New Formulation and Evaluation of Multi-particulate Antibiotic … · 2019. 2. 9. · for Oral...

Formulation and Evaluation of Multi-particulate Antibiotic

Alternatives for Oral Delivery to Livestock Animals to Target

Gut Pathogens

By

Yin-Hing Ma

A thesis submitted in conformity with the requirements

for the degree of Doctor of Philosophy

Graduate Department of Pharmaceutical Sciences

University of Toronto

© Copyright by Yin-Hing Ma (2016)

ii

Formulation and Evaluation of Multi-particulate Antibiotic Alternatives

for Oral Delivery to Livestock Animals to Target Gut Pathogens

Yin-Hing Ma

Doctor of Philosophy

Graduate Department of Pharmaceutical Sciences

University of Toronto

2016

Abstract

This thesis explored strategies for enhanced oral delivery of antibiotic alternatives to the

livestock intestinal tract to target enteric pathogens. The first part dealt with formulating a liquid

essential oil (EO) into solidified granules and subsequently enteric coated for intestinal release.

Trans-cinnamaldehyde (CIN), a model EO compound with strong antibacterial activity was

formulated into granules using a three-component phase-diagram, yielding granules with high

loading and stability, and enteric coated by fluid-bed processing. Antibacterial activity of CIN

remained intact throughout manufacturing, after addition of 1% eugenol to prevent autoxidation

of CIN which lasted for at least 1 year under ambient storage. In animal trials, coated granules

delivered CIN in higher concentrations than free oil to the lower intestinal regions where

pathogens colonize.

The second part explored bacteriophage microencapsulation for oral, bio-control of intestinal

Salmonella in chickens. Preliminary studies revealed the phage intestinal distribution and fecal

excretion pattern following a single oral dose, and optimized the Ca-alginate-whey protein

formulation of microbeads for quicker phage release. Phage lytic activity studies against a target

iii

pathogen showed single phages were insufficient at maintaining reduction of STT104NalR, rather

phage cocktails were required to maintain reduction in vitro and in animal digesta, which

achieved reductions of several log10CFU/g digesta. Two trials were carried out, testing the

effectiveness of pathogen reduction by phage cocktails administered to chicks experimentally

infected with STDT104NalR. Trial #1 showed that cocktail 1 did not significantly affect the

colonization levels even though in vitro activity was good. Trial 2 re-tested CT1 along with an

expanded cocktail 2, with larger animal groups. CT1 yielded a similar lack of efficacy, while

CT2 achieved significant reductions of STDT104NalR in ileum, ceca, and colon after 7 days,

twice-daily administration with feed. This study suggests that phage cocktails

microencapsulated for continuous oral administration with feed could reduce intestinal pathogens.

iv

Acknowledgements

I would like to acknowledge my supervisors Dr. Shirley X. Y. Wu and Dr. Qi Wang for the

opportunity to work on this project and their enduring support and funding. In addition, Dr.

Parviz Sabour, Dr. Jim Chambers, Dr. Joshua Gong and Dr. Hai Yu at the Guelph Food Research

Centre, Agriculture-Agri-Food Canada, and the kind assistance of Marta Hernandez, Shailja Baxi,

my Guelph lab mates and co-workers for their much appreciated assistance in performing

experiments: Golam Islam, Dr. Xianhua Yin, Ling Yuan, Yonggang Zhang, Ben Huang, and Dr.

Zhenxing Tang and my UofT lab mates and co-workers Jamie Lugtu-Pe, Dr. Alireza Shalviri, Dr.

Michael Chu. Also I’d like to thank my committee members Dr. Ping Lee, Dr. Ian Crandall, Dr.

Heiko Heerklotz for their helpful criticism and suggestions over the course of my research.

v

Publications from this thesis

Peer reviewed Journal Papers:

1. Ma, Y., Wang, Q., Gong, J., Wu, X.Y. Formulation of granules for site-specific

delivery of an antimicrobial essential oil to the animal intestinal tract. Journal of

Pharmaceutical Sciences. (2016) 105(3):pp.1124-1133.

2. Ma, Y., Wang, Q., Gong, J., Wu, X.Y. In vivo performance of coated essential oil

granules for targeted delivery to the animal GI tract. (in preparation)

3. Ma, Y., Islam, G., Wu, Y., Sabour, P., Chambers, J., Wang, Q., Wu, X.Y., Griffiths, M.

Temporal Distribution of Encapsulated Bacteriophages during passage through the

chick gastro-intestinal tract. Poultry Science. Published online September 2016 doi:

10.3382/ps/pew260

4. Ma, Y., Yin, X., Islam, G., Wang, Q., Sabour, P., Chambers, J., Wu, X.Y.

Encapsulation of bacteriophages for enhanced delivery to the chicken intestine in

vivo performance. (in preparation for submission to Antimicrobial Agents and

Chemotherapy)

Conference Abstracts and Poster Presentations:

1) Ma, Y., Wang, Q., Gong, J., Wu, X.Y. Formulation and in vitro/in vivo evaluation of

coated granules for site-specific oral delivery of an antimicrobial essential oil to livestock

GI tract. AAPS Annual Meeting, Orlando, Fl. Oct. 2015.

2) Ma, Y., Islam, G., Wang, Q, Sabour, P.M., Chambers, J.R., Griffiths, M, Wu, X.Y.

Encapsulation of bacteriophage for enhanced delivery to the chicken intestine: in vitro vs

in vivo. CIFST 2012, Niagara Falls, ON.

3) Ma, Y., Wang, Q., Tang, Z.X., Sabour, P.M., Chambers, J.R., and Griffiths, M. (2012).

"Microencapsulation of bacteriophage for enhanced delivery to chicken intestine.” 10th

Annual OMAFRA Food Safety Research Forum, Guelph, ON, Canada, May 2, 2012.

4) Tang, Z.X., Ma, Y., Wang, Q., Sabour, P.M., and Wu, X. Y. (2012). "Microencapsulation

of bacteriophages for oral delivery to chicken intestines.", 11th International

vi

Hydrocolloids Conference, Stewart Center, Purdue University, West Lafayette, IN, USA,

May 14-18, 2012.

5) Wang, Q., Ma, Y., Zhang, Y., Gong, J., Sabour, P.M., Chambers, J.R., and de Lange,

C.F.M. (2012). "Application of microencapsulation in the development of alternatives for

antibiotics in food animal production.", 20th International Conference on

Bioencapsulation, Orillia, ON, Canada, September 21-24, 2012

6) Ma, Y., Wang, Q., Sabour, P.M., Chambers, J.R., and Wu, X. Y. (2012). "Encapsulation

of bacteriophage for enhanced delivery to chicken intestine in vitro vs in vivo", 50th

National Conference of the Canadian Institute of Food Science and Technology (CIFST):

Innovation meets Commercialization, Niagara Falls, ON, Canada, May 27-29, 2012.

7) Ma, Y., Wang, Q., Gong, J., and Wu, X.Y. (2011). "Towards replacement of antibiotics

in feed: encapsulation of cinnamaldehyde oil for delivery to the gastrointestinal tract of

pigs.", Canadian Society of Microbiologists (CSM) 61st Annual Conference / Conférence

annuelle de la Société Canadienne des Microbiologistes (SCM), Memorial University of

Newfoundland, St. John's, NL, Canada, June 20-23, 2011, AE25.

8) Tang, Z.X., Ma, Y., Huang, X.Q., Baxi, S., Wang, Q., and Sabour, P.M. (2010). "In-vitro

survival of dried encapsulated Staphylococcus aureus phage K at different temperature",

2010 Canadian Institute of Food Science & Technology (CIFST) and AAFC Conference,

Fairmont Winnipeg Hotel, Winnipeg, MB, Canada, May 30-June 1, 2010.

9) Ma, Y., Wang, Q., Gong, J., Wu, X.Y. Site-specific oral delivery of essential oil granules

to the swine GI tract as antibiotic alternatives, AAPS Annual Meeting, 2010.

vii

Table of Contents

Abstract .......................................................................................................................................... ii

Acknowledgements ...................................................................................................................... iv

Publications from this thesis ........................................................................................................ v

List of Tables ............................................................................................................................... xii

List of Figures ............................................................................................................................. xiv

Abbreviations commonly used in this thesis ............................................................................ xvi

Chapter 1. Introduction and literature review ........................................................................... 1

1.1 The growing problem of antibiotic resistance .................................................................. 1

1.2 Causes of antibiotic resistance: the widespread overuse and misuse of antibiotics ......... 2

1.2.1 Antibiotic growth promoters ..................................................................................... 2

1.3 Evaluating alternatives to antibiotic use ........................................................................... 7

1.3.1 Essential oils as potential antibiotic alternatives ...................................................... 8

1.3.2 Potential resistance mechanisms to EO compounds ............................................... 11

1.3.3 In vitro activity of EO’s review .............................................................................. 12

1.3.4 In vivo studies ......................................................................................................... 15

1.4 The lower GIT of animals: a reservoir for pathogen colonization ................................. 17

1.4.1 Challenges in the oral delivery of EOs and compounds ......................................... 19

1.5 Delivery approaches applicable to EO compounds ........................................................ 19

1.5.1 Review of suitable EO delivery methods ................................................................ 19

1.6 Bacteriophage as antibiotic alternatives ......................................................................... 23

1.6.1 Mechanism of action of bacteriophage ................................................................... 24

1.6.2 Phage for bacterial reduction/Phage therapy .......................................................... 28

1.6.3 Considerations and strategies for encapsulated phage for oral delivery ................. 32

1.7 Microencapsulation of phages using Ca-alginate matrix based methods ...................... 33

1.7.1 Ca-alginate gelation process ................................................................................... 33

1.8 Hypothesis and specific objectives of thesis .................................................................. 36

Chapter 2. Formulation of granules for site-specific delivery of an antimicrobial essential

oil to the animal intestinal tract ................................................................................................. 40

Abstract ........................................................................................................................................ 41

2.1 Introduction .................................................................................................................... 42

2.2 Materials and Methods ................................................................................................... 44

viii

2.2.1 Determination of oil adsorbing capacity of powders .............................................. 45

2.2.2 Formulation selection by phase diagram ................................................................ 46

2.2.3 Preparation of core granules ................................................................................... 46

2.2.4 Enteric coating of core granules ............................................................................. 48

2.2.5 Assay of cinnamaldehyde and cinnamic acid content in granules .......................... 48

2.2.6 In vitro release of CIN from granules ..................................................................... 49

2.2.7 Antimicrobial activity assay with pure culture in liquid growth media .................. 49

2.3 Results ............................................................................................................................ 50

2.3.1 Formulation and properties of CIN core granules .................................................. 50

2.3.2 Selection of antioxidant to stabilize CIN in the core granules against atmospheric

oxidation (autoxidation) ........................................................................................................ 57

2.3.3 Properties of coated granules .................................................................................. 59

2.3.4 In vitro release of CIN from core granules and coated granules ............................ 61

2.3.5 Antimicrobial activity of CIN against E. coli K88: MIC and MBC determination 62

2.4 Discussion ...................................................................................................................... 64

2.4.1 Importance of formulation on the properties of granules ....................................... 64

2.4.2 Antibacterial activity of and storage stability of CIN granules .............................. 67

2.5 Conclusions .................................................................................................................... 68

2.6 Acknowledgements ........................................................................................................ 69

Chapter 3. In vivo performance of essential oil granules for targeted delivery to the animal

GI tract ......................................................................................................................................... 70

Abstract ........................................................................................................................................ 71

3.1 Introduction .................................................................................................................... 72

3.2 Materials and Methods ................................................................................................... 73

3.2.1 Preparation of core granules and enteric coating .................................................... 73

3.2.2 Oral administration to 2 week old chicks ............................................................... 73

3.2.3 Pig trial .................................................................................................................... 75

3.2.4 CIN determination in digesta .................................................................................. 76

3.2.5 Gas chromatography ............................................................................................... 76

3.2.6 Antimicrobial activity in animal digesta ................................................................. 77

3.2.7 In vitro release of CIN from granules ..................................................................... 78

3.2.8 Statistics .................................................................................................................. 78

ix

3.3 Results ............................................................................................................................ 78

3.3.1 Chicken Study ......................................................................................................... 78

3.3.2 Pig Study ................................................................................................................. 82

3.3.3 Antimicrobial activity of CIN against E. coli K88 in animal digesta ..................... 84

3.4 Discussion ...................................................................................................................... 85

3.4.1 Chicken trial ............................................................................................................ 85

3.4.2 Pig trial .................................................................................................................... 86

3.4.3 Antimicrobial activity ............................................................................................. 87

3.4.4 Release in vitro ....................................................................................................... 89

3.5 Conclusions .................................................................................................................... 89

3.6 Acknowledgements ........................................................................................................ 90

Chapter 4. Temporal distribution of encapsulated bacteriophages during passage through

the chick gastro-intestinal tract ................................................................................................. 91

Abstract ........................................................................................................................................ 92

4.1 Introduction .................................................................................................................... 93

4.2 Materials and methods ................................................................................................... 95

4.2.1 Chicks used ............................................................................................................. 95

4.2.2 Bacterial strain used for phage enumeration ........................................................... 96

4.2.3 Bacteriophages used ................................................................................................ 96

4.2.4 Bacteriophage encapsulation into Ca2+-alginate-whey protein gel beads ............... 96

4.2.5 Bacteriophage enumeration assay ........................................................................... 98

4.2.6 In vitro release of encapsulated phage from Ca2+-alginate-whey protein gel beads98

4.2.7 In vitro phage incubation with chick feed under simulated gastric and intestinal

conditions ............................................................................................................................... 99

4.2.8 In vivo phage distribution following oral administration to chicks ........................ 99

4.2.9 Fecal phage excretion profile following oral phage administration to chicks ...... 100

4.2.10 In vitro phage incubation with chick fecal material and buffer solutions ............. 101

4.2.11 Microscopic observation ....................................................................................... 101

4.2.12 Statistical Analysis ................................................................................................ 101

4.3 Results .......................................................................................................................... 102

4.3.1 In vitro incubation of phage FO1 with chick feed and digestive enzymes. .......... 103

x

4.3.2 Phage distribution within the chicken digestive tract following single oral dose in

absence of host bacteria. ...................................................................................................... 104

4.3.3 Fecal excretion of an oral phage dose over time. .................................................. 108

4.3.4 In vitro incubation of FO1 in chick feces, MBS and SM buffer. .......................... 110

4.4 Discussion .................................................................................................................... 111

4.5 Acknowledgements ...................................................................................................... 116

Chapter 5. Encapsulation of bacteriophages for enhanced delivery to the chicken intestine

in vivo performance ................................................................................................................... 117

Abstract ...................................................................................................................................... 118

5.1 Introduction .................................................................................................................. 119

5.2 Materials and methods ................................................................................................. 120

5.2.1 Bacterial strains used ............................................................................................ 120

