaXaOmPSOse
.Vallaassee
silot P V
_-,rtW


t^ Xiv
*30 a '


Java


p oN


45A'00 "asc


u-I-I-I-I-


I -p-p -p-pp
-!1P1 PP P -
~ n -
a -
I I ~I~ I I ~I~ I I I -


u- -


~ I -i-i-i-i-i-
I -! P


- ~ ~ ~ ~ ~ -


0 10 20 30 40 50 60
% Mortality Within Serotype


70 80 90 100


Figure 2-5. Mortality distribution, within serovar, of non-surviving
cases 1999-2002

30-

25- z

20-


E 1
z


equine salmonella


B C1 C2 D E F
Salmonella Group


Figure 2-6. Mortality by salmonella group in 88 cases with known outcomes


%o3T`IOod
W^tllo


I









Table 2-13. Effect of salmonella grou) on mortality in 88 cases of equine salmonellosis*
Odds Lower Upper
Group Died Lived % Dead Ratio 95%CI 95%CI
B 26 20 56.5 15.7 1.90 129.25
C1 2 3 40.0 6.0 0.42 85.25
C2 7 10 41.2 7.0 0.74 65.95
E 2 4 33.3 6.0 0.42 85.25
D 1 12 7.7 1.0 0.06 17.90
F 0 1 0.00 ---- ---- ----
* Pearson Chi-square value of 12.129 with 5 degrees of freedom, p = 0.033.

Within group B organisms, S. Typhimurium was associated with the highest

mortality rate (75.0% mortality), followed by 4,5,12:i-monophasic (66.7% mortality), S.

Saintpaul (60.0% mortality), S. Typhimurium var. Copenhagen (50.0% mortality), and S.

Java (45.0% mortality). Four serovars (Tallahassee, Reading, Meleagridis, and Agona)

may appear more virulent due to low numbers of cases.

Plasmid Profiling

Plasmid profiles were completed for 104 clinical salmonella isolates. Several

isolates in the main database were not analyzed due to equivocal identification, inability

to culture the provided isolate sample, or loss during storage. The majority of examined

isolates, 64.4% (67/104), contained at least one large (> 20-kb) plasmid. Several isolates

had additional smaller plasmids and some had more than one large plasmid. The

breakdown of plasmid carriage by serovar is listed in Table 2-14.

Table 2-14. Plasmid-positive salmonella isolates by serovar 1999-2002
Serovar Number Plasmid % Positive of % Positive
Positive / Number Total Isolates Within
Serovar Serovar
Java 23/25 22.12 92.00
Newport 5 / 13 4.81 38.46
Typhimurium 8 /8 7.69 100.00
Typhimurium var. 7/ 8
Copenhagen 6.73 87.50
Javiana 3 /5 2.88 60.00
Miami 1 /7 0.96 14.29









Table 2-14. Continued
Serovar Number Plasmid % Positive of % Positive
Positive / Number Total Isolates Within
Serovar Serovar
Muenchen 6 /6 5.77 100.00
Saintpaul 2 /6 1.92 33.33
Unidentified 2 /4 1.92 50.00
Anatum 2 /4 1.92 50.00
4,5,12:i-monophasic 3/3 2.88 100.00
Newington 0 /2 0.00 0.00
Hartford 2 /2 1.92 100.00
London 1 /2 0.96 50.00
Mbandaka 1 /2 0.96 50.00
Agona 0/ 1 0.00 0.00
Braenderup 0/ 1 0.00 0.00
Infantis 0 / 1 0.00 0.00
Meleagridis 0 / 1 0.00 0.00
Reading 0/ 1 0.00 0.00
Rubislaw 1 / 1 0.96 100.00
Tallahassee 0 / 1 0.00 0.00
TOTAL 67/104 64.42%

Selected examples of results from agarose gel electrophoresis plasmid profiles are

shown in Figures 2-7 through 2-13. Figure 2-7 shows the plasmid profiles of several

isolates extracted using the Birnboim and Doly method.135 All isolates were considered

plasmid-positive; however, lanes 3, 5, 6, and 7 show plasmids that were slightly larger

than the 100-kb plasmid of X3306. Three of those four were S. Typhimurium var.

Copenhagen isolates of bovine origin, and the fourth was an equine S. Newport. Figure

2-8 shows the plasmid profiles of Case 71 in lane 4, Case 66 in lane 5, and Case 77 in

lane 6. All of these isolates were determined to be plasmid-negative. These isolates were

identified as S. Miami, S. Newport, and S. Miami respectively. Figure 2-9 shows the

plasmid profile of five cases which all demonstrated large (> 16-kb) plasmids, but

varying in size relative to the 100-kb plasmid of the control strain. Only Case 89

(identified as S. Typhimurium var. Copenhagen) possessed the spy genes-all others









were negative. Interestingly, that plasmid appears very close in size to the 100-kb

virulence plasmid of the control S. Typhimurium strain, while the other four isolate

plasmids are large, but not necessarily equivalent in size. Figure 2-10 shows the plasmid

profile of Case 44 in lane 5 and Case 43 in lane 6. Both isolates appeared to possess a

single large (z 100-kb) and a single small plasmid. Both of these isolates were identified

as S. Typhimurium var. Copenhagen from horses having antibiotic administration prior to

admission, and neither of these isolates possessed the spy genes. Figure 2-11 shows the

plasmid profile of Case 63 in lane 5. Case 63 appeared to possess multiple plasmids

ranging in size from approximately 2-kb to > 100-kb but did not possess the spy genes.