5.2.2 Enumeration from samples containing STDT104 ................................................ 121

5.2.3 Bacteriophages used .............................................................................................. 122

5.2.4 Phage propagation and purification ...................................................................... 123

5.2.5 Phage enumeration ................................................................................................ 124

5.2.6 Bacteriophage in vitro activity screening assay .................................................... 124

5.2.7 Bacteriophage encapsulation into alginate-whey protein gel beads ..................... 125

5.2.8 In vitro phage-bacteria incubations in animal digesta. ......................................... 126

5.2.9 In vivo Trials of oral phage therapy effectiveness in broiler chicks ..................... 126

5.2.10 Animals used for experiment and care .................................................................. 126

5.2.11 Fecal phage excretion after a single oral dose ...................................................... 128

5.2.12 Data presentation and statistical analysis. ............................................................. 129

5.3 Results .......................................................................................................................... 129

5.3.1 Animal Trial #1. .................................................................................................... 136

5.3.2 Animal Trial #2 ..................................................................................................... 141

5.4 Discussion .................................................................................................................... 143

5.4.1 Fecal excretion trial ............................................................................................... 145

5.4.2 Trial #1 discussion ................................................................................................ 145

5.4.3 Trial #2 discussion ................................................................................................ 147

5.4.4 Effect of treatment length of time and multiple dosing, timing of administration 148

5.4.5 Phage concentration in vivo MOI ......................................................................... 149

xi

5.4.6 Future outlook of phage therapy ........................................................................... 150

5.5 Acknowledgements ...................................................................................................... 152

Chapter 6. Conclusions and Future Research Discussion ..................................................... 154

6.1 Original Contributions and Conclusions from this Thesis ........................................... 154

6.2 Discussion of Limitations of Thesis Work and Future Research Directions ............... 156

References .................................................................................................................................. 160

xii

List of Tables

Table 1.1 Possible antibiotic alternatives listed from current literature sources ........................... 8

Table 1.2. Various essential oils with antimicrobial activity and their origin ................................ 9

Table 1.3. Examples of antibiotic growth promoting (AGP) concentrations used and effective EO

concentrations ............................................................................................................................... 13

Table 1.4. Properties of trans-cinnamaldehyde the major active component found in cinnamon

bark oil .......................................................................................................................................... 14

Table 1.5. Alimentary tract characteristics of swine and poultry ................................................. 18

Table 1.6. Chemical and physical properties of commercial enteric (Eudragit®) polymers ......... 22

Table 1.7. The advantages and disadvantages to using bacteriophages as biocontrol agents ...... 27

Table 2.1. The oil adsorbing capacity of various powders. Values shown as AVE ± SD (n ≥ 3) 51

Table 2.2. Granule properties and process characteristics ............................................................ 56

Table 3.1.Average CIN concentrations in the digesta of chickens following oral gavage of two

froms of CIN ................................................................................................................................. 80

Table 3.2.Peak concentrations observed from individual chicks within each group .................... 80

Table 3.3. CIN concentration in pig digesta after 5 hours post-feeding of feed containing free oil

or coated granules at 1500 µg/g (PPM) of feed ............................................................................ 83

Table 3.4. Antibacterial activity of CIN against E.coli k88 in different digesta types (pig) and

formulation from 5 hour incubation, data presented as mean log10 CFU/mL .............................. 85

Table 4.1. Compositions of the different calcium alginate-whey protein formulations used for

phage encapsulation ...................................................................................................................... 98

Table 4.2. Incubation of phage FO1 with and without chick feed and digestive enzymes in

simulated gastro-intestinal conditions. Values are presented in log10 (PFU/mL) (AVE ± SE, n=3).

FP = free phage, EP = encapsulated phage. Means with significant difference (at P < 0.05 level)

as determined by a two-way ANOVA and Tukey’s correction, are indicated by different letters

(a,b,c) across row and by different subscript numbers (1,2) within column. .................................. 104

Table 4.3. Incubation of encapsulated and free phage with buffer in chick feces and its effect on

phage survival over time. Data shown are in log10 PFU, (AVE ± SE, n=3) .............................. 111

Table 5.1. Effect of phage cocktail CT1 (phage FO1+F24) at MOI 10 on Salmonella growth in

raw chicken digesta isolated from different region of GI tract over time. Data presented as

AVE+ SE of log CFU/g digesta. ................................................................................................. 134

Table 5.2. Treatment groups and sample sizes used in trial #1 .................................................. 137

xiii

Table 5.3. Salmonella Typhimurium DT104 NalR recovery after 2 and 7 days continuous phage

treatment in trial #1 (AVE ± SE, N=5-6) log10 (CFU/g of tissue or digesta) ............................. 138

Table 5.4. Bacteriophage recovery after phage treatment in trial #1 (AVE ± SE, N=5-6) log10

(PFU/g of tissue or digesta) ........................................................................................................ 139

Table 5.5. Treatment groups and sample sizes used in trial #2 ................................................. 141

Table 5.6. Trial #2, Salmonella Typhimurium DT104 NalR counts recovered from digesta and

tissue samples after 2 and 7 days continuous phage treatment (AVE ± SE, log10 (CFU/g of tissue

or digesta) .................................................................................................................................... 142

Table 5.7. Trial #2, Bacteriophage counts recovered from treatment groups treated with two

phage cocktails (AVE ± SE, N=11-13) log10(PFU/g of tissue or digesta) .................................. 142

xiv

List of Figures

Figure 1.1 Typical life cycle of lytic phage .................................................................................. 26

Figure 1.2. Egg-box like structure of Ca-alginate gelation process forming network .................. 35

Figure 2.1. Schematic diagram for granule production process ................................................... 48

Figure 2.2. Pictures of Neusilin US2 at various CIN oil loading: top 0%, middle 62%, and

bottom 78% ................................................................................................................................... 52

Figure 2.3. Phase diagram for a 3-component mixture at molten (fatty acid) and cooled

temperatures. Dot indicates optimized formulation. Outline indicates range of formulation

components that resulted in acceptable granules. ......................................................................... 54

Figure 2.4. Effect of formulation composition changes on the size distribution of granules while

keeping batch size, stirring rate constant. ..................................................................................... 57

Figure 2.5. (a) Stability of CIN in uncoated granules with and without antioxidant when stored at

23 and 4 °C. (b) Stability of coated granules with antioxidant (1% EUG) stored at 23 and 4 °C

for up to 1 year .............................................................................................................................. 59

Figure 2.6. (a) Lauric acid granules uncoated (left) and after coating (right) and (b) palmitic acid

granules with 42% w/w CIN, coated with Eudragit L100 and sub-coat with Kollicoat® IR ....... 60

Figure 2.7. (a) Release of CIN in SIF (pH 6.8) from core granules formulated with different fatty

acid types and release of CIN from coated granules under two-stage dissolution. Values represent

AVE±SD ....................................................................................................................................... 61

Figure 2.8. (a) Inhibitory activity of CIN vs CA towards E. coli K88 grown in TSB culture based

on change in OD600 over 6 hrs @ 37° C with 200 rpm shaking. (b) Antibacterial activity of CIN

oil and granules against E. coli K88 in TSB medium. Data points represent AVE±SD .............. 63

Figure 3.1 . Schematic of chicken trial ......................................................................................... 75

Figure 4.1. Release profiles of the six encapsulation formulations compared to the 1.5%

Alginate-3% whey protein formulation, in simulated intestinal fluid without digestive enzymes.

F1-F6 refer to the different encapsulation formulations found in Table 4.1. (AVE±STDEV, n=3)

..................................................................................................................................................... 102

Figure 4.2. The in vivo distribution of free/released FO1 after single oral doses of encapsulated

phage (EP) and free phage (FP) in SM buffer after (a) 1 h, (b) 2 h, and (c) 4 h. (AVE ± ST DEV,

n =4). *indicates significant difference between means at p < 0.05 level after two-way ANOVA

and Tukey’s correction. Chicks were given a dose of (FP) 3 1010 PFU/chick and (EP) 4.6-9

109 PFU/chick. ............................................................................................................................ 107

xv

Figure 4.3. Images of chicken GI tract contents after oral gavage of encapsulated phages. In crop

contents (a) some blue beads (encapsulated phage) are observed after 2 hours and (b) beads are

still visible after 4 hours. Numerous blue beads are seen among ingested feed in gizzard contents

after (c) 2 hours and (d) 4 hours following gavage. .................................................................... 108

Figure 4.4. Fecal excretion of FO1 over time following gavage of a single oral dose of 200 mg

beads delivering 6x109 PFU/chick (AVE ± SEM, n=6). ............................................................ 109

Figure 4.5. Fecal excretion levels of phage FO1 given orally in liquid suspension containing

~6x109 PFU/chick (AVE ± SEM, n=6). ..................................................................................... 110

Figure 5.1. Effect of single bacteriophages: FO1, V6, V16, F3, F4, F24, F32 on STDT104 NalR

growth at various initial MOI in TSB growth medium and initial cell density of 104 CFU/mL 132

Figure 5.2. Optical density (@ 600 nm) readings showing in vitro lytic activity of phages CT1,

YJ104 and CT2 at MOI: 1, compared with a control bacterial growth curve (initial 106 CFU/mL

of STDT104). Shown are AVE+SE values from experiments repeated twice with n≥6 per

experiment. .................................................................................................................................. 133

Figure 5.3. Optical density (OD600) reading of CT1, YJ104 and CT2 at MOI of 10 with an initial

106 CFU/mL of STDT104. Note difference in Y-axis scale compared with Figure 5.2. Shown

are AVE+SE readings of experiments replicates ≥6 per experiment. ........................................ 133

Figure 5.4. Excretion profile of phage in feces after a single oral dose of free phage cocktail (~5

x 109 PFU/chick of CT1 in 0.2 mL SM buffer) in 4 day and 9 day-old broiler chicks (AVE±SE,

N=3) ............................................................................................................................................ 135

Figure 5.5. Excretion profile of phage in feces after a single oral dose of encapsulated phage

cocktail (~109 PFU/chick of CT1 in 0.2 g wet beads) in 4 and 9 day-old broiler chicks (AVE±SE,

N=3) ............................................................................................................................................ 136

xvi

Abbreviations commonly used in this thesis

AGP – antibiotic growth promoter(s)

AR – antimicrobial/antibiotic resistant

ACSSuT – ampicillin, chloramphenicol, streptomycin, sulfonamide, tetracycline

ATCC – American type culture collection #

ATP – adenosine tri-phosphate

AVE – average

AX – antioxidant

BHT – butylated hydroxy-toluene

BGS – brilliant green sulpha

CA – cinnamic acid

CAl – cetyl alcohol

CIN– cinnamaldehyde

CFU – colony forming units

DNA – deoxyribonucleic acid

ECK88 – E. coli K88 , F4 fimbriae producing

ESBL – extended spectrum beta-lactamase

EO – essential oil

ETEC – enterotoxigenic E. coli.

EUG – eugenol

EU – European Union

FA – fatty acid

FO1 – bacteriophage Felix O1

GC – gas chromatography

GC-FID – gas chromatography flame ionization detector

GC-MS – gas chromatography – mass spectrometry

GIT– gastro-intestinal tract

GI – gastro-intestinal

i.v. – intravenous injection

i.m. – intramuscular injection

i.p. – intraperitoneal injection

xvii

LA – lauric acid (dodecanoic acid)

LPS – lipopolysaccharide

MA– myristic acid

MDR – multi-drug resistant

MIC – minimum inhibitory concentration

MBC –minimum bactericidal concentration

MRSA – methicillin resistant Staphylococcus aureus

Nal – nalidixic acid sodium salt

NDM-1 – New Delhi metalloproteinase-1

NE – necrotic enteritis

OD600 – optical density at 600 nm

PA – palmitic acid

PFU – plaque forming units

PMF – proton motive force

RNA – ribonucleic acid

RT – room temperature

SA – stearic acid

SE – standard error

STDT104 NalR – Salmonella Typhimurium definitive type-104, nalidixic acid resistant

Tet – tetracycline

TSA – tryptic soy agar

TSB – tryptic soy broth

US2 – Neusilin® US2

UV-vis – ultraviolet- visible light spectroscopy

VRE – vancomycin resistant Enterococci

WHO – world health organization

1

Chapter 1. Introduction and literature review

1.1 The growing problem of antibiotic resistance

Ever since the discovery of penicillin in 1929 and the ensuing mass production and widespread

use of subsequent classes, antibiotics appeared as the miracle drugs that kept deadly bacterial

infections in check for many decades. Over time bacterial diseases returned stronger, due to their

adaptability and acquisition of various mechanisms to circumvent the effects of antibiotics

rendering them (and earlier generations) ineffective and newer antibiotics needed to be isolated

and/or developed. Over many decades of exposure, many pathogenic bacteria have acquired

resistance to all earlier generations of the antibiotics such as sulfonamides, tetracycline,

penicillin, and streptomycin, suggesting that bacterial antimicrobial resistance is inevitable.

Many bacteria exhibiting a multidrug resistance (MDR) phenotype are now simultaneously

resistant to ≥3 different classes of antibiotics, and ongoing antibiotic resistance surveillance

studies reported that the incidence of MDR isolates (or “superbugs”) from hospitals are on the

rise across Canada, the US and globally [1] with some considering the age of antibiotics already

in decline due to the current widespread proliferation of antibiotic resistant bacteria and the lack

of new antibiotics being developed or available for treatment [2, 3].

Bacterial strains commonly causing food-borne illness come from the species Salmonella

enterica, Escherichia coli , Campylobacter, and Listeria [4-6]. Such infections can affect people

of all ages and can lead to vomiting, nausea, fever, severe diarrhea, and sometimes death when

treatment fails [7]. E. coli serotype K88 (F4) a type of enterotoxigenic E. coli (ETEC), was

found to be responsible for human and animal outbreaks, which have been isolated from farms

across Ontario, and showed resistance to spectinomycin, tetracycline, ampicillin, neomycin, and

carbenicillin [8]. Piglets infected with E. coli K88 experience symptoms ranging from severe

2

diarrhea to sudden death, with mortality rates of 50% or higher [9, 10]. Salmonella enterica

serotype Typhimuruim DT104 is another pathogen readily transmitted between turkeys, cattle,

swine, poultry and humans [11] that can cause diarrhea and even death in severe infections [12].

Since the early 80’s this Salmonella strain has already been reported to be penta-resistant and

with growing resistance to fluoroquinolones. These antibiotic resistant pathogens are being

consistently isolated among the bacterial flora of food animals and humans of which their

digestive tracts are now major reservoirs of pathogens [13, 14].