This isolate was identified as S. London from a horse with prior antibiotic administration

(penicillin G). Interestingly, this isolate was not considered to be one of the more multi-

drug resistant strains to the 19 antimicrobials tested (resistance to more than 13/19 typical

in multi-drug resistant strains). The isolate from Case 63 was resistant to clindamycin,

doxycycline, erythromycin, oxacillin, penicillin, rifampin, and tetracycline. Figure 2-12

shows the plasmid profile of Case 83 in lane 5 and Case 40 in lane 6. Case 83 did not

appear to have any plasmid of any size visible on the gel, while Case 40 appeared to

possess both a small and large plasmid (between 4-kb and > 100-kb respectively). Case

83 was identified as S. Reading from a necropsy large intestine specimen with unknown

cause of death. Case 40 was identified as S. Typhimurium var. Copenhagen from a horse

with prior antibiotic administration (trimethoprim-sulfamethoxazole, metronidazole and

penicillin), and this was one of the few isolates of this serovar to not possess the spy

genes. Figure 2-13 shows the plasmid profile of Case 36 in lane 3. This isolate was

serotyped as the only group F Salmonella identified in horses (S. Rubislaw). Group F









salmonellae have not been reported as common equid isolates in the literature. This

isolate appeared to possess two large and two small plasmids but did not carry the spy

genes and was obtained from 3/5 fecal samples submitted.








100 kb




Figure 2-7. Plasmid profiles of 9 clinical salmonella isolates. Refer to Appendix C for
specific isolate information. Lanes: 1) Previously extracted 100-kb plasmid of
x3306, 2) Case 8, 3) Bovine isolate of S. Typhimurium var. Copenhagen, 4)
Case 11, 5) Bovine isolate of S. Typhimurium var. Copenhagen, 6) Case 12,
7) Bovine isolate of S. Typhimurium var. Copenhagen, 8) Case 6, 9) Case 10,
10) Case 7.

spy Gene Analysis

Of the 67 isolates found to be plasmid-positive 19.4% (12.5% of all isolates) were

also PCR-positive for the spy genes examined. All positive isolates generated expected

PCR product for all genes examined, they were an "all or none" result. Also, all isolates

that were spy positive were also plasmid-positive, and it was assumed that the genes were

located on a plasmid. Figures 2-14 through 2-18 show examples of the results obtained

for the spy gene analyses of clinical isolates. All isolates positive for the spy genes were

exclusively group B salmonellae. Within this group they were also limited to only three

serovars, Typhimurium, Typhimurium var. Copenhagen, and 4,5,12:i-monophasic, an

antigenically close relation to S. Typhimurium.















100 kb



16.21 kb









2.06 kb


Figure 2-8. Plasmid profiles of 4 clinical salmonella isolates. Refer to Appendix C for
specific isolate information. Lanes: 1) supercoiled marker DNA, 2) 100-kb
plasmid of 3306, 3) Case 78, 4) Case 71, 5) Case 66, 6) Case 77, 7) 100-kb
plasmid of 3306, 8) blank.

In Figure 2-14, there are multiple background bands in the spvA gel as well as one

band in the negative control lane of the spvC gel. Since none of these bands were of the

same intensity as the control, nor were they an appropriate size, they were considered

artifacts. No clinical isolate tested in Figure 2-14 was considered positive for the spy

genes. There appears to be a faint band of appropriate size in lane 9 of the spvC gel;

however, since the product band was of low intensity (compared to the positive control)

and there was no corresponding positive result in the spvA or spvR (not shown) gels, the

isolate was determined to be negative.
















100 kb



16.21 kb







2.06 kb


Figure 2-9. Plasmid profiles of 5 clinical salmonella isolates. Refer to Appendix C for
specific isolate information. Lanes: 1) supercoiled marker DNA, 2) 100-kb
plasmid of 3306, 3) Case 89, 4) Case 92, 5) Case 85, 6) Case 93, 7) Case 96,
8) 100-kb plasmid of X3306.







100 kb










Figure 2-10. Plasmid profiles of 4 clinical salmonella isolates. Refer to Appendix C for
specific isolate information. Lanes: 1) blank, 2) 100-kb plasmid of X3306, 3)
Aged monthnt) plasmid extract of 3306, 4) Case 46, 5) Case 44, 6) Case 43,
7) Case 53, 8) blank.













100 kb


16.21 kb








2.06 kb



Figure 2-11. Plasmid profiles of 4 clinical salmonella isolates. Refer to Appendix C for
specific isolate information. Lanes: 1) supercoiled marker DNA, 2) 100-kb
plasmid of X3306, 3) Aged monthnt) plasmid extract of X3306, 4) Case 41,
5) Case 63, 6) Case 64, 7) Case 65, 8) supercoiled marker DNA.























Figure 2-12. Plasmid profiles of 3 clinical salmonella isolates. Refer to Appendix C for
specific isolate information. Lanes: 1) supercoiled marker DNA, 2) 100-kb
plasmid of 3306, 3) 100-kb plasmid of k3306, 4) Case 82, 5) Case 83, 6)
Case 40, 7) 100-kb plasmid of X3306, 8) blank.














100 kb










Figure 2-13. Plasmid profiles of 7 clinical salmonella isolates. Refer to Appendix C for
specific isolate information. Lanes: 1) 100-kb plasmid of 3306, 2) Case 32,
3) Case 36, 4) Case 37, 5) Case 91, 6) Case 90, 7) Case 87, 8) Case 86.


1000 bp







1000 bp
1000 bp


Figure 2-14. PCR product results for spvA and spvC genes in 9 clinical salmonella
isolates, with positive and negative controls. Refer to Appendix C for specific
isolate information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) X3337
spy negative control, 3) x3306 spy positive control, 4) Case 86, 5) Case 87, 6)
Case 88, 7) Case 117, 8) Case 100, 9) Case 101, 10) Case 102, 11) Case 103,
12) Case 104, 13) 1-kb ladder DNA marker (Promega), 14) blank.


1 2 3 4 5 6 7 8 9 10 11 12 ^^131















1000 bp



Figure 2-15. PCR product for asd gene in 9 clinical salmonella isolates (same isolates
and orientation as Figure 2-14). Refer to Appendix C for specific isolate
information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) x3337 spy
negative control, 3) x3306 spy positive control, 4) Case 86, 5) Case 87, 6)
Case 88, 7) Case 117, 8) Case 100, 9) Case 101, 10) Case 102, 11) Case 103,
12) Case 104, 13) 1-kb ladder DNA marker (Promega), 14) blank.

Figure 2-15 shows the PCR products of the salmonella asd gene for the same

isolates (and same orientation in the gel) as Figure 2-14. Note that there is no product

visible for the isolate in lane #7 of either Figure 2-14 or 2-15. This isolate was positively

identified as S. Newport previously; however, the sample taken for template DNA in the

PCR mixture on this day was taken from a HE plate-where Salmonella spp. are

identified based on their ability to produce hydrogen sulfide. Normally these samples

were taken from isolates growing on LB plates. Apparently the hydrogen sulfide or some

other compound in the culture medium interfered with the PCR reaction, which would

have caused a false-negative result to be generated had this control not been run

simultaneously. An isolate was only evaluated for spy genes pending positive

determination of the asd gene, which essentially validated that the isolate was a

Salmonella spp.

Figure 2-16 demonstrates five clinical salmonella isolates that were positive for the

spvA gene. Serovars represented by these five isolates include S. Typhimurium and

4,5,12:i-monophasic.














1000 bp


Figure 2-16. PCR product results for spvA genes in 11 clinical salmonella isolates, with
positive and negative controls. Refer to Appendix C for specific isolate
information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) x3306 spy
positive control, 3) x3337 spy negative control, 4) lost isolate, 5) Case 8, 6)
Case 7, 7) Case 12, 8) Case 13, 9) Case 10, 10) Case 9, 11) Case 5, 12) Case
3, 13) Case 11, 14) Case 4.


1000 bp


wE


Figure 2-17. PCR product results for spvC genes in 11 clinical salmonella isolates, with
positive and negative controls. Refer to Appendix C for specific isolate
information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2) x3306 spy
positive control, 3) x3337 spy negative control, 4) Case 21, 5) Case 19, 6)
Case 16, 7) Case 22, 8) Case 27, 9) Case 26, 10) Case 25, 11) Case 24, 12)
Case 23, 13) Case 15, 14) Case 14.

Figure 2-17 demonstrates four clinical salmonella isolates that were positive for the

spvC gene. Serovars represented by these four isolates include S. Typhimurium var.


Copenhagen, S. Typhimurium, and 4,5,12:i-monophasic.


1] 2] 3 0 1 1 3 1


MbI














1000 bp

1000 bp


Figure 2-18. PCR product results for the asd gene in 11 clinical salmonella isolates
(same isolates and orientation as Figure 2-17). Refer to Appendix C for
specific isolate information. Lanes: 1) 1-kb ladder DNA marker (Promega), 2)
x3306 spy positive control, 3) x3337 spy negative control, 4) Case 21, 5) Case
19, 6) Case 16, 7) Case 22, 8) Case 27, 9) Case 26, 10) Case 25, 11) Case 24,
12) Case 23, 13) Case 15, 14) Case 14.

Figure 2-18 shows the PCR products of the salmonella asd gene for the same

isolates (and same orientation in the gel) as Figure 2-17. This is the typical appearance of

Salmonella spp. probed with the asdprimer set. All of these isolates could subsequently

be examined for the spy genes since they were positively determined to be salmonellae.

Outcome by Presence of the Virulence Plasmid or spy Genes

Short-term outcome was examined with respect to the presence or absence of the

virulence plasmid or spy genes. Results are detailed in Table 2-15.

Table 2-15. Summary outcome as determined by presence of the virulence plasmid and
spy genes in 98 equine salmonella cases
spv spv TOTAL
Lived 2 47 49
Died 9 28 37
Unknown Outcome 2 10 12
TOTAL 13 85

There was a significant correlation between the presence of spy genes and mortality

in this study population (p=0.001). Table 2-16 and Figures 2-19 and 2-20 depict the

differences in short-term outcome between cases with respect to the virulence plasmid

and spy genes. Nine out of 11 cases (81.8%) with spy-positive salmonella strains had a


1 2 3 4 5 6 7 8 9 10 11 12 13 14





owl









fatal outcome as opposed to 28/75 (37.3%) of the cases with spv-deficient strains. The

spy genes were restricted to group B salmonellae including serovars Typhimurium,

Typhimurium var. Copenhagen, and 4,5,12:i-monophasic. Horses infected with spy

gene-positive salmonella serovars were 12.3 times more likely to die than if they were

infected with a spy negative strain. Also, if the organism was detected outside the

gastrointestinal tract it was significantly more likely to be spy positive.