1.2 Causes of antibiotic resistance: the widespread overuse and misuse of

antibiotics

The underlying cause of antibiotic resistance is the repeated exposure of bacteria to sub-lethal

concentrations which allows them to acquire resistance mechanisms over time and to proliferate,

and by looking back at the historical practices of antibiotic use over many decades, two main

culprits were identified as being [15, 16]: 1) human misuse/overuse and 2) misuse overuse of

antibiotics in animal farming for disease prevention (e.g. prophylaxis against weaning diarrhea,

respiratory/enteric disease, coccidiosis, etc.) and growth promotion, or so-called “antibiotic

growth promotion” (AGP). The sub-therapeutic use of antibiotics in farm animals began from as

early on as the post WWII era [2] when observations were made that intensively farmed animals

receiving supplemented antibiotics experienced better weight gain, feed conversion efficiency,

and were healthier overall, thus produced higher yield and profits for the farmers. Other types of

farming like aquaculture and agriculture continue to use antibiotics.

1.2.1 Antibiotic growth promoters

AGPs have been used predominately in swine, cattle and poultry farming. The mechanisms of

action of AGP are believed to involve the following positive effects [10, 16-18]:

3

- reduction of harmful bacterial flora in the gut results in lower fermentation and diversion

of energy from feed, fewer enteric diseases

- a healthier gut lining with reduced atrophy from inflammation leads to a thinner mucosal

wall and more efficient digestion and absorption of nutrients from feed.

- better feed uptake and conversion efficiency leads to faster growth, weight gain, and

healthier animals.

AGPs are typically used [10] via direct supplementation in the feed and/or drinking water of

animals for ease of distribution to the entire herd or flock at sub-therapeutic levels (see Table 3)

to protect against unwanted disease outbreaks. The positive effects of AGPs are more prominent

when used in stressed and vulnerable animals, while healthy animals in sanitary facilities

experience less growth promoting effects [16]. Piglets of post-weaning age (3-5 weeks old) face

a challenging time when they are separated from the sow and transferred to growing farms, and

antibody protection supplied from the sows’ milk becomes cut off. Their young immune systems

and gut flora undergo major changes during exposure to a new environment [19]. AGPs can

suppress invasion of the livestock gut by harmful bacteria (E.coli) and reduce disease severity so

that intestinal atrophy is minimized [20], allowing better nutrient absorption and utilization.

When outbreaks of disease are detected, the antibiotic levels are increased to therapeutic

concentrations and intravenous administration may be used. As an indication to the scale of

antibiotic use, it was estimated in 2005 that in the US alone between 17-25 million pounds of

antibiotics were used per annum in livestock (e.g. swine, poultry, cattle, sheep) production for

non-therapeutic purposes, far outweighing the amounts used for human therapeutic uses of

around 4.5 million pounds. The millions upon millions of tonnes of antibiotics used over many

decades [2, 14] in agriculture have repeatedly exposed bacteria to sub-therapeutic levels, which

4

puts selective pressure on them to adapt and acquire resistance mechanisms and to continue to

proliferate while susceptible bacteria are suppressed. Many examples (e.g. tetracyclines,

penicillins, aminoglycosides) show that antibiotic use on farms directly correlates to the

appearance of resistant strains [2, 21, 22] in addition to potentially developing cross-resistance to

other classes of antibiotics [21, 23].

Due to proliferation of antibiotic resistance, pathogens can spread unhindered between

susceptible hosts through transfer by cross-contamination of farms, clothing, equipment surfaces,

irrigation run off, soil, among animals by the fecal-oral or respiratory route, and by humans

through contact with carcasses during harvest and processing, leading to potential outbreaks [24-

26]. Therefore, successful pathogen reduction at the farm and host (animal gut) levels would

greatly reduce the human exposure rate to AR pathogens. For example, in a risk assessment

model for campylobacteriosis from chicken producers, a ~2 log reduction of C. jejuni at the

carcass level could reduce the number of cases of human illness by 30 fold due to mitigation of

risk at the initial sources of cross-contamination [27].

1.2.1.1 Mechanisms of antibiotic resistance

There are four general mechanisms that bacteria may possess to counteract the effects of

antibiotics:

Many resistance determinants have been identified to be encoded in the DNA of bacteria and can

be transferrable via plasmids [26]. There are over 1000 known variant genes of the β-lactamases

[2, 28], and the recently identified New Delhi Metallo-B-lactamase (NDM-1) enzymes confer

resistance to many β-lactams like carbapenems and cephalosporins while membrane efflux

pumps remove antibiotics and antimicrobials from the cell interior [29]. These resistance

determinants can be passed on via horizontal (between allowable species) and vertical (from one

5

generation to the next) gene transfer events like conjugation or from external environment

transformation [26].

1. Efflux pumps – are proteins expressed at the membrane of bacteria able to transport

specific and broad ranged antibiotic substrates out of the cell using energy dependent

processes. Tet A/B/C/D pumps can remove tetracyclines, MefA and MsrA can efflux a

range of macrolides and type B streptogramins, while Mex A/CD/EF/XY is responsible

for cross resistance to ciprofloxacin, nalidixic acid, tetracycline, and chloramphenicol.

Some types of pumps can remove more than one type of antibiotic at the same time and

many other pumps exist for different species of bacteria. Usually, bacteria become

resistant when expression of these efflux pumps are upregulated through substrate

induction or gene regulation [30].

2. Deactivation or modification by enzymes – e.g. β-lactamases and cephalosporinases are

able to cleave the structure (β-lactam ring) of antibiotics rendering them unable to bind to

the target cell wall trans-peptidases. In other examples, Erm, an erythromycin methylase

modifies the ribosomal target and prevents binding by the drug, while aminoglycoside

phosphotransferases (APH’s) and methyltransferases are bacterial encoded enzymes able

to confer resistance to aminoglycosides via enzymatic modification [29].

3. Reduced membrane permeability to drugs – by changing the lipid composition density

of membrane porin channels, bacteria can reduce the passive diffusion of small molecule

antibiotics to the interior, and when combined with other forms of resistance such as

efflux pumps, antibiotic resistance is further augmented.

4. Mutation/modification of the target site – results from random mutations of the

sequence encoding the target protein, thus altering the binding affinity of antibiotic, e.g.

6

fluoroquinolone resistance. A more recent, comprehensive listing of known resistance

mechanisms characterized for specific species of bacteria can be found elsewhere [29]

1.2.1.2 Current situation and costs of antibiotic resistance (Banned the further use of

antibiotics)

Due to world health authorities recognizing the dangers of unrestrained use of antibiotics [22],

many countries have banned the non-therapeutic and limited the use of therapeutic antibiotics in

animal farming including such drugs as enrofloxacin in poultry production, tylosin phosphate,

and spiramyin. The EU has since 2006 prohibited non-therapeutic use of antibiotics including

penicillins, tetracyclines, and streptogramins in livestock production, while other countries (in

North America) are continuously evaluating which antibiotics to regulate depending on the level

of importance (Class I-IV) to human medical use, with class I being of highest importance which

includes ceftriaxone, ceftiofur (3rd generation cephalosporins), amoxicillin/clavulanic acid, and

ciprofloxacin and enrofloxacin (fluoroquinolones) [21] in efforts to curb worldwide spread of

resistance and cross resistance [23, 31]. Individuals infected by bacteria resistant to these drugs

will have very limited treatment options. The pathogens of concern include Salmonella

Typhimurium with the ACSSuT phenotype (penta-drug resistant: ampicillin, chloramphenicol,

streptomycin, sulfonamide, tetracycline), vancomycin resistant Enterococcus (VRE), methicillin-

resistant Staphylococcus aureus (MRSA), fluoroquinolone- and extended spectrum

cephalosporin-(ESBL) resistant pathogens.

1.2.1.3 Impact of antibiotic withdrawal on farming practices

In some cases the sudden withdrawal of antibiotic use on farms [17, 32] have led to noticeable

reduction in specific strains of antibiotic-resistant pathogens [16], but there have also been

negative consequences on livestock production including: increased microbial contamination and

7

sick animals susceptible to diarrhea, more serious disease outbreaks (diarrheal) on farm, and

lower quality and yield of animals or death, which has consequently resulted in greater use of

antibiotics for therapeutic purposes [16, 33]. The efficiency of food animal production has

declined as animals are more susceptible to certain diseases since pathogens are not controlled

[34]. Therefore, there is urgent need to identify and implement viable antibiotic alternatives and

treatments. They should be cost-effective for farmers to apply for protecting the health of

livestock animals, yielding “growth promoting” effects, and counter the spread of antibiotic

resistant pathogens between animals and humans by avoiding use of medically important

antibiotics for farming [3].

1.3 Evaluating alternatives to antibiotic use

Many alternatives have been proposed for controlling pathogen levels in animals including, but

not limited to: changes in animal husbandry such as animal housing practices, proper waste

management, use of natural chemicals or biological additives such as plant essential oils, organic

acids, inorganic compounds, bacteriophages, probiotics, prebiotics, and bacterial

antigen/antibodies administration to control undesirable bacteria. A listing of the many

alternatives are summarized below in Table 1.1 [3, 10, 17, 35-40].

8

Table 1.1 Possible antibiotic alternatives listed from current literature sources

Improvements to

husbandry and

farm biosecurity

Better waste management, biosecurity measures in place

Housing and hygiene/disinfection practices

high temperature feed pelleting

Biological

additives

or methods

Immune modulators (cytokines, antibodies)

Vaccination/immunization against known pathogens using bacterial antigens

Bacteriophages, phage lytic enzymes (lysins)

Probiotics (competitive exclusion: Lactobacillus, Bifidobacterium, Bacteriodes,

Bacillus cereus)

Bacteriocins, nisin, other antimicrobial peptides

(bio)chemical

additives or

methods

Plant extracts/essential oils

Prebiotics (e.g. fructooligosaccharides, mannose-oligosaccharides,

isomaltooligosaccharides, inulin, etc.)

Enzymes as digestion aids, xylanase

Organic acids/acidifiers (VFAs: acetic, lactic, propionic, butyric acids)

Inorganic compounds (zinc, copper salts)

Supplements (vitamins, amino acids, herbs, etc.)

1.3.1 Essential oils as potential antibiotic alternatives

Of the various alternatives, plant EOs are one of the more readily available and safe choices that

are being investigated [39, 41, 42]. Essential oils are the volatile, lipophilic extracts obtainable

from a wide species of aromatic plants from processes like steam distillation or solvent

extraction. EO from a plant may consist of between 20-60 individual chemical components

(sometimes more) through identification by GC-MS analysis, but many EOs consist of only 2 or

3 predominant compounds of between ~20-85% composition that generally reflects the EOs

activity [42-44]. Chemically, EOs are made up of mono-, sesqui- and diterpenoids,

phenylpropanoids, fatty acids, beneznoids, phenolics, and related small molecular weight

compounds that are biosynthesized in the plant as secondary metabolites. Some common EOs of

interest and their plant of origin are found in Table 1.2. Several of the main components of EOs

have been identified and are produced industrially via synthetic pathways.

9

Table 1.2. Various essential oils with antimicrobial activity and their origin

Essential oil Plant source Major component(s)

cinnamon Bark/leaves of cinnamon plant (Cinnamomum spp.) trans-cinnamaldehyde, eugenol

clove Flower buds of Syzygium aromaticum eugenol

oregano Leaves of Origanum vulgare (oregano) carvacrol, thymol, p-Cymene

thyme Leaves of Thymus vulgaris thymol, carvacrol

lemon myrtle Leaves of Backhousia citriodora (Lemon myrtle) , citral, geraniol

coriander Coriandrum sativum

Coriander/cilantro

linalool

sage Salvia officinalis Camphor, α-,β-pinene, α-tujone

rose Rosa gallica petals geraniol

EOs and their components have been traditionally found in many applications as food

flavourings, fragrances/perfumes, spices, and in natural medicinal/homeopathic preparations for

over hundreds of years. The known effects of EOs include: antibacterial (bactericidal), antifungal,

antioxidant, insecticidal, anaesthetic, analgesic, anti-inflammatory and anti-cancerogenic when

tested in laboratory conditions [39, 44]. The antibacterial activity of EOs has spurred recent

investigations into their use in applications ranging from food preservatives to antibiotic

substitutes. Plant derived EOs are listed by the Food and Drug Administration as generally

recognized as safe (GRAS) for use as food additives [39] making them potentially valuable and

safe substitutes to agricultural antibiotics. Some veterinary uses include EOs as an appetite

stimulant which leads to production of more digestive juices in animals during feeding [45].

The bio-activity of EOs can be attributed to their roles in the plants themselves. Plants produce

these volatile, secondary metabolite compounds to serve various ecological and protective

functions and may prevent parasite and fungal invasion or deter insect and herbivore feeding [46].

The mechanism(s) behind the antibacterial activity of EOs are not fully understood due to their

complex compositions and the potential for multiple sites of action (membrane or cytoplasmic

proteins/enzymes, cell wall constituents, or DNA/RNA). Some studies provide insight into the

action of specific components of EOs (e.g. carvacrol, thymol and cinnamaldehyde) active against

10

gram-positive and gram-negative bacteria [47-49]. The components of EOs with strongest

antimicrobial activity fall into three major categories: phenolics, organic acids, and aromatic

aldehydes. Each has a different mechanism of action but EO compounds are generally lipophilic

and can similarly act at the surface or interior of membranes or by traversing the cell wall and

cytoplasmic membrane to gain entry into the interior of bacteria.

Phenolics: (carvacrol, thymol, eugenol) are pro-oxidant and can inhibit enzymes and cellular

components by causing damage to the structure and function of cell membranes, leakage of ions

(K+, H+) and ATP leading to dissipation of the proton motive force and membrane associated

components like porins and LPS, while at higher concentrations they can cause precipitation of

cytoplasmic proteins and cell death [50] [51-53]. Pathogens susceptible to these compounds

include E. coli, S. aureus, and L. monocytogenes at 5 mM concentrations. The location of the –

OH group on the phenol was determined to affect the degree of antimicrobial activity[54].

Organic acids: (acetic, benzoic, propionic, butyric, sorbic acids) are typically weak acids and act

as a proton shuttler bringing protons and the counter anion (via undissociated form) into the

cytoplasm causing a lowering of intracellular pH and affecting the proton motive force,

interfering with the energy needs of the cell. In a food matrix, the acidifying action lowers the

pH of environment and becomes less favourable for microbial growth. Energy is expended while

attempting to restore gradients thus hampering the cell growth activities [18, 52]. The

antibacterial activity of an organic acid was found to be stronger if the pKa was higher [52].

Aromatic aldehydes: (salicyaldehyde, benzaldehyde, trans-cinnamaldehyde, citral) The

aldehyde functional group present in these compounds are reactive towards nucleophilic targets

such as amines, thiols, and hydroxyls and are able to form Schiff bases with pendant groups of

proteins and/or DNA, thus able to deactivate membrane enzymes and transporters. This action is

11

similar to the alkylating cross-linking agents glutaraldehyde and formaldehyde. By interfering

with cell wall synthesis enzymes and electron transport chain proteins, processes such as cell

wall synthesis and energy production can be hindered or even kill the bacteria in higher

concentrations. Trans-cinnamaldehyde was found to be bactericidal to L. monocytogenes at

30mM concentrations [55] through interfering with the glucose uptake, utilization and energy

production processes (dispersed PMF), and small ion leakage at the membrane but not ATP,

while at 2 mM this compound was inhibitory to E. coli O157:H7 and S. Typhimurium growth

[55] [53] .

1.3.2 Potential resistance mechanisms to EO compounds

Bacteria exhibiting resistance to EO compounds are not well described in literature [51] possibly

from their absence in antibacterial applications, but in general, bacteria may naturally lack the

target of antimicrobial action (cell wall enzymes) or can have an outer membrane acting as a

natural permeability barrier, or they can have efflux transporters to remove antimicrobials [52],

and finally porin proteins are believed to allow rapid exchange of small lipophilic molecules with

the periplasmic space [51, 53]. Some general resistance mechanisms against antimicrobials

include: i) innate insusceptibility – intrinsic, non-induced resistance mechanisms such as lacking

the target of action (e.g. cell wall enzymes) or possess outer membrane or spore coat

permeability barriers. ii) Efflux systems – enzyme pumps present in the membrane enable

removal of antimicrobials that have accumulated inside a cell, for example the yeast

Saccharomyces cerevisiae possesses a multi-drug resistance pump Pdr 12 and a H+-ATPase that

export intracellular preservative anions and protons, preventing their accumulation. iii) Acquired

resistance occurs when genes encoding the efflux systems or inactivating enzymes (e.g.

penicillinase gene) are transferred within or between species of bacteria via gene transfer events

12

(naked DNA uptake, conjugation, infection by phage, etc.) and the genetic material can be used

by the previously antimicrobial-susceptible bacteria due to high genetic plasticity and short

generation cycle of microbes. iv) Inactivation by enzyme modification - specific enzymes that

degrade the antimicrobial may exist (e.g. cytoplasmic enzymes acetylating tetracyclines,

penicillinases hydrolyzing ß-lactams, and so forth). v) Stress adaptation – some bacteria are

known to induce energy-dependent processes to restore homeostasis that has been disrupted by

antimicrobial. S. Typhimuruim and E. coli have an acid tolerance response to allow survival at

pH 3. Yeasts can reduce membrane permeability to benzoic acid, resulting in altered diffusion

coefficient. The extent and variability of potential resistance to EO compounds requires further

study to ensure their long-term effectiveness.

1.3.3 In vitro activity of EO’s review

The potency of EOs against food pathogens has been well described in literature through

evaluation of their minimum inhibitory concentrations (MIC) and minimum bactericidal

concentrations (MBC) against pathogens in vitro [41, 54, 56]. Two common methods employed

are the paper disk diffusion method that measures the size of a zone of inhibition on an agar plate,

and the broth dilution method that monitors bacterial growth over time in the presence of EOs in

a liquid growth medium at various concentrations [42, 44, 57]. Potency of EOs can vary widely

depending on the EO used, its composition, how it is used and against which bacteria. Table 1.3

compares the MICs of some antibiotics and some essential oil compounds and the susceptible

bacteria [41, 58].

13

Table 1.3. Examples of antibiotic growth promoting (AGP) concentrations used and

effective EO concentrations

Antibacterial type Active concentration

range (g/kg feed)

Prophylaxis against/Treatment for

Avilamycin

Tylosin sulphate

Spiramycin

Aureo S-P 250 premix

-chlortetracycline

-sulfamethazine,

-penicillin

Baytril (enrofloxacin)

bacitracin

virginiamycin

0.04

0.044-0.110

0.110

0.110

0.055

Necrotic enteritis (NE) Clostridium. perfringens

Swine dysentery, porcine enteropathy, ileitis,

Lawsonia intracellularis,

Salmonellosis and bacterial necrotic enteritis

caused by S. cholerasius

Reduction of incidence of cervical abscesses

caused by group E streptococci

Coccidiosis, mycoplasmas, E. coli , respiratory

diseases

Enteric diseases, mastitis, salmonellosis

Cinnamon oil

Cinnamaldehyde

Carvacrol

Thymol

Pancosma Xtract®

(carvacrol,

cinnamaldehyde,

capsicum oleoresin mix)

0.200

0.056-0.200

0.200-0.255

0.015

0.009

0.006

E. coli O157:H7, K88

S. Typhimuruim DT104

Increased animal growth characteristics

Numerous studies performed in vitro demonstrated the antibacterial activity of EOs, supporting

their potential use in controlling bacterial growth in liquids and foods preservation. In liquid

based foods like unpasteurized apple cider, E. coli O157:H7 was reduced effectively by

cinnamon and clove oils when combined with pH adjustment or mild heat [59] and in carrot

broth, growth of Bacillus cereus was completely inhibited by many different EO compounds for

up to 60 days [60]. EOs of bay, cinnamon, oregano, clove were inhibitory towards (prolonging

mean generation time and lag growth phase) Bacillus cereus, Listeria monocytogenes, E. coli, S.

Typhimurium , and gram-positive organisms were found to be more susceptible at 1.2% vs 3.5 %

for gram-negative species, and furthermore pH and salt conditions could increase the efficacy of

EOs [43]. Helander et al. explored the effects of essential oil compounds (carvacrol, thymol, and

trans-cinnamaldehyde) against gram-negative bacteria and found the MICs to be in the range of

14

1-3 mM against E. coli O157:H7 and S. Typhimurium in liquid culture, which caused effects

including membrane disruption, lysis, release of LPS, and ATP pool depletion [53].

Chorianopoulos et al. demonstrated that carvacrol and thymol rich essential EOs effectively

inhibited growth of and were able to kill 5 well known food-borne pathogens S. aureus, B.

cereus, S. enteritidis, E. coli, and L. monocytogenes at concentrations of 1% v/v in liquid form or

as 5 µL EO added to a paper disk [42]. Cinnamon oil was reported to be one of the most potent

EOs against the drug-resistant pathogens S. Typhimuruim DT104, E. coli K88 , and E. coli

O157:H7 [41] with trans-cinnamaldehyde (CIN) being the most active component in cinnamon

EO that was found to have broad spectrum activity against 9 bacterial species (including MRSA,

K. pneumoniae, Salmonella) in addition to mosquito larvae, molds, dermatophytes, and yeasts

when used at 250-500 μg/mL in liquid culture [57, 61, 62]. Due to this strong reported activity

of CIN, it was selected for development into a formulation for in vivo delivery targeted towards

intestinal pathogens. Table 1.4 summarizes some properties of this natural compound. Other

EO compounds with good antibacterial activity include carvacrol and eugenol [41, 49, 62, 63].

Table 1.4. Properties of trans-cinnamaldehyde the major active component found in

cinnamon bark oil

MW: 132.16

Density 1.048-1.052 g/mL (25°C)

MP : -7.5°C

BP : 246°C

FP: 111°C

Water solubility: 1 in 700 parts water (1.4g/L @25°C)

Miscible with ether, alcohol, chloroform

LD50 (rat): 2.2g/kg

15

1.3.4 In vivo studies

Although specific EOs exhibited strong activity in vitro, the EOs were mostly effective when

pure or emulsified forms were applied directly to liquid or solid food media, which enables

uniform contact between EO and pathogens. Shelef discussed some properties of food matrices

that were found to limit the potency of EOs such as the water content, level of fat/oil, and

amount of solids present [64]. Lower water and higher fat contents increased the amount of EO

needed to exert antibacterial effects due to the transfer of EO components out of the aqueous

phase and becoming trapped in oily phases. Further difficulty arises when EOs are to be

delivered in vivo to target pathogens in the animal GI tract. EOs are highly lipophilic and can

naturally partition across mucous membranes along the GI tract, causing reduced availability at

the lower regions of highest pathogen density and activity. Michiels et al. administered four

separate EO compounds (carvacrol, thymol, eugenol, and trans-cinnamaldehyde) orally to

groups of piglets (~20 kg) after mixing them directly into the corn starch fraction of feed at

concentrations between 12-13mg/kg to study the kinetics of oral absorption of these compounds.

The compounds underwent rapid absorption in the stomach and proximal small intestine, leaving

only a small fraction of the initial dose at the lower regions of the GI tract, with microbial

degradation accounting for additional loss of the compounds as well [58].

In a follow up study, the authors compared different formulations of carvacrol and thymol at 500

and 2000 mg/kg dose levels given simultaneously with feed. The free compounds (first adsorbed

to bulk material) were again found to be almost completely absorbed in the stomach and

proximal small intestine of piglets of dose leading to reduced concentrations and expected

effectiveness further down the GIT (<10% of dose). While an experimental micro-encapsulated

formulation (by coacervation method) was intended to prevent early absorption, the formulation

16

failed to yield any retardation of release or absorption in this study, since the actual

concentrations detected in the proximal small intestine digesta were low (<10 mg/kg for low

dose and <30 mg/kg for high dose) regardless of the formulation, however, their results showed

positive effects of these EO compounds on gut health [65].

In a similar ex-vivo study using piglet jejunum, researchers showed that thymol early absorption

could be diminished by conjugation with water-soluble molecule (β-D-glycopyranoside) which

would be subsequently released by microbial glucosidase activity in the lower tract. Otherwise,

thymol experienced rapid absorption in the proximal small intestine (jejunum), calling for

alternate methods to deliver small molecular weight, hydrophobic EO compounds to the lower

intestine [66]. The above findings were consistent with other studies which reported that much

higher doses (3-10X) were required with oral administration to achieve activities comparable to

those observed in vitro, because EOs may also be sequestered by foods rich in protein and fat

during digestion, reducing the effective aqueous concentrations [39, 51, 62].

In a study performed with 20-day old broilers, it was found that CIN or eugenol added to feed as

free oil at concentrations of 0.5-1% v/w, were able to reduce colonization of Salmonella

enteritidis in the ceca after continuous feeding for 10 days, (however they did not report the

actual concentrations of the EO compounds detected in the lower GIT), but also their inclusion

significantly reduced overall feed intake due to the strong volatility and organoleptic properties

of the EO compounds, evidenced in lower final body weights [67].

Considering these previous in vitro and in vivo studies, therefore, formulation of a site-specific

oral delivery vehicle for EOs could enhance the dose delivered to the lower GIT of animals and

minimize the strong organoleptic properties of EOs affecting feed intake levels, by coatings that

allow release only at the desired regions through control by the dissolution pH. Other

17

advantages include turning the volatile EOs that are irritant, difficult to handle and distribute,

into a more stable solid granular product that is easier for storage and handling.

1.4 The lower GIT of animals: a reservoir for pathogen colonization

Antibiotic resistant strains of foodborne pathogens are typically isolated among the natural flora

of animals in the small intestine, cecum and colon and excreted in the feces. Salmonella

enteritidis and Campylobacter jejuni are common pathogens associated with poultry products

[68] that that are able to colonize (asymptomatic infection) the intestines and cecum of the birds

and can lead to contamination of meat and eggs during processing [69]. E. coli O157:H7

concentrates in the ileo-cecal junction, proximal colon and sometimes the tonsils of pigs [13].

Such pathogenic strains of E. coli rapidly colonize along the epithelia of the jejunum and ileum

of pigs at up to 109 CFU/g of tissue after adhering to specific intestinal epithelial receptors [37].

GI infection by Salmonella in pigs is characterized by increased mucosal secretions,

inflammation, and necrosis starting from the stomach and duodenum to ileum, and more damage

to the mucosal and epithelial tissues of the cecum and colon caused by secretion of toxins and

causes diarrhea [10, 70]. Pigs are well known reservoirs for many strains of Salmonella and E.

coli [8] and worldwide outbreaks of Salmonellosis on farms have frequently occurred in the past

[70]. The common cause of post-weaning diarrhea in pigs is enterotoxigenic E. coli K88 (ETEC).

This K88 strain can rapidly multiply to around 109 CFU/g tissue along the length of the jejunum

and ileum by binding of bacterial adhesins to epithelial receptors found on the absorptive

enterocytes [37]. Once there, the bacterium colonizes the mucous/epithelium and produces

toxins that disrupt the intestinal wall integrity causing diarrhea [10]. Another strain, E. coli

O157:H7 is another pig borne strain that can cause profuse, watery diarrhea, rapid dehydration,

acidosis and death is common.

18

Targeting the delivery of EOs towards the jejunum-ileum-cecum segments of the lower gut

would thus be appropriate to effectively target these pathogens in vivo. To design a site-specific

delivery formulation to target these areas of the GIT, knowledge of its characteristics is needed.

The GI physiology and anatomy of pigs are comparable to that of humans and pig models of

human GI disease are frequently employed, while that of poultry has its unique characteristics.

The digestive tract characteristics of both swine and poultry are given in the Table 1.5.

Table 1.5. Alimentary tract characteristics of swine and poultry

Region pH range Retention time (mins)

Pig (swine), 4 week old

Stomach 1.7-4.3 45-120

Small intestine

-duodenum

-jejunum

-ileum

4.1-6.1

5.5-5.9

6.2-6.8

150-180

Cecum 5.6-6.4 120-180

Colon 5.9-6.8 360-720

Rectum 6.4-7.1

Chicken (poultry) (1.75kg)

Crop 5.5 50

Proventriculus 2.5 30

Gizzard 3.5 60

Duodenum

Jejunum

Ileum

5-6

6.5-7

7

5-8

20-30

50-70

Ceca 7.5 >60

Colon/rectum 7.5-8 20

[38, 70-73]

To trigger release one can exploit the pH variations along of the GIT of animals to release actives

at the desired site by applying surface coatings of pH dependent- soluble polymers [74]. The

capacities and residence times can be observed to design the dose size and release rate of the EOs.

Knowing the sites of pathogen colonization and the associated pH range of those regions enables

a site-specific delivery formulation of EOs to be designed.

19

1.4.1 Challenges in the oral delivery of EOs and compounds

Almost all essential oils are volatile and lipophilic, being based upon the basic structures of

isoprene/terpenes (mono-, sesqui-, and di-terpenes), terpenoids, aromatic compounds and most

are chemically reactive [44], with properties comparable to flavours used in food processing.

Such compounds require protection against oxidation, interactions with other food components,

and evaporative loss on storage. Although specific EOs exhibit good in vitro antimicrobial

activity where the compounds come in direct contact with the bacteria, difficulty arises when

EOs are delivered in vivo. There are several factors to consider for delivering high

concentrations these hydrophobic compounds to the lower GI tract of livestock animals: (also see

[43])

- EOs are lipophilic and readily penetrate the mucous membrane [75] and be absorbed

rapidly in the upper GI tract

- Studies showed that higher doses (3x-10x) were needed to achieve similar effects in vitro

because EOs may be trapped by partially digested foods rich in protein or fat [39, 51, 62]

- Cost of materials and preparation must not be too high

- EOs are liquid and volatile, highly odorous which may limit the concentrations of free oil

to be used (organoleptic properties affecting feed intake)

- EOs are susceptible to oxidation/degradation and evaporative loss, requiring protection

1.5 Delivery approaches applicable to EO compounds

1.5.1 Review of suitable EO delivery methods

Protection of EO compounds by may be achieved through microencapsulation.