Table 2-16. Effect of spy gene presence on mortality in 86 cases of equine salmonellosis
where outcome was known*
Odds
Exposure Died Survived % Dead Ratio Lower 95%CI Upper 95%CI
spy Positive 9 2 81.8 12.30 2.59 58.41
spy Negative 28 47 37.3 1.00 0.52 1.91
Pearson Chi-square value of 14.070 with 1 degree of freedom, p=0.001.


spv +


SLived MDied


Figure 2-19. Outcome in equine salmonella cases, as influenced by presence of the spy
gene locus.










spy -




63%





37%




SLived Died



Figure 2-20. Outcome in equine salmonella cases, as influenced by absence of the spy
gene locus.

Effect of Clinical and Laboratory Parameters on Outcome

A large number of independent clinical and laboratory variables were investigated

with respect to predicting outcome in horses infected with Salmonella. Individual

variables with a p-value less than 0.2 were included in the original model. There was no

significant difference between survivors and non-survivors with respect to total serum

protein (TSP) at admission (p=0.197) or TSP at death or discharge (p=0.198), however

both factors significantly impacted outcome when investigated using forward stepwise

logistic regression. The median TSP at admission and discharge for the survivors was

6.25 (mean = 6.45, with 6.1 to 6.6 95%CI) and 6.15 (mean = 6.1, with 5.9 to 6.4 95%CI)

respectively. The median TSP at admission and death for the non-survivors was 6.8

(mean = 6.7, with 6.4 to 7.0 95%CI) and 5.5 (mean = 5.7, with 5.3 to 6.2 95%CI),









respectively. The total white blood cell and neutrophil counts at presentation did not

predict outcome.

Regardless of clinical presentation or syndrome, the presence or absence of

diarrhea was recorded where available for each case. True to the predominantly enteric

nature of this disease, 74 out of 85 (87.1%) cases exhibited diarrhea at some point during

hospitalization. Eleven cases (12.9%) did not develop diarrhea at any point during

hospitalization, and the information could not be determined for 21 cases. The presence

of absence of diarrhea did not predict survival.

Forward stepwise logistic regression analysis indicated that four categorical

predictor variables had a significant impact on outcome: spy gene status, TSP at

admission, TSP at death or discharge, and days of hospitalization were all related to

outcome. The average number of days spent hospitalized was 10.2, with a minimum of

one day (6 cases) and a high of 48 days. Of those cases that had a 3-day or less

hospitalization period, there was a 91.66% mortality rate-these cases likely were

admitted with severe disease, and were euthanatized due to expense, prognosis or

complications, with the Salmonella not being confirmed until after death. The significant

variables with test statistics are included in Table 2-17.

Table 2-17. Logistic regression model with variables predictive of outcome
Variable B SE Wald Statistic df Significance Exp (B)
spy genes -2.710 1.077 6.332 1 0.012 0.067
TSP at admission 1.050 0.398 6.967 1 0.008 2.859
TSP at death or discharge -1.099 0.356 9.516 1 0.002 0.333
Days of hospitalization -0.169 0.061 7.633 1 0.006 0.845









Relationship Between Proportion of Positive Fecal Salmonella Cultures and
Outcome

A Kruskal-Wallis test was used to investigate the percentage of submitted samples

that were positive with respect to outcome. There was a significant difference (p=0.042)

between those that lived and those that died. Horses that survived had a median of 60%

of their fecal cultures that were positive (mean = 62.71%, with 53.8% to 71.6% 95%CI),

as compared to horses that died spontaneously or that were euthanatized, where the

median percentage of positive cultures was 100% (mean = 77%, with 65.8% to 88.2%

95%CI).

Antibiotic Resistance Profiles

Antibiograms were obtained for 101 isolates. Complete MIC and resistance data

for all isolates can be found in Appendix E. A summary of the antibiotic susceptibilities

of 101 cases is displayed in Table 2-18 and the complete antibiogram of Case 78,

demonstrating the only isolate with reduced sensitivity to the fluoroquinolone

enrofloxacin, is shown in Table 2-19.

Table 2-18. Antibiotic susceptibilities for 101 equine salmonella isolates. The reported
% susceptible, % intermediate, and % resistant, are only for those isolates
with data for that antibiotic.
Antibiotic % Susceptible % Intermediate % Resistant
Clindamycin 1.1 0 98.5
Erythromycin 1.15 0 98.5
Penicillin 0 0 100.0
Oxacillin 1.1 0 98.9
Rifampin 1.1 0 98.9
Doxycycline 58.9 0 41.0
Tetracycline 61.0 0 39.0
Trimethoprim- 61.4 0 38.6
Sulfamethoxazole
Amoxicillin-Clavulanic Acid 67.4 0 32.6
Ampicillin 68.3 0 31.7
Ceftiofur 69.0 0 31.0
Cefazolin 70.0 2.0 28.0