Microencapsulation is a process used to entrap a core material by another outer wall material to

20

obtain microcapsules or microparticles of between 1-1000 µm diameter [76, 77] and is widely

used in the chemical, fragrance/cosmetic, and pharmaceutical industries to encapsulate a wide

variety of actives of liquid, solid or gas states and includes flavours, pesticides, pharmaceuticals,

inks, nutrients, and cells [78]. Common wall materials used in encapsulation include natural

polymers like gums, starch, maltodextrin, cellulose derivatives, alginates and others. Although

these materials are safe and food-grade, they not entirely suitable for site specific release (i.e. pH

dependent release), lacking the reproducibility of synthetic polymers. Moretti et al used a

coacervation technique with gelatin and glutaraldehyde as the encapsulating material to form

microcapsules with high loading, but the release of EO was prolonged for up to 30 days since

their application was for insecticidal purposes [79].

Another method to physically entrap hydrophobic drugs is by microparticle/microcapsule

formation. Microparticles have a matrix structure where the core material is evenly dispersed or

dissolved within the matrix material while microcapsules have the core material concentrated at

the center while being surrounded by an outer wall material [80]. Methods such as hot-melt

fusion to prepare microparticles (<10 µm) can be cost-effective and simple. Solid lipid

microparticles are prepared by emulsifying a melted lipid containing drug into a continuous

phase and then dispersed into droplets of desired size within an incompatible phase, followed by

cooling to solidify into solid particles, furthermore lipophilic drugs or excipients may be

introduced at the molten lipid stage. Materials employed are typically safe and well tolerated

physiological or food-grade lipids such as fatty acids and their glycol esters [81]. Lipid based

microparticles have been used for parenteral drug delivery and are typically much smaller in size,

so methods can be adapted to make granules of suitable size for mixing with feed and taken

orally.

21

Ukita et al, prepared granules loaded with essential oils using a high-shear melt granulation

method but their loading level was low with <3% EO [82]. Pellets containing similar volatile

plant-extract compounds were prepared by a rotary fluidized bed granulation for sustained

release but incorporation of >10% actives reduced the quality (size and shape) of cores

drastically [83].

Since lipid microparticles are primarily directed at parenteral delivery, a similar preparation

method can also be adapted to increase their size to millimeter sized granules. Larger granules

and pH-sensitive release will be employed together with high EO loading to formulate controlled

release granules for the oral route. This process is easy to scale up and does not require

expensive high-pressure homogenization to yield nano-sized particles or other complex

machinery. A similar process for entrapping EOs is explored for the current purposes.

Considering the above factors from literature reviews, a form of encapsulation that allows for

high loading and achieves rapid release at the desired intestinal region, is used with the following

chosen components:

- An inert powder material for high oil retention capacity

- Melt-solidification technique to form granules suitable for coating , preferably at low

temperature to minimize evaporation of EO compounds

- A polymeric coating to allow targeted release, site specific delivery

The conveniences of a multi-particulate formulation includes ease of handling, prolonged

stability and storage, better distribution within feed and the animal digestive tract when

compared with the pure liquid EO.

22

1.5.1.1 pH sensitive polymers (enteric polymers)

To achieve site specific drug release, polymer coatings can be applied to dosage forms which

dissolve in environments at and above the dissolution pH. One widely used group of polymers

with a pH dependent solubility is the Eudragit® polymers that have neutral, cationic, or anionic

types depending on the functional groups present. The anionic types are synthesized from

(meth)acrylic acid monomers (acting as weak acids), which imparts pH dependent solubility

behaviour. They are linear polymers made with the co-monomers ethyl acrylate, methyl

methacrylate, or methacrylate. For example Eudragit® L-100-55 is designed to dissolve in

aqueous environments when the pH is above 5.5-5.6 [84]. Below the threshold pH the polymer

is insoluble so that contents completely coated with this polymer will not be released in the upper

GI tract, allowing the dose to be delivered to the lower GI tract where the pH is between 5.5-7.2.

Granules coated with Eudragit® L100 for example can target release of EOs in the regions of pH

≥ 6 starting at the jejunum and continue in ileum and cecum of animals. In order to have

complete protection, the coated objects require appropriate coating thickness and uniformity,

devoid of defects which may result in incomplete protection. Other factors affecting the

performance of coated dosage forms include level and type of plasticizer(s) used, shape and size

of dosage form, properties of enteric polymer blends, and so forth.

Table 1.6. Chemical and physical properties of commercial enteric (Eudragit®) polymers

Polymer

Type

Chemical

makeup

Dissolution

pH

MW

(Da)

(g/mol)

Tg

(ºC)

MFT

(ºC)

mols of OH

groups per

polymer

mmols OH

groups per

g polymer

% of OH groups

hydrolyzed to

form clear

polymer

solution

L100-55 50:50

EA:MAA

≥5.5 250,000 110 25 1344 5.4 50%

(pH 5.7)

L100 50:50

MMA:MAA

≥ 6.0 135,000 >150 >100 725 5.4 67-73%

(pH 6.3)

S100 70:30

MMA:MAA

≥ 7.0 135,000 160 >100 471 3.5 81-86%

(pH 7.2-7.4)

*MMA- methyl methacrylate, MAA-methacrylic acid, EA-ethyl acrylate

23

Polymers shown in Table 1.6 are designed to dissolve completely at pH 5.5, 6 and 7 that enables

the targeted release in the duodenum, jejunum/ileum and the colon, respectively. Enteric

coatings are formulated nowadays as aqueous emulsions, and the final mixture is applied to solid

dosage forms. Small spherical or granular particles and tablets can be coated using spray coating

equipment and can be scaled up to large batches, but requires optimization to the type of dosage

forms being coated.

1.6 Bacteriophage as antibiotic alternatives

Bacteriophages or ”phages” are naturally occurring viruses of bacteria found in all sorts of

environments [3]. They were observed around 1915 by Twort and previous researchers, and then

named ultimately by Felix D’Herelle in 1917 [85]. Ever since their discovery, phages have been

fundamental to the advancements in molecular biology (e.g. gene regulation/insertion, phage

display, and so forth), while on the other hand, they were proposed as a way to control

undesirable bacteria in humans, animals (e.g. cholera, bubonic plaque, dysentery, and other

bacterial infections) as well as plants. Phages are estimated to be the earths’ most abundant

organisms (~1031 total) from sampling of diverse natural ecosystems where potential host

bacteria are present, including the ocean, sewage, lakes, soil, the foods we eat, and also various

niches in and around animals such as the alimentary tract, oral cavity, and skin surface [86].

Phages can significantly alter bacterial populations in many eco-systems, but their activity is not

readily apparent unless there is close observation [87]. They have been implicated in affecting

geothermal cycles and are even found in hot springs, being responsible for between 20-50%

bacterial mortality/turnover in the environment causing release of organic matter to feed

biogeochemical cycles, although the total bacterial biomass is not changed considerably due to

renewed bacterial growth [3]. As another example of phage in action, in many industrial

24

fermentations which produce dairy products (cheeses, milk, and yogurt), vinegar, and antibiotics,

a recurring problem is the contamination of production by lytic phages which can effectively kill

off the culture bacteria (e.g. Lactobacteria spp., Bacillus spp., Acetobacter spp.) such that the

fermentation product yields are problematic. The source of contamination typically can come

from whey proteins, raw milk or water supplies, so that phage concentrations need to be

constantly monitored during production to ensure desired products are obtained [88].

As virus particles, phages range in sizes between 30-300 nm in length and consist of two

basic components – the bulk of which is protein forming the overall structure including capsid,

tail, and fibers, followed by nucleic acid as genetic material usually in the form of double

stranded DNA. The DNA is tightly coiled and packed inside the capsid head which serves as a

protective layer against degradative enzymes and harsh environments [89]. Some phages

produce lysozyme which can be used during entry and lysis of host cells by partial break down of

the cell wall structure. Functional proteins found on the exterior (tail fibers) are important in

host cell receptor recognition and binding (adsorption). Phages can be tailed or tail-less

depending on the family they are part of (most are tailed) and their diversity is enormous in terms

of hosts selectivity, genetic variation, genome size and function, while many phages remain

unexplored due to our inability to culture the host bacteria for phage isolation.

1.6.1 Mechanism of action of bacteriophage

Since phages can only replicate in susceptible living host cells, they are termed obligate

intracellular parasites and can be thought of as the natural predators of bacteria. There are two

life cycles of phage. One is the lytic (or ``virulent``) phage cycle, which is desirable for bio-

control applications because of its ability to carry out lysis of bacteria once production of phage

25

progeny is complete. The other is a temperate phage (or “lysogenic``) cycle and is not preferred

from a bio-control standpoint for reasons discussed below.

Lytic phage cycle: phages encode the genes necessary to carry out the lytic cycle, which

involves a cascade of events starting with adsorption by recognition proteins of the phage (e.g.

tail fibers) that bind to receptor sites on the bacteria. Phages can recognize a variety of surface

structures present on their hosts that may include: components of lipoprotein layer,

lipopolysaccharide layer, glycolipids, flagella, pilus, O-antigens, and peptidoglycan layer. The

typical lytic cycle for T4-like phages and related coli-phages outlined in the Figure 1.1 [85, 90]:

26

Figure 1.1 Typical life cycle of lytic phage

Temperate phages have evolved the ability to integrate their genome within bacterial

chromosomes to become “pro-phage” , or entering a dormant state, and can subsequently

reproduce in parallel with bacterial cell division, typical examples being lambda (λ) phage which

infects E. coli K12 [90] and P22 which infects Salmonella hosts [91]. Even though they are able

to reproduce and induce lysis of host cells, majority of the time temperate phages are in the latent,

non-infectious pro-phage state, coexisting with the bacteria host. Their genome contains other

genes encoding proteins involved with a lysogenic cycle, and they may de-integrate from the

host and choose reproduction only when the phage-host pair (termed “lysogen”) is exposed to

appropriate environmental conditions or signals. These signals can be chemical agents, UV

irradiation, ionizing radiation, and carcinogenic/mutagenic compounds [90]. The main concern

Phage adsorption to host cell, injection of phage DNA into bacterial host

Host cells’ enzymes redirected or newly induced to produce viral proteins (“early genes” expressed), superinfection blocked

Viral genome (DNA) replication occurs, host DNA may be degraded for use. Expression of “middle genes” involved with phage production (structural proteins)

Protein synthesis continues, “late genes” expression to produce final structural proteins or enzymes (capsid, tail, base plate, holin, lysozyme,etc.)

structural assembly together with DNA packaging occurs to produce final, mature phage particles

Fully assembled phage particles are released by lysis using lysozyme/holin resulting in burst release or via budding without lysis

27

with using lysogenic phages against pathogens is that they are well known-mediators of genetic

transduction in bacteria, and have the potential to transfer virulence factors between strains,

insert or delete toxin genes (e.g. stx, tst genes), and aid in the overall evolution of bacteria

towards increased pathogenicity towards animal hosts by altering genes associated with

colonization/adhesion, invasion, immune evasion, exotoxin production, antibiotic susceptibility,

and transmissibility, due to temperate phages’ intimate connection with the bacterial host [92].

In order to avoid contributing to pathogenicity of targeted bacteria inadvertently, temperate

phages must be avoided by careful genetic screening and exclusion of potential phages that

contain known antibiotic resistance and virulence-associated genes [85].

Advantages and disadvantages of using bacteriophages for pathogen control are listed in Table

1.7 [3, 85, 93]:

Table 1.7. The advantages and disadvantages to using bacteriophages as biocontrol agents

Advantages to using Phages Disadvantages to using Phages

-Highly host specific, narrowly targets strain

(minimal disruption of flora) by bacteria specific

receptors: flagella, O-antigen of LPS, pili,

peptidoglycan, etc.

-Can replicate at the site of infection under high

bacterial load, “self-replicating”

-Non-toxic towards eukaryotic cells

-Relatively easy to isolate and select for new

phages in a short time

-Continuous co-evolution alongside bacterial hosts

in natural cycle

-Naturally high abundance in nature and selection

for targeting of pathogenic bacteria will likely

minimize negative effects on environmental

ecology (vs. broad spectrum antibiotics)

- have a large database of known phages, but a

large diversity remains to be discovered and

characterized

-Need to know the specific strain causing

problems/infection (etiologic agent) , can use broad

host range or cocktail of phages.

-Exponential growth of phage occurs only under

specific set of conditions (timing and dosing

interval)

-There is potential to release endotoxins during

lysis and cause toxicity indirectly

-Resistant clones may emerge due to phage attack

-Pharmacokinetics of phage and reproduction in

host cells are not clearly understood and can vary

with their physiological state, environment

(intracellular invasion), and phage-host

combination

- Safety of phages, need to be monitored for

unfavourable genes; purification techniques need to

ensure phage preparations are pyrogen and antigen

free. Over time antibodies to phage may develop in

the blood.

28

Lastly, there are two modes of phage therapy, active where phage continues to reproduce under

sufficiently high host density and growth conditions, and maintaining exponential growth

requires the accurate timing of phage administration after bacteria establishes its infection [94].

Active phage therapy may or may not occur in host animals even though phage growth occurs in

vitro. The second mode of phage therapy is passive, where lysis of target bacteria is occurring

and exponential growth of phage is of secondary importance; in this case the starting dose of

phage, or more precisely the phage to bacteria (MOI) ratio is more important [95]. There are over

1000 Salmonella specific phages known, and Felix O1 is a virulent phage with a double stranded

DNA genome and is a member of the Myoviridae family [91]. FO1 is considered a broad host-

range phage within this genus and is well characterized and would be safe to use as a model

phage [96].

1.6.2 Phage for bacterial reduction/Phage therapy

The earliest therapeutic use of phages was in humans to treat hemorrhagic dysentery, cholera,

and bubonic plague, with phages simply isolated from waters and rivers in areas with disease

outbreaks [85]. Phage therapy in humans continued in Eastern Europe, Poland and the former

Soviet Union (now Georgia). Recent studies have demonstrated that phages administered via

oral or injection routes, were safe in human volunteers [97, 98] and different animal models.