Table 2-18. Continued
Antibiotic % Susceptible % Intermediate % Resistant
Ceftazidime 70.0 10.0 20.0
Gentamicin 75.2 9.9 14.9
Chloramphenicol 85.0 1.0 14.0
Amikacin 92.9 0 7.1
Enrofloxacin 99.0 1.0 0
Imipenem 100.0 0 0
Nitrofurantoin 100.0 0 0

Table 2-19. Antibiotic susceptibility report for Case 78, with intermediate resistance to
enrofloxacin
Drug MIC RS Drug MIC RS Drug MIC RS
Amikacin <=2 S Amox/Clav >16 R Ampicillin >16 R
Cefazolin >16 R Ceftazidime 32 R Ceftiofur >4 R
Chloramp 32 R Clindamy. >2 R Doxycyc. >4 R
Enroflox. 1 I Erythrom. >4 R Gentamicin >8 R
Imipenem <=1 S Nitrofur. <=32 S Oxacillin >4 R
Penicillin >16 R Rifampin >4 R Tetracyc. >16 R
TMP-Sulfa >4 R

Antibiotic Resistance Transformation

Electroporation transformation was successful in transferring cefazolin resistance

from the three salmonella isolates; but arcing due to the presence of buffer salts could

have potentially damaged the plasmid DNA to the extent of generating false-negative

results. Chemical transformations were performed with successful transference of

cefazolin, ceftiofur and ampicillin resistance from all three isolates. Figures 2-21 through

2-24 show the gel electrophoresis results for these analyses. Cefazolin-resistant E. coli

was shown to contain a new plasmid equivalent in size to the original cefazolin-resistant

transforming salmonella isolate (Figure 2-21). The untransformed E. coli did not possess

a plasmid, and is included for comparison.


















100 kb







Figure 2-21. Plasmid profiles of 3 clinical salmonella isolates and E. coli transformed
with plasmid DNA from those isolates. Refer to Appendix C for specific
isolate information and Appendix E for antimicrobial susceptibilities. Lanes:
1) 100-kb plasmid of 3306, 2) Untransformed E. coli DH5a, 3) E. coli DH5a
transformed with Case 97, grown in CEF, 4) E. coli DH5a transformed with
Case 92, grown in CEF, 5) E. coli DH5a transformed with Case 98, grown in
CEF, 6) Transforming plasmid DNA from Case 97, 7) Transforming plasmid
DNA from Case 92, 8) Transforming plasmid DNA from Case 98.


100 kb


Figure 2-22. Plasmid profiles of 2 clinical salmonella isolates and E. coli transformed
with plasmid DNA from those isolates. Refer to Appendix C for specific
isolate information and Appendix E for antimicrobial susceptibilities. Lanes:
1) Untransformed E. coli DH5a, 2) 100-kb plasmid of X3306, 3) E. coli DH5a
transformed with Case 98, grown in AMP, 4) E. coli DH5a transformed with
Case 98, grown in NAX, 5) E. coli DH5a transformed with Case 98, grown in
CEF, 6) E. coli DH5a transformed with Case 92, grown in AMP, 7) E. coli
DH5a transformed with Case 92, grown in CEF, 8) blank.



















100 kb









Figure 2-23. Plasmid profiles of 2 clinical salmonella isolates and E. coli transformed
with plasmid DNA from those isolates. Refer to Appendix C for specific
isolate information and Appendix E for antimicrobial susceptibilities. Lanes:
1) Untransformed E. coli DH5a, 2) 100-kb plasmid of 3306, 3) E. coli DH5a
transformed with Case 97, grown in AMP, 4) E. coli DH5a transformed with
Case 92, grown in NAX, 5) E. coli DH5a transformed with Case 97, grown in
NAX, 6) E. coli DH5a transformed with Case 97, grown in CEF, 7) blank, 8)
blank.

In Figure 2-23, lane 5 shows an isolate (Case 97) that transferred resistance to

NAX, CEF, and AMP. The plasmid transferring resistance to ceftiofur is larger than the

other two transforming plasmids (which appear to be the same size). Looking at the

plasmid profile of Case 97 in Figure 2-24, lane 5-there are three large plasmid bands

visible (2 smaller than 100-kb and 1 larger). This isolate most likely is carrying the AMP

and CEF resistance genes on the same plasmid and the NAX resistance gene on another

larger plasmid. On closer examination of Figure 2-21-red box, the second larger

plasmid is visible in Case 97 (lane 6), along with the other transforming plasmids of

homogenous size.


















100 kb






Figure 2-24. Plasmid profile of Case 97-lane 5. The red box delineates 3 large plasmid
bands that are visible in the upper part of the lane. This isolate transferred
ceftiofur, cefazolin, and ampicillin resistance via two different plasmids (the
lower two).

These experiments demonstrate that cefazolin, ampicillin, and ceftiofur resistance

in Cases 92, 97, and 98 were carried on plasmids which are smaller than the 100-kb

virulence plasmid and do not contain the spy genes.

Site of Salmonella Isolation

The majority of isolates in this study were obtained from fecal samples (80.0%).

Isolates from various segments of the gastrointestinal tract were examined separately and

as part of the group of gastrointestinal isolates. If isolates from all enteric sites are

considered together, the proportion of gastrointestinal isolates in the study rises to 93.3%.

Isolate distribution by site of infection is summarized in Table 2-20.

Table 2-20. Clinical salmonella isolates from 105 equine cases by location of cultured
specimen
Anatomic Site Number of % of Total Cases
Cases
Feces 84 80.0
Small Intestine (necropsy or surgery) 7 6.7
Large Intestine (necropsy) 4 3.8
Synovial (joint) Fluid 2 1.9
Lung (necropsy) 1 1.0









Table 2-20. Continued
Anatomic Site Number of % of Total Cases
Cases
Duodenum (necropsy) 1 1.0
Gastric Reflux 1 1.0
Abscess 1 1.0
Rectal Biopsy 1 1.0
Blood 1 1.0
Liver (necropsy) 1 1.0
Physis (necropsy) 1 1.0

The relationship between serovar and site of infection is summarized in Table 2-21.

Figure 2-25 illustrates the systemic isolates compared to the gastrointestinal isolates by

group.

Table 2-21. Systemic sites of salmonella infection in horses by serovar
Serovar Number and % of
Systemic Isolates
Hartford 1/1 100.0
Typhimurium 3/8 37.5
Muenchen 1/5 20.0
Typhimurium var. Copenhagen 1/7 14.3
Newport 0/13 0.0
Java 0/23 0.0
Javiana 0/7 0.0
Miami 0/7 0.0
Saintpaul 0/6 0.0
Anatum 0/4 0.0
4,5,12:i-monophasic 0/3 0.0
Newington 0/2 0.0
London 0/2 0.0
Mbandaka 0/2 0.0
Agona 0/1 0.0
Braenderup 0/1 0.0
Infantis 0/1 0.0
Meleagridis 0/1 0.0
Reading 0/1 0.0
Rubislaw 0/1 0.0
Tallahassee 0/1 0.0




















0% Systemic
*% Gastrointestinal


B C1 C2 D E F
Salmonella Group


Figure 2-25. Systemic equine salmonella isolates compared to gastrointestinal isolates by
group

Extra-intestinal isolates were 16.18 times more likely to carry the spy genes on a

virulence plasmid than enteric isolates (p=0.001). All salmonellae that contained the spy

genes also carried a large plasmid; no isolates were plasmid-negative and spy-positive.

This was significant with a 95% confidence interval of 3.54 to 74.06. Data are

summarized in Table 2-22.

Table 2-22. Relationship of the virulence plasmid and spy genes to isolate location in 98
cases of equine salmonellosis*


Site of Odds Lower Upper
Isolation spy Positive spy Negative % Positive Ratio 95%CI 95%CI
Extra-intestinal 4 3 57.1% 12.15 2.34 63.10
Intestinal 9 82 9.9% 1.00 0.38 2.65
* Pearson Chi-square value of 18.994 with 1 degree of freedom, p=0.001.






65


Multi-Serovar Salmonella Infections

Six horses had more than one serovar of Salmonella isolated from them during

hospitalization. All of the horses with multi-serovar salmonella infections developed or

were admitted with diarrhea, all survived, and all of the isolates were obtained from fecal

specimens. Table 2-23 details the groups, serovars, and relevant case information from

these 6 horses.










Table 2-23. Details of multi-serovar salmonella infections in six horses 1999-2002
Case Specimen Salmonella Clinical Ou e
Sex Age Serovar Outcome Diarrhea
ID Origin Group Syndrome
80 M Feces C2 muenchen Diarrhea,
M 7y Lived YES
36 Feces F rubislaw Fever

70 Feces B java Post-Op
69 M 6y Feces D javiana Colic Lived YES
Diarrhea

52 Feces C2 newport Diarrhea, Lid
F 3m Lived YES
58 Feces D miami Fever

75 Feces B java .
S M 3m Feces B java Diarrhea Lived YES
71 Feces D miami

73 F 6 Feces Cl hartford Diarrhea, Lid
F 6y Lived YES
39 Feces C2 newport Fever

81 Feces B Multiple serovars Colic -
M 3m (NVSL sample) Diarrhea, Lived YES
37 Feces C2 muenchen Chronic
Diarrhea









Discussion

Risk factors for the development of salmonella infection in horses have been well

described and were not investigated in the present study. The original aim of the

proposed study was to contrast Salmonella spp. shed from diseased animals in a hospital

setting with those recovered from a population of asymptomatic animals at pasture. This

comparison was to focus on isolate serovar, grouping, plasmid, and spy gene status.

Unfortunately, despite extensive and repeated culturing of animals at pasture we were

unable to isolate Salmonella from any asymptomatic animal. This finding was a surprise,

even in the face of a low prevalence (0.8%) reported in the recent NAHMS survey of

North American horses.95 With prevalence estimations ranging between 0% to 70% of

the horse population, depending on the risk group being sampled and the type of

diagnostic test used, it was expected to find at least one horse asymptomatically shedding

Salmonella in their feces. The sampling was done over several seasons to ensure that the

influence of temperature, weather patterns and time of year was minimized, and several

samples were taken from each animal over a period of time to maximize the possibility of

identifying periodic shedding episodes. There are several explanations for this negative

result: too few samples examined per horse, culture techniques too insensitive to identify

the low levels of bacteria shed by healthy horses, or more likely, that the true prevalence

was so low that insufficient numbers of animals were sampled. The sample collection

procedure and culture techniques were validated with samples from hospital patients

known to be shedding Salmonella. Enrichment (with sodium selenite + cystine, or

tetrathionate broth) and culture is currently the gold standard for diagnosing Salmonella

from fecal samples in horses, nevertheless the technique is not 100% sensitive.50









The study focus shifted towards a closer examination of the hospitalized

population, including descriptive data, risks factors within this population associated with

outcome, including organism group, serovar and spy gene status. With respect to the

descriptive data we were restricted by an inability to obtain accurate hospital population

demographics for the period in question. The breed distribution likely reflected the

regional and hospital population. No breed predilection has been reported for non-host

adapted salmonella infection in similar populations of hospitalized horses.44;136 The

mean age of affected animals in the present study was low in comparison to published

values, but likely reflects the referral horse population in North Central Florida. This

teaching hospital has a large caseload of young horses and foals due to close proximity to

breeding farms, and this factor likely contributed significantly to the low mean age.