They have been studied for use against bacteria on foods like frankfurters, ready to eat meats,

fresh produce (tomatoes, peppers) to control unfavourable pathogens like Salmonella enterica,

Listeria spp., E. coli O157:H7, Campylobacter spp., as well as in aquaculture, [3, 99]. A few

products already approved by the FDA include Listex P100 for use in cheese manufacturing and

the meat industry for controlling Listeria as a source of food poisoning, while the EPA approved

use of phages against bacterial diseases and Salmonella in pepper and tomato plants. The human

29

gut has been shown to be readily exposed to phages naturally found in drinking water and food

supplies, such that mammalian organisms are constantly in contact with phages in the

environment [86]. Furthermore, the phages found within the gut were temperate phages with

less variability and remain dormant unlike phages found in aquatic environments which

participate in constant bacterial population renewal cycles driven by naturally occurring virulent

phages [3]. In humans, infections of the lungs, nose, throat, ear canal by antibiotic resistant

Pseudomonas aeruginosa and Klebsiella pneumoniae have been experimentally treated with

intranasal phage with some success [3] while intestinal pathogens have been targeted by using

oral phages as well. Studies showed that a cocktail of 9, T4-like phages against E. coli given

orally to healthy human volunteers showed no adverse effects on the normal physiology or organ

functions and phages were detected in the stool (~1% of dose) and were also eliminated by the

spleen and liver [98, 100]. Other routes of administration with various efficacies and success to

treat sepsis include i.v. or i.p. administration and phage incorporated into creams for treating

topical skin infections have been reported as well.

1.6.2.1 Studies utilizing phages for bio-control of bacteria in animals

Various phages against E. coli strains infecting mice, calves, piglets and lambs were found to

successfully reduce the incidence of diarrhea and death and improved the outcomes of disease,

however, some resistant colonies emerged that were also less virulent [101]. Further

investigation revealed that phages were differentially deactivated by temperature shifts, pH

(acid), and antibodies present in colostrum and co-factors such as calcium were needed for lysis

to occur [102].

Indeed, phages given orally could be found in the blood, urinary tract, and organs of animals and

humans while i.v., i.m. and i.p. injections have allowed phages to be detected in the brain, past

30

the blood brain barrier during experimental bacterial meningitis or encephalitis of animals [86,

98]. Vibrio cholera in a rabbit model showed decreased shedding after treatment with a 5-phage

cocktail [103].

In other models, mice infected with lethal doses of V. vulnificus strains were rescued from

mortality and severe disease through administering phage in high doses (108 PFU) and optimum

timing [104]. Phages isolated from finishing pig farms were found to partially reduce the

salmonella levels in the gut and only significantly in the rectum of animals, concluding that

optimizations to the phage cocktail and dosing schedule were needed [105].

1.6.2.2 Phage therapy in poultry and other livestock

In 2-day old chicks, it was found that lytic phages against S. Typhimurium could reduce

mortality rates only when high numbers (>1010 PFU/ml) were used, and phages were able to

multiply in the alimentary tract (ceca) of the animals; however, the transient levels of phage and

bacteria were short lived necessitating the search for more effective phages [106]. Huff et al.

showed that colibacillosis in poultry could have reduced mortality when given phage treatment to

the lungs via aerosol spray and by i.m. injection [107]. The same group then showed that higher

titers (108 PFU) significantly reduced mortality rates of chickens when phages were used i.m. in

a single dose [108]. Follow-up research indicated that previously exposed chickens could

produce antibodies (IgG type) to phages that decreased their efficacy in older chickens [109].

Phage cocktails against Salmonella enteritidis (SE) NalR infection of chickens were able to

partially cause reductions in the cecal counts and shedding [110]. Chickens receiving E. coli

phages via different routes (oral, spray, i.m.) experienced penetration into the bloodstream,

organs, and infection sites to various extent, concluding that higher doses were more effective at

lowering bacterial loads [111]. Phages isolated from chicken farms were combined in a 3-phage

31

cocktail, and when used in broilers produced lasting effects on reducing the cecal counts of SE

PT4 [112]. While SE reductions were achieved when sewage-sourced phages were used in

combination in spray or oral forms after 10 days, reducing the proportion of infected animals

[113]. A cocktail of phages was found to effect reductions of SE in the lower intestine of

chickens but for only a short period of time (less than 24 hours) [114]. C. jejuni and C. coli

infections in chickens can lead to human campylobacteriosis and a phage cocktail given orally

and in feed to chickens was found to reduce the counts in the feces by around 2 log10CFU/g and

could prove useful for control in commercial poultry production, although resistant colonies were

isolated at low frequency [40] similar to the results from an earlier study[115]. Atterbury et al.

concluded that high titer (~11 log PFU) of phage by oral passive therapy would be needed to

significantly reduce Salmonella colonization in the ceca since active therapy requires a phage

proliferation threshold that appears to be impractical to achieve in vivo due to presence of non-

specific binding to phage, low host density and less favourable phage-bacterial interactions than

in vitro. To prevent acid deactivation of phage, co-administration with CaCO3 was done.

Furthermore, differences in the effectiveness of their phages against various Salmonella strains

was observed, hence suggesting identification of more varied phages and their host receptors in

order to reduce the chance of resistant- mutants arising and allowing more precise targeting [32].

Phage Felix O1 was proposed as a treatment to reduce the S. Typhimuruim load in pigs just

before slaughter within a short time frame [116]. A good review of phages in potential

applications against AR pathogens and zoonotic agents in cattle, poultry, pigs is provided

elsewhere [117]. Even though there have been extensive literature reports on the encouraging

potential of phages to effect reduce pathogen levels, the varying results of multiple studies have

yet to prove phages to be completely successful because of the different experimental parameters

32

used such as: the target strain(s) of bacteria, the source of phage(s) used against them, the

timescale and dose of therapy (MOI, concentration), routes and timing of administration, etc. as

exemplified in the studies reviewed.

1.6.3 Considerations and strategies for encapsulated phage for oral delivery

1.6.3.1 Considerations for oral delivery of phages

Bacteriophages in contrast to chemical agents are biological in nature so that maintenance of

their viability is of primary importance during delivery to the desired sites of action. In terms of

the requirements for phage therapy, the formulation should allow for delivery of high doses (high

loading efficiency) and release to the lower alimentary tract of high target pathogen density,

similar to the delivery of EOs. The process should be versatile (can used with phage

combinations) and yield a product that is easier to use and store than liquid phage suspension.

Since the oral delivery of phages involves potential exposure to feed, digestive enzymes and

processes, encapsulation methods that can limit the deactivation of phages would be helpful with

the following features:

-high phage loading efficiency, good preservation of phage viability/infectivity

-can control the release of phage to enhance oral delivery efficiency

-easy to administer, store, modify with new phages

Conventional oral phage delivery methods have included phage suspended in buffer solution, or

together with antacids (CaCO3) applied to the animal feed or gavaged, while non-oral methods

have included such forms as coarse, aerosol sprays to the respiratory tract, or feed, and i.m., i.v.,

and i.p. administration to animals. When phage suspensions are applied to various surfaces or

33

environments they could be deactivated by temperature, sunlight, humidity, and pH exposures

[3] where some researchers have found around ~2 log PFU reduction after simply adding to dry

feed [110].

There have been limited published methods of formulating phages with biopolymers. Previous

work in our group microencapsulated the model phage FO1 using chitosan-alginate beads with a

matrix structure. This work showed that there was good encapsulation efficiency (>93%) of FO1

and protected the phage against acidic deactivation in simulated gastric and intestinal incubations

[118]. Attempts were made to dry the beads and protect the phage with a stabilizing agent but

viability was insufficiently maintained. Addition of pectin to the alginate gel structure has also

been tested [119] and have been shown to more effectively protect phage against acid

deactivation.

1.7 Microencapsulation of phages using Ca-alginate matrix based methods

Alginate (or alginic acid) is a linear, polymer hydrocolloid obtained from a variety of natural

sources, primarily seaweed plants [120] consisting of alternating β-D-mannuronic acid and α-L-

guluronic acid blocks to form a block-copolymer. Depending on the source and extraction

process of the polymer material, there are different molecular weights (80-290 kDa) as well as

different relative compositions (ratio of G to M) of the blocks, which spans a wide range of

solution viscosities. The material is readily available in the acidic form or sodium salt as alginic

acid or Na+-alginate, respectively.

1.7.1 Ca-alginate gelation process

The water soluble, sodium alginate is used to prepare solutions containing an encapsulant.

Relatively inert and mild gelation conditions are used in forming gel beads that can contain up to

34

95% water, and been used previously to entrap, immobilize or control the release of proteins of

various molecular weights, DNA, drugs, hormones, and whole cells and other sensitive

biological materials [120]. During exposure to acidic conditions such as pH 1 in stomach, the

alginate matrix can undergo shrinkage and constriction of the pores to slow down release,

potentially protecting the inner contents against acid exposure [120]. This is due to formation of

the unionized form alginic acid (pKa of 3.38 and 3.65 for the G and M monomers) causing

decreased ionic repulsion between chains [121].

When the alginate solution is extruded or dropped into a Ca2+ bath the droplets rapidly turn into a

gel through cross-linking caused by the divalent ions in solution. The gelation process results in

a ribbon-like, egg-box shaped structure from polymer chains coming in to close contact from

dimerization and junctions of strong cooperative interactions with Ca2+ (exchanged with Na+)

and the α-D-guluronic acid residues involving >20 monomer units to form the gel network as

shown in Figure 1.2 [120]. The pore sizes of the gels interior can range from 5-200 nm while on

the surface typically 12-16 nm pore sizes are observed. Small molecules like ethanol, glucose,

small molecules etc. readily diffuse in/out of the network while diffusion is limited for larger

molecules like insulin, fibrinogen, antibodies, depending on the molecular weight and charge.

Phages (100 nm) would be expected to be entrapped within the gel.

35

Figure 1.2. Egg-box like structure of Ca-alginate gelation process forming network

The size and shape of resultant beads are affected by the nozzle size, viscosity of the solution, air

pressure, temperature, and by the encapsulant if it is charged and competes with the ionic

interactions involved in gelation. Thus, polycationic proteins can potentially compete with

calcium ions for binding the COO_ sites of alginate, decreasing the cross-link density of the final

gel.

The addition of whey protein to the Ca-alginate gel formulation has been reported as well [122,

123], which can modify the release of entrapped substances further. The inclusion of whey

protein isolate (>90% pure protein) allows for the action of digestive proteases to enhance the

degradation of the beads. Coacervate formation between alginate and protein can also decrease

the cross-linking density of the gel by decreasing the availability of cross-linking sites [120].

The release of phages from such beads would then be dependent on the following processes:

36

1. Degradation of the gel network: ion exchange of Ca2+ ions for monovalent ions in

solution, removing the cross-linking between alginate chains, opening up pores and

erosion of the gel.

2. Digestive enzymes (pepsin, pancreatin) breaking down the protein fraction of the beads

causing erosion of the network.

3. Diffusion outward once the pores have become large enough for phage particles to move

through.

We chose to continue developing the Ca-alginate-whey protein based beads for phage

encapsulation due to the simplicity, good biocompatibility, and high yield encapsulation

efficiency as previously reported [118, 123]. One potential limitation of microencapsulation is

that phage concentrations would be diluted from the initial phage suspension concentration after

encapsulation, so that phage concentrations would not be expected to be higher than this

theoretical limit unless drying of the beads were achieved without loss of viability.

1.8 Hypothesis and specific objectives of thesis

Hypothesis: The oral delivery of selected antibiotic alternatives (EOs compounds and phages)

can be improved over conventional delivery methods, through the formulation of such antibiotic

alternatives using low-cost, functional excipients and methods (considering the chemical or

biological requirements of the actives) to achieve site specific release in the livestock animal

alimentary tract.

The thesis consists of two main projects: firstly based on delivery of essential oils and secondly

on phage delivery. The objectives for the EO project were to:

37

1. Select an EO compound with good antibacterial activity (CIN) for formulation into a

multi-particulate/granular product suitable for oral delivery to livestock animals (chicken,

pig) and is also convenient for storage and handling.

2. Develop solid granules containing liquid EO into a coated, solid dosage form, and is

optimized for loading, release and stability in vitro and in vivo, and preservation of

antibacterial activity.

3. Evaluate the site specific formulation to achieve enhanced delivery to desired regions of

the animal intestine compared to un-encapsulated free oil in animal trial(s) and to

quantify the in vivo release of the active compound.

Objectives of the phage project were to:

1) Enhance the delivery of phage, prolong the viability on storage, improve the storage and

handling options as compared to a liquid phage suspension,

2) Optimize phage cocktail and test antibacterial activity in different environments in

digesta effectiveness against a target pathogen,

3) Study the transit time and distribution of phage following oral administration to broiler

chicks.

4) Evaluate the in vivo ability of encapsulated phage formulations to cause reductions in

pathogen colonization of chickens.

Summary of the thesis chapters

Chapter 2 involves the initial selection and formulation of granules using a model EO compound,

trans-cinnamaldehyde. Screening of suitable materials and phase diagram analysis allowed the

38

maximum oil loading to be determined and production of core granules with up to ~45% w/w

loading. The granule cores were coated with an aqueous enteric coating using a fluid-bed spray

coating method to impart pH dependent dissolution. Furthermore, the CIN was stabilized against

autoxidation by addition of eugenol. The final granules maintained the antibacterial activity

against the test pathogen E. coli K88.

Chapter 3 investigated the in vivo performance of the coated granules, specifically the release of

CIN in the alimentary tract. The contents from 7 sections of the alimentary tract were collected

every half hour for a period of 3 hours post-gavage and then quantified using solvent extraction

and GC methods. The coated granules formulation was compared to a conventional method of

administration (by direct addition of EO to feed) and the results showed that the coated granule

formulation was able to deliver higher concentrations further down the alimentary tract of

chickens and in pigs.

Chapter 4 explored various aspects of oral delivery of encapsulated phages to young broilers.

After a single oral dose, the temporal distribution and persistence levels of viable phage in the

digestive tract of chicks was quantified, also providing evidence that encapsulated phage was

successfully reached the lower gut, even in the absence of host bacteria. Low phage levels were

found continuously excreted in feces. The movement of the dose appeared non- homogeneous

with some portions moving faster than others. The lytic activity of FO1 was screened with other

Salmonella phages in combination to identify more effective cocktails. Although FO1 is

considered broad host range, its lytic activity was found to be insufficient at suppressing growth

of the target pathogen STDT104, and a two phage cocktail was found to be more effective.

Cecal concentrations achieved by oral dosing were likely limited by the selective uptake of the

39

liquid fraction of the digestive contents, so that multiple dosing of phage could be more effective

to achieve stable phage levels in vivo, in the absence of active phage amplification.

To obtain an optimal timing interval for dosing/administration, it was observed that more than 4

hours was required for the majority of the dose to move through the chicks’ alimentary tract

following an oral dose. Lastly, various exposure conditions (fecal, feed, buffers, and simulated

intestinal conditions) suspected of deactivating the viability of phage during oral administration

were not found to affect phage FO1 survival.

Chapter 5 covered further optimization of a phage cocktail against STDT104NalR and was

compared to the previous two-phage cocktail. Two animal trials of phage cocktails in broiler

chicks were carried out. The first cocktail failed to show effectiveness in reducing pathogen

level. In the second trial, an expanded cocktail was used (containing more phages) and was found

effective in achieving reductions in the ceca, colon and ileum of chicks, suggesting that

optimized phage cocktails could be useful in reduction of pathogen levels in broiler chicks, while

further research is proposed to explore the effect of combining multiple unique phages into

cocktails.

40

Chapter 2. Formulation of granules for site-specific delivery of an antimicrobial essential

oil to the animal intestinal tract

Yin-Hing Ma1,2, Qi Wang1, Joshua Gong1, Xiao Yu Wu2*

1Guelph Food Research Centre, Agriculture and Agri-Food Canada, 93 Stone Road West,

Guelph, Ontario, Canada N1G 5C9

2Advanced Pharmaceutics and Drug Delivery Laboratory, Leslie Dan Faculty of Pharmacy,

University of Toronto, 144 College St., Toronto, Ontario, Canada M5S 3M2

*Correspondence to: X.Y. Wu (Telephone: 416-978-5272; Fax: 416-978-8511; email:

[email protected])

(All experimental work was carried out by YHM under the supervision of QW, JG and XYW,

with the exception that core granules were delivered to The Toronto Institute of Pharmaceutical

Technology for development of the enteric coating by fluid bed)

Reprinted with permission from J. Pharm. Sci. (2016) Vol: 105(3) pp:1124-33. Copyright 2016.

(See copyright acknowledgement)

mailto:[email protected]

41

Abstract

Owing to proliferation of antibiotic-resistant bacteria, the use of antibiotics for livestock growth

promotion is banned in many countries and alternatives to in-feed antibiotics are needed.

Cinnamon essential oil exhibits strong in vitro antibacterial activity, however, direct addition of

essential oils to animal feed has limited practicality due to their high volatility, odor, fast

decomposition, and poor availability in the lower intestines. To solve these problems, we

formulated trans-cinnamaldehyde (CIN) with an adsorbent powder and fatty acid via a melt-

solidification technique. Core granules of an optimized composition contained up to 48% w/w

CIN. The granules were then coated with an enteric polymer to impart site-specific release of

CIN. CIN was mostly retained in simulated gastric fluid, and released rapidly (>80% under 2 hrs)

in simulated intestinal fluids. Rapid CIN autoxidation into cinnamic acid was inhibited by

adding 1% v/v eugenol, which maintained CIN stability for at least one year. The granule

formulation increased the antimicrobial activity of CIN against Escherichia coli K88 slightly

with a minimum bactericidal concentration of 450 µg/mL for CIN in LA- based granules

compared to 550-600 µg/mL for PA-based granules and free CIN, respectively. These results

encourage the potential use of encapsulated CIN for control of animal enteric pathogens by oral

in-feed administration.

(Keywords: Formulation, Site-specific delivery, Anti-infectives, Antioxidants, Coating, Oral

drug delivery, Stability)

42

2.1 Introduction

The practice of supplementing antibiotics to animal feed at sub-therapeutic levels to promote the

growth of food producing animals and prophylaxis has been banned across European Union

members[124] in an effort to curb spreading of antibiotic resistance among pathogens caused by

selective pressure under intensive farming environments.[125] Alternative compounds to in-feed

antibiotics are thus pressingly needed to help maintain economical livestock production. Many

plant-derived essential oils (EOs) exhibiting strong antimicrobial activity in vitro [39, 41, 44] can

be used as alternatives of antimicrobial drugs, shifting the use and resistance development away

from more medically important antibiotic drugs. Of various EOs, trans-cinnamaldehyde (CIN),

the most abundant component of cinnamon oil, shows antimicrobial effects (ranging from

inhibitory to bactericidal) against various bacterial pathogens at concentrations between 56 to

385 mg/L (µg/ml) in liquid media.[41, 63, 126] Additionally, CIN has shown some potential use

in dairy cows to affect rumen microbial fermentation.[127] Hence CIN is a good candidate for an

alternative antimicrobial agent. Nevertheless, the effectiveness of direct addition of free EOs to

the feed is questionable due to several limiting physical-chemical factors such as the high

hydrophobicity, volatility, odor, oxygen sensitivity, and low availability of the EOs in the lower

gastrointestinal tract (GIT) of animals.[58, 65]

In order to improve the availability of EOs to exert their antimicrobial activity in vivo, it is

desirable to develop a formulation that is amenable to oral delivery to the lower GIT of food-

producing animals. For this purpose, solid granules containing liquid EOs would be a good

choice as they can be mixed within animal feed more readily than liquid EOs, and can be further

processed to reduce volatility and degradation of EOs, and to allow site-specific release of EOs

in the GIT by applying a suitable coating to the granules. The granulation and coating can

43

potentially control the effects of EOs on feed palatability for animals by reducing oil volatility

and masking odor.[45]

To obtain EO granules with high loading levels while minimizing the evaporation and odors of

the oils, we have designed a new granule formulation in this work by soaking the oil in an

absorbent powder and then incorporating the powder in a melted lipid to form solid granules.

Various pharmaceutically acceptable powdery, inert absorbents were investigated including

magnesium aluminum silicate, microcrystalline cellulose, and wheat bran powder. The powder

possessing the highest oil retention capability was then selected for further development of

granules with a lipid binder.

The success of developing solid granules containing lipid and EO-absorbed powder depends on

many factors that influence the granule properties, e.g. granule size and strength, EO loading

levels and release kinetics. These factors include the compatibility of the powder with the lipid,

the melting temperature of the lipid and its mixture with the EO-absorbed powder, and the

droplet forming properties that can affect the size and shape of solidified granules. Although

meltable lipids have been employed to prepare solid lipid particles previously to encapsulate

various drugs, such dosage forms were primarily designed for parenteral drug delivery with

small (nano) sizes and low drug loading levels, and in the absence of powdery materials. [82,

128-132] Therefore, we conducted in depth studies on the molten and re-solidified properties of

mixtures in relation to the composition of the three component system and rationally selected the

type of lipid. Simple saturated fatty acids with well-defined melting points from processed oils

of palm or coconut as well as fatty alcohols were tested, as they have a melting range from 44-70

°C and have been successfully applied in melt-pelletization and melt-granulation processes. [82,

133-135]

44

Despite its good antimicrobial activity, the practical application of CIN as an alternative

antibiotic product is limited by its poor stability. CIN is known to undergo rapid oxidation upon

exposure to atmosphere, converting to non-bioactive cinnamic acid (CA). Therefore, a series of

antioxidants were screened to stabilize CIN. Eugenol, an effective antioxidant present in natural

EOs was selected for further study of CIN in granules and its effect on CIN stability was

monitored at room temperature and 4 C for one year.

To protect actives from acidic pH and achieve site-specific drug release in the lower GIT, an

enteric polymer was introduced in the core and surface of granules. The surface of core CIN

granules was coated using a pH-responsive polymer, Eudragit® L 100. Structurally, this polymer

contains acidic monomers (methacrylic acid) co-polymerized with non-ionizable monomers

(methyl/ethyl acrylate) and is identified in the USP under the general term “methacrylic acid co-

polymers.” It dissolves in an aqueous medium at pH > 6. Here we describe a systematic

investigation for the first time on the formulation development, granule production, stability

testing, and enteric coating of CIN core granules intended for releasing CIN in the lower GIT

regions to target intestinal pathogens. Finally the antimicrobial activity of CIN granules was

examined against a multidrug resistant bacterial strain E. coli K88[136] as compared to free CIN.

2.2 Materials and Methods

Essential oil compounds trans-cinnamaldehyde (CIN) (99%) and eugenol (99%), fatty acids

(FA): lauric acid (C12), myristic acid (C14), palmitic acid (C16), stearic acid (C18), and the fatty

alcohol palmitic alcohol (C16), were obtained from Sigma-Aldrich (Oakville, Ontario, Canada).

Enteric polymers Eudragit® (Evonik Industries) L 100 and S 100 are methacrylic acid-

methylmethacrylate copolymers in the ratio 1:1 and 1:2, respectively, obtained in powder form

from Almat Pharmachem Inc, (Concord, Ontario, Canada). Oil adsorbent powders Neusilin®

45

US2 and UFL2 (magnesium aluminum silicate) were bulk samples obtained from Fuji Chemical

Industries (Fuji Health Science, Burlington, New Jersey, USA), microcrystalline cellulose

Avicel® grades PH102, RC-591, CL-611 were obtained from FMC Biopolymer, and white wheat

bran powder (50 mesh) was provided by Hayhoe Mills, Ltd. (Woodbridge, Ontario, Canada).

Carvacrol, butylated hydroxy toluene (BHT), t-butyl methyl phenol were obtained from Sigma-

Aldrich chemicals (Oakville, Ontario, Canada).

2.2.1 Determination of oil adsorbing capacity of powders

Two types of pharmaceutical excipient powders, i.e., microcrystalline cellulose powder (Avicel®

PH102, Avicel® RC-591, Avicel® CL-611) and magnesium aluminum silicate powder (Neusilin®

UFL2, Neusilin® US2), and a food powder (wheat bran) were tested for their oil adsorbing

capacity. Approximately 0.5-1 g of a powder was accurately weighed into a 15 mL conical

polypropylene screw-cap tube (BD Falcon) and then CIN was added until excess was visible.

After allowing the oil to completely soak the powder for more than 10 min, the excess oil was

then separated by centrifugation for 10 min at 3000 g. The excess oil was poured off and a

pipette was used to draw any remaining superficial oil after inverting the tube and allowing

drainage. After all the excess oil was removed, the tube containing powder and adsorbed oil was

weighed again. The weight of adsorbed oil was obtained by weight difference. The oil loading

capacity was calculated by the following equation:

% w/w oil = 𝑤𝑒𝑖𝑔ℎ𝑡 𝑜𝑓 𝑜𝑖𝑙

𝑤𝑒𝑖𝑔ℎ𝑡 𝑜𝑓 𝑜𝑖𝑙+𝑤𝑒𝑖𝑔ℎ𝑡 𝑜𝑓 𝑝𝑜𝑤𝑑𝑒𝑟× 100%

46

2.2.2 Formulation selection by phase diagram

Based on the test of oil adsorbing capacity of powders described above, Neusilin® US2 powder

yielded the highest CIN oil absorbing capacity among six studied powders and was selected for

further formulation development. Oil loaded core granules were formulated with three

components: CIN oil, adsorbent powder, and fatty acid. A phase diagram was constructed to

better define composition ranges that yielded: 1) at molten temperatures, a mixture in the liquid

state for ease of droplet formation in suspension, and 2) when cooled to room temperature (23

°C), a solid with absence of excess oil on the surface. Compositions (see Supplementary Figure

S1) were tested by making 2 g of mixtures containing different weight % of each component in a

10 mL glass vial. Melting and solidification of the mixtures was done using a hot water bath and

ice water bath, respectively. The CIN oil was mixed with the powder first and then blended into

molten fatty acid with a spatula and subsequently solidified by cooling in a water bath.

Qualitative observations were made (see Supplementary Table S1) of the mixtures at different

temperatures and classified as powder, solid (homogeneous), paste, or liquid.

2.2.3 Preparation of core granules

The CIN oil was encapsulated into spherical granules by melt-solidification in an aqueous

suspension. To prepare 100 g of oil-containing granules, CIN oil (40-50 g) was first adsorbed

with Neusilin® US2 powder (US2) (0-13 g) by slow addition and mixed gently in a beaker to

ensure even distribution of oil. In a separate beaker, the fatty acid was melted in a water bath set

at 45 to 67 °C (i.e. up to 5°C higher than the melting point of the fatty acid chosen). The oil-

powder mixture was then added to the molten fatty acid and mixed with a spatula until

homogeneous. In order to stabilize the molten droplets, an aqueous solution of methacrylic acid-

methyl methacrylate copolymer (Eudragit® S 100) neutralized with NaOH to pH 7.5 was used at

47

a final concentration of 4-5% w/w. The aqueous dispersion medium, (200-300 mL) was heated

to the same temperature as the molten fatty acid (51-52 °C for LA 66-67°C for PA) in a 600 mL

beaker with a water bath to ensure the CIN-US2-FA mixture remained molten when dispersed.

Tween 80 (at 0.5% v/v of the aqueous phase) was found to slightly improve the round shape of

granules. An overhead stirrer (Caframo, Ontario, Canada) was used to disperse the oily phase

into the aqueous phase at 150-375 rpm using a stainless steel 4-bladed round edged propeller to

form a two-phase, liquid-liquid suspension. After droplet size and shape were visually

acceptable (~30 sec – 1 min) stirring was stopped and the beaker was transferred to an ice-water

bath to solidify the droplets into granules. Afterwards, granules were suction filtered and washed

with small portions (10 mL) of distilled water, and then allowed to air dry on aluminum pans for

45 min or until free-flowing. For palmitic acid (PA) granules, higher initial temperature was

used (67-70°C) to maintain molten state before blending with oil-soaked powder. Figure 2.1

summarizes the steps in the granule production process. Control granules were made with

10%US2:90%FA by physically breaking down the re-solidified mixture into small sized particles.

48

Figure 2.1. Schematic diagram for granule production process

2.2.4 Enteric coating of core granules

Core granules were first sub-coated with an ethanolic dispersion of Kollicoat IR (polyvinyl

alcohol-polyethylene glycol graft co-polymer) to resist further evaporation of oil by acting as a

sealant layer. The outer enteric polymer coating was applied using a fluid bed machine (Glatt

laboratory fluid bed, Germany) in 100-500 g batches of core granules. The polymer for enteric

coating was Eudragit® L 100 which dissolves above pH 6 giving targeted release in the jejunum-

ileum regions of the intestinal tract.

2.2.5 Assay of cinnamaldehyde and cinnamic acid content in granules

Granules were weighed (50-100 mg) and methanol (10-20 mL) was added into glass vials. A

sonication bath and shaking was used to disintegrate granules to fully release CIN. Triplicate

samples from individual batches were taken for determining loading and stability of CIN

contained in individual batches. Due to the oxidation of CIN (λmax 291 nm) to cinnamic acid

(λmax 272 nm) a mixture of CIN and CA was present in some cases. Thus calibration equations

49

of the two standards were obtained with both pure compounds and mixtures and then used in

determining unknown concentrations of both compounds simultaneously using a simple additive

rule. [137] Readings were obtained on a UV-vis spectrophotometer (Lambda, Perkin-Elmer,

USA).