Twenty-eight cases (28.57%) were in horses less than 6 months old, consistent with the

opportunistic nature of Salmonella in the very young, immunosuppressed, or geriatric

animals.107 Olsen et al. showed a very similar distribution regarding isolation rates by

age in humans, with over 48% of 441,863 isolates coming from individuals less than 19

years of age.137 The unbalanced distribution of the case population may also be a

reflection of compounded risk factors associated with age (e.g., younger horses may

undergo surgery more often than older ones, or younger horses are kept in larger groups

and may have an increased exposure to pathogens relative to solitary individuals).

As expected and reported in the literature, the largest number of cases in the present

study occurred during the warmer months of the year, 68% from April through

September. The seasonal predominance of salmonellosis in horses is typically highest

during the warmer summer months9599 and this seasonality was likely extended due to









the warm Florida climate. Thirty-year average minimum temperatures in Gainesville,

Florida remained above 60.3 F (15.7C) during the months of May through October

(Table 2-10).The wide spectrum of Salmonella recovered in this population is consistent

with previous studies in horses. There were significant associations between salmonella

grouping and spy gene presence and mortality. Animals infected with group B

Salmonella were nearly 16 times more likely to die than infected with the common Group

D bacteria. It was not surprising that the highest percentage of non-survivors occurred in

the group B organism S. Typhimurium and related serovar groups S. Typhimurium var.

Copenhagen and 4,5,12:i-monophasic. S. Typhimurium is a serious pathogen worldwide,

with higher mortality rates than many other serovars, even within the group B. S.

Typhimurium and S. Typhimurium var. Copenhagen were shown to cause significantly

higher fatality rates than all other serovars in two studies of hospitalized horses.99;138

This effect could likely be attributed to the presence of virulence plasmids, other

antimicrobial resistance factors, or undetermined virulence factors significant in horses.

Plasmid-bearing, spy gene positive organisms were restricted to group B Salmonella.

Eighty seven and a half percent of S. Typhimurium isolates were spy gene positive; 29%

of S. Typhimurium var. Copenhagen isolates were spy gene positive; and all 3 isolates of

4,5,12:i-monophasic contained spy virulence genes. It is important to point out however

that many group B Salmonella do not carry spy genes. This includes S. Java (none of 23

isolates), S. Saint Paul (0 of 6), S. Agona and S. Reading (0 of 1, respectively).

Extra-intestinal isolates were limited to groups B, Cl, and C2. The serovars

recovered from those isolates included S. Hartford, S. Typhimurium, S. Muenchen, and S.

Typhimurium var. Copenhagen. S. Typhimurium was the only serovar with more than









one systemic isolate, and more than 37% of all S. Typhimurium isolates were from

systemic sites. Systemic isolates had a significantly higher potential of carrying the spy

genes. This finding may indicate a similar role for the salmonella virulence plasmid and

these genes in horses, as demonstrated in calves,26 humans,22;139 and mice.21 Montenegro

et al. showed that virulence plasmids were detected in nearly 100% of extra-intestinal

isolates from human blood as well as cattle or swine internal organs.23

In summary, spy gene-containing isolates in horses are likely restricted to certain

group B salmonellae, are more likely to be recovered outside the intestinal tract, and are

more commonly associated with a negative outcome than non-spy gene-containing

isolates. The fact that all spy positive isolates were Group B salmonellae is also in

agreement with published reports. Eleven different serovars have been reported to carry

virulence plasmids (including S. Typhimurium); however, not all isolates within those

serovars will necessarily contain a virulence plasmid.140 In the present study, one S.

Typhimurium and six S. Typhimurium var. Copenhagen isolates did not possess the spy

genes.

Verification that the spy genes were located on the plasmid (and not integrated into

the chromosomal DNA) was not performed, but could be determined by transferring the

gel electrophoresis products to solid membranes, and then DNA-DNA hybridization to

the plasmid band (Southern blot). Chromosomal integration of the spy genes has only

been reported in subspecies II, IIIa, IV, and VII which typically infect cold-blooded

vertebrates.141;142 These subspecies do not infect warm-blooded vertebrates-only

subspecies I isolates have demonstrated mammalian pathogenicity.134 It was also shown

in a mouse-avirulent subspecies IV isolate that the chromosomally integrated spy genes









were not normally expressed and complementation with the entire virulence plasmid

from S. Typhimurium did not cause the isolate to become mouse virulent.

The ability to recover through bacterial culture, Salmonella spp. from fecal

samples, correlated with outcome. In general, animals with significant enteric disease

and higher mortality were more likely to return a larger proportion of positive cultures

than those with milder disease. This may be related to the number of organisms being

shed and/or to the immune status of the animal, with immunocompromised individuals

unable to significantly respond to the organism.

A recent retrospective study determined that low serum total protein concentrations

were associated with failure to survive in horses admitted for acute diarrhea.143 Using

limited clinical and laboratory data we performed a stepwise logistic regression analysis

in order to unmask factors that may be important in predicting outcome in horses with

salmonella infection. We also concluded that total plasma protein was an important

determinant of outcome, in addition to spy gene status, and duration of hospitalization.

Unfortunately none of these factors, with the exception of total plasma protein at

admission, could be used reliably to predict outcome in the clinical setting. The protein

concentration at admission was higher in the non-surviving group, likely reflecting more

severe hemoconcentration in those horses associated with acute fluid losses. Although

this finding is of clinical interest it is unlikely by itself to influence the decision to pursue

treatment. Overall, spy gene-containing isolates in horses are likely restricted to group B

organisms, more likely to be recovered outside the intestinal tract, and more commonly

associated with a negative outcome than spy gene-negative isolates.









Examination of in vitro salmonella sensitivity data is an important facet of clinical

practice. Not only does sensitivity data guide therapy but also is important in terms of

monitoring for drug resistance. The recognition of fluoroquinolone resistance in resident

strains of Salmonella is particularly important. A recent report detailed an outbreak and

general increase in the number of multidrug-resistant S. Newport being isolated from

humans.144 These isolates were resistant to amoxicillin/clavulanate, ampicillin, cefoxitin,

ceftiofur, cephalothin, chloramphenicol, streptomycin, sulfamethoxazole, and

tetracycline, and the resistance can be attributed to the presence of plasmids carrying a

blacmy gene, which produces AmpC-type enzymes that confer resistance and are termed

Newport MDR-AmpC strains. One isolate (case 14) was from a horse previously treated

with antibiotics (penicillin and trimethoprim-sulfamethoxazole) that developed diarrhea

attributed to S. Newport. This isolate had a resistance pattern strikingly similar to the

multidrug-resistant S. Newport described in the report (resistant to

amoxicillin/clavulanate, ampicillin, cefazolin, ceftazidime, ceftiofur, chloramphenicol,

clindamycin, doxycycline, erythromycin, oxacillin, penicillin, rifampin, tetracycline, and

trimethoprim-sulfamethoxazole) and also carried a large plasmid that did not contain the

spy genes. These strains of Salmonella are commonly associated with dairy farms, sick

cows, and unpasteurized milk or cheese.144

Only four serovars of Salmonella accounted for the 22 multidrug-resistant isolates

identified in this study (resistant to >8 drugs out of the 12 clinically relevant drugs

tested). S. Java accounted for ten, S. Typhimurium var. Copenhagen accounted for five,

S. Javiana accounted for two, and S. Newport accounted for one. Interestingly, S.

Typhimurium var. Copenhagen was also isolated from four hospitalized cows during the









same time period, and similar to what was reported for S. Newport, dairy cattle could be

reservoirs as well as modulators of resistance pressure in this serovar. It should be noted

that in vitro sensitivity data does not directly correlate with in vivo susceptibility due to

the normally intracellular location of this organism. This is particularly true for non-lipid

soluble antibiotics such as gentamicin. Based on this population of organisms and their

susceptibility data, clindamycin, erythromycin, penicillin, oxacillin, and rifampin cannot

be recommended for therapeutic treatment of salmonella infections, due to more than

95% of all isolates being resistant to each of those drugs. This is expected based on the

mode of action and gram-positive bacterial spectrum of these drugs. Drugs typically used

for peri-operative prophylaxis such as the first generation cephalosporin cefazolin or the

aminoglycoside gentamicin, had approximately 28% and 15% resistance respectively.

Amikacin, enrofloxacin, imipenem, and nitrofurantoin each had greater than 92% of all

isolates susceptible. A popular antibiotic selection for treatment of salmonella infections

in adult horses is the fluoroquinolone enrofloxacin. Enrofloxacin has an excellent gram-

negative spectrum, is accumulated within macrophages, and is effective against

intracellular organisms such as Salmonella. In this study, a single isolate demonstrated

intermediate resistance to this drug (Case 78); all others were susceptible. Case 78 was

from an 18y old Welsh Pony that presented for a gastric impaction. Post-operatively, this

horse developed diarrhea associated with a group B-S. Java, and was later euthanatized.

This patient was treated with penicillin, gentamicin, and metronidazole during

hospitalization and also was treated with enrofloxacin for the Salmonella. Surveillance

for fluoroquinolone resistance in Salmonella (especially veterinary isolates) is extremely

important, as these drugs are reserved for life-threatening infections in humans.145-147









Antibiotic resistances shown to be carried on plasmids include beta-lactamases and

extended-spectrum beta-lactamases,148 ampicillin,149 tetracycline,150 quinolones,151

trimethoprim and sulfonamides.149;152 The 3 cases demonstrating successful transference

of antibiotic resistances were all the same serovar (S. Java), a prevalent serovar in the

hospital during that time period. Prescott reported that multiple-antibiotic resistance is a

problem only in S. Typhimurium and not in other salmonella serovars.153 This was found

not to be the case, as most of the multiple-antibiotic resistant serovars in this study were

not S. Typhimurium, but S. Java. Preliminary data generated by this study supports the

contention that the majority of large plasmids associated with clinical isolates of

Salmonella from horses are likely antimicrobial resistance plasmids or R plasmids.