The EO loading efficiency in prepared granules was calculated in the following manner:

Theoretical loading = 𝑤𝑡.𝑜𝑓 𝐶𝐼𝑁 𝑎𝑑𝑑𝑒𝑑

𝑡𝑜𝑡𝑎𝑙 𝑤𝑡. 𝑜𝑓 𝑎𝑙𝑙 𝑔𝑟𝑎𝑛𝑢𝑙𝑒 𝑚𝑎𝑡𝑒𝑟𝑖𝑎𝑙𝑠∗ 100%

Actual loading = 𝑤𝑡.𝑜𝑓 𝑡𝑜𝑡𝑎𝑙 𝐶𝐼𝑁 𝑒𝑥𝑡𝑟𝑎𝑐𝑡𝑒𝑑

𝑡𝑜𝑡𝑎𝑙 𝑤𝑡.𝑜𝑓 𝑔𝑟𝑎𝑛𝑢𝑙𝑒𝑠∗ 100%

Loading efficiency = 𝑎𝑐𝑡𝑢𝑎𝑙 𝑙𝑎𝑜𝑑𝑖𝑛𝑔 𝑜𝑓 𝐶𝐼𝑁 𝑖𝑛 𝑔𝑟𝑎𝑛𝑢𝑙𝑒𝑠

𝑡ℎ𝑒𝑜𝑟𝑒𝑡𝑖𝑐𝑎𝑙 𝑙𝑜𝑎𝑑𝑖𝑛𝑔 𝑜𝑓 𝐶𝐼𝑁 𝑔𝑟𝑎𝑛𝑢𝑙𝑒𝑠∗ 100 %

2.2.6 In vitro release of CIN from granules

Simulated gastric fluid (SGF, adjusted with 0.1 M HCl, 0.2%w/v NaCl) and simulated intestinal

fluid (SIF, 50mM KH2PO4, adjusted with 0.1 M NaOH) were prepared at pH of 1.2 and 6.8,

respectively. Release of CIN was monitored by UV detector at two wavelengths (291 and 272

nm) at 10 min intervals via flow-through cuvettes with up to 6 replicates per run. Granules were

tested for 2 h in SGF (~250 mL) and 6 h in SIF (~900 mL) and the CIN concentrations were

converted into amounts to calculate % released. Solutions were maintained at 37 °C and stirring

rate was 100 rpm using a dissolution testing apparatus (Erweka, Germany).

2.2.7 Antimicrobial activity assay with pure culture in liquid growth media

Antimicrobial activity of pure compound CIN, its oxidation product CA, and CIN-encapsulated

core granules was assessed. A study of the growth inhibition by EO compounds was performed

against a pure culture of E. coli K88 strain JG280 undergoing exponential growth at 37 °C in

50

tryptic soy broth (TSB) media as described previously[41]. The three treatments were: control

(no EO compounds), free EO, and EO granules at different EO concentrations in 5 mL of TSB

inoculated with E. coli K88 at a concentration of 104 CFU/mL. The bacteria were then allowed

to grow with agitation at 200 rpm on an orbital shaker-incubator at 37°C. The optical density

(OD) of the suspensions were monitored at a wavelength of 600 nm in comparison with the

control OD [62]. The experiment was stopped when the control OD reached 1.2-1.3 or typically

between 5-6 h of incubation. The MIC90 of CIN against this E. coli K88 strain was defined as

the lowest concentration of the EO compound that produced a 90% inhibition of the bacteria

undergoing log growth. [41, 138]

𝐼𝑛ℎ𝑖𝑏𝑖𝑡𝑖𝑜𝑛 % = 100 ∗ (𝑂𝐷𝐶𝑜𝑛𝑡 − 𝑂𝐷𝐸𝑂

𝑂𝐷𝐶𝑜𝑛𝑡 )

ODCont represents the OD600 of a bacterial culture grown with the same initial 104 CFU/mL, in

the absence of any EO (negative control). The minimum bactericidal concentration (MBC) of

CIN against this E. coli K88 strain was taken as the lowest concentration that produced no

observable growth after transferring to fresh growth media for up to 6 h, followed by plating

dilutions (20-100 µL) onto TSA agar plates incubated overnight (16-24 h) to identify the number

of surviving colony forming units (CFU/mL). These experiments were repeated at least twice.

2.3 Results

2.3.1 Formulation and properties of CIN core granules

2.3.1.1 Selection of adsorbent powder with highest capacity

Oil retention capacity of three types of pharmaceutical or food-grade powders was evaluated to

find a suitable powder with the highest oil adsorption. Table 2.1 compares the oil retention

51

capacity of these powders. Among the tested powders, Neusilin® US2 shows the highest

adsorbing capacity of 84%, attributable to its high specific surface area (Table 2.1), and was thus

selected for further study. Although the powders could hold a large amount of oil, once above

~78% w/w the powder surface appeared wet with oil due to powder saturation. Furthermore,

powders containing oils were fairly soft and weak, and appeared almost like a wet paste (Figure

2.2), so a third component was added to allow formation into granules, that being a meltable fatty

acid binder.

Table 2.1. The oil adsorbing capacity of various powders. Values shown as AVE ± SD (n ≥ 3)

Powder type Grade Oil adsorbing capacity

% w/w CIN Specific surface* (m2/g)

Avicel® PH102 1.532 ± 0.083 61.66 ± 0.99% 1-1.35

Avicel® RC-591 0.609 ± 0.022 39.00 ± 0.84% -

Avicel® CL-611 0.676 ± 0.010 41.52 ± 0.34% -

Neusilin® UFL2 4.023 ± 0.018 80.86 ± 0.07% 300-339

Neusilin® US2 5.074 ± 0.107 84.19 ± 0.28% 300-339

Wheat bran - 0.970 ± 0.039 50.4 ± 1.00% -

*values obtained from[139, 140] and product bulletin

52

Figure 2.2. Pictures of Neusilin US2 at various CIN oil loading: top 0%, middle 62%, and

bottom 78%

53

2.3.1.2 Selection of fatty acid binder to obtain sufficiently strong granules

Saturated fatty acids were chosen as the binder for their well-defined melting point to allow

quick encapsulation by rapid cooling and their lipophilic property. Lauric acid (LA) was initially

chosen for its low melting point at ~44 C and being easily re-solidified at room temperature.

When molten, LA blended easily with the oil soaked powder and when cooled and dried

overnight, spherical granules were obtained. However, LA-based granules were mechanically

too weak to withstand further processing in a fluid bed coating machine due to problems such as

attrition and fusing of the granules. In order to address these issues two approaches were

undertaken: blending with higher melting temperature fatty acids or fully substituting with

another fatty acid type. It was found that blending of different fatty acids yielded mixtures that

were more difficult to solidify than pure fatty acid. After experimenting further with other

excipients it was concluded that the granule strength was greatest when a single fatty acid was

used instead of mixtures.

2.3.1.3 Optimizing formulation of core granules via phase diagram

A phase diagram was used as an aid to optimizing oil loading in the final mixture and to find

compositions that would allow uniform, completely re-solidified granules that were suitable for

further processing, as shown in Figure 2.3. The phase regions/boundaries were drawn as a result

of the processabilty limits of the mixtures (see supplementary Table and Figure S1 for actual

composition points tested). There were four types of phase state that were observed, based on

the physical processability of the mixture at molten temperatures or at room temperature (after

cooling and/or re-solidification). Regions of phase diagram were classified as:

54

Figure 2.3. Phase diagram for a 3-component mixture at molten (fatty acid) and cooled

temperatures. Dot indicates optimized formulation. Outline indicates range of formulation

components that resulted in acceptable granules.

Powder – the mixture was loose and powdery due to under-saturation of the US2 and

insufficient inter-powder binding. At molten temperature, the mixture was not dispersible in the

aqueous continuous phase unless further liquid adsorption occurred. After cooling down, there

was insufficient inter-powder binding by the fatty acid component to hold mixture together.

Paste – mixture appeared wet, weakly held together and not free-flowing, containing over-

saturation of the US2 powder by liquid (CIN and fatty acid) at molten temperatures. After

mixture cooled an excess of liquid CIN remained and weakened binding due to insufficient

amount of fatty acid.

Liquid – mixture was liquid-like, i.e. free-flowing under its own weight. Mixture was easy to

disperse into uniform droplets at above molten temperature.

55

Solid – sufficient amount of re-solidified fatty acid to hold mixture together with little to no

excess oil visible. Upon cooling, granules became uniformly solid; when dry were free flowing

and round in shape due to good dispersability as droplets in the molten state.

For ease of dispersibility into small droplets and forming granules, a molten mixture with liquid

behaviour was preferred, corresponding to >80% liquid and <16% powder. When the liquid

saturation of US2 was approached or exceeded, the mixture became more easily dispersible into

droplets. But in order for re-solidification of granules to occur upon cooling, sufficient fatty acid

needed to be present in the mixture to bind the powder together upon re-solidification. The

region where a paste occurred was between 15-20% US2 and 80-84% liquid and such mixtures

were not processable into droplets and resulted in large globs being formed, even under high

speed stirring (>1000 rpm). The powder region (blue) was where there was ≤80% liquid and

≥20% powder, and due to under-saturation of the powder, the mixture was not dispersible within

a liquid continuous phase, which was unfavourable for forming CIN-containing droplets. The

region of the phase diagram with maximal yield of granules with high EO loading is outlined in

red. The type of fatty acid used to make granules did not affect the oil loading properties (i.e. a

similar phase diagram could be used for the other FAs). It was observed that use of a higher

melting temperature fatty acid resulted in harder granules that could be felt by grinding the

particles between fingers, since neither the oil nor the powder possessed appreciable binding

ability and hence the granule solidification was essentially resulting from re-solidification of the

fatty acid component.

2.3.1.4 Process conditions and properties of core granules

Table 2.2 presents the process and granule characteristics of batches prepared under a range of

conditions. From these studies optimized conditions were found and then used to produce 0.8-1

56

kg batch granules for subsequent coating. The shape and size of granules were found to be

highly affected by the fluid properties of the molten mixture. Cooling of the mixture after

stirring stopped yield of more spherical granules while addition of cooling water into the

suspension disrupted droplet shape during cooling.

Table 2.2. Granule properties and process characteristics

Oil loading, % w/w (actual/theoretical)

16.9-45%/25-48%

Oil loading/encapsulation efficiency

~93%

RPM range 150-375

Granule batch size range 0.5-5 mm

Batch size 20-200 g

Process yield >90%

molten:aqueous phase ratio 0.13-0.66

Formulation components ranges:

CIN 20-50% US2 0-10% PA 40-60%

Optimum ratio of CIN:US2:PA 48:10:42

The sieved size distribution of finished batches prepared at fixed conditions (300 rpm, 150 g

batch, 300 mL continuous phase in 600 mL vessel) but different formulation composition is

shown in Figure 2.4. It was found that when the US2 remained constant (10%), increase in PA

concentration from 40-45% produced higher proportion of larger granules; similarly, when the

CIN concentration kept constant (45%), increasing the US2 concentration from 10-13%

produced higher fractions of larger granules, especially those over 1400 μm. The encapsulation

efficiency was similar among the formulations tested (~93%). The formulation

48%CIN:10%US2:42%PA yielded hard granules, with a high loading efficiency and had the

57

highest weight fraction of granules in the range of 600-1400 μm, which are suitable for

processing in fluid bed coating and is in a product form convenient for mixing with animal feed.

Thus this formulation was selected for further study. Formulations with more than 10% US2

produced much larger fractions of granules over 1400 μm due to the apparent viscosity

enhancing effect of US2 on the molten mixture (and other suspensions in general).

Figure 2.4. Effect of formulation composition changes on the size distribution of granules

while keeping batch size, stirring rate constant.

2.3.2 Selection of antioxidant to stabilize CIN in the core granules against atmospheric

oxidation (autoxidation)

After incorporating pure CIN in initial batches of granules, the EO was unstable as it underwent

autoxidation into cinnamic acid (CA) fairly rapidly upon exposure to the atmosphere which led

to loss of antimicrobial activity as explained in the following section. As shown in Figure 2.5a

the rapid conversion of CIN into CA started as soon as after leaving granules exposed to

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

50:10:40 48:10:42 45:10:45 45:12:43 45:13:42

Wei

ght

frac

tio

n o

f to

tal

Weight % CIN:US2:PA in formulation

>1400 um

1000-1400

850-1000

600-850

<600 um

58

atmospheric oxygen overnight. Without any antioxidant protection, more than 70% CIN (of the

initial loading) degraded into CA within 5 days at room temperature and within 30 days at 4 ºC;

the rate of decomposition was slower (Figure 2.5a). To counteract this undesirable degradation,

a series of potential anti-oxidative compounds were screened. Among the tested compounds

(BHT, carvacrol, eugenol, and t-butyl methyl phenol), it was found that proton-donating

compounds (phenols) were the most effective in slowing down/inhibiting autoxidation. Eugenol

was selected for further study because of its natural presence in cinnamon oils, primarily in the

plants leaves[141, 142]. It was found that a concentration of 1 % v/v eugenol added to pure CIN

was sufficient for protection against CIN autoxidation in granules for at least a year at both

temperatures. After addition of the antioxidant to CIN, granules made with PA and LA showed

the same degree of stability. Coated granules containing CIN with 1% eugenol were also stable

for at least a year when stored either at room temperature or refrigerator temperatures as seen in

Figure 2.5b.

59

Figure 2.5. (a) Stability of CIN in uncoated granules with and without antioxidant when

stored at 23 and 4 °C. (b) Stability of coated granules with antioxidant (1% EUG) stored at

23 and 4 °C for up to 1 year

2.3.3 Properties of coated granules

Two types of core granules produced from LA and PA were coated by a fluid bed process. The

first type of granules made using LA are shown in Figure 2.6a. The CIN in these granules did

60

not contain an antioxidant so the CIN content was low in the final product (~6% w/w) and

granules were solid enough to be coated due to conversion to the solid CA. However, after

inclusion of an antioxidant (1% eugenol) in subsequent batches of LA granules, the granules

were not strong enough to withstand the fluid-bed coating process so higher melting temperature

fatty acids were investigated and used. The stronger granules based on PA were thus produced

and coated. The core granules contained 42% CIN and the coated granules contained 20% CIN

(with 1% eugenol) due to a weight gain of ~30-35% from coating. Both types of granules were

fairly round and the coating uniform as shown in Figure 2.6.

a

b

Figure 2.6. (a) Lauric acid granules uncoated (left) and after coating (right) and (b)

palmitic acid granules with 42% w/w CIN, coated with Eudragit L100 and sub-coat with

Kollicoat® IR

61

2.3.4 In vitro release of CIN from core granules and coated granules

Figure 2.7a compares release profiles of CIN from uncoated core granules prepared with

different fats in SIF at pH 6.8 at 37 ºC. LA based granules showed the highest rate and extent of

release, followed by myristic acid (MA), palmitic acid (PA), stearic (SA) and cetyl alcohol (CAl).

Without coating CIN was released rapidly from granules with over 80% being released from LA,

MA, and PA-based granules within 1 h. SA and CAl-based granules showed lower extent of

release probably attributable to their higher hydrophobicity.

a

b

Figure 2.7. (a) Release of CIN in SIF (pH 6.8) from core granules formulated with different

fatty acid types and release of CIN from coated granules under two-stage dissolution.

Values represent AVE±SD

62

Core granules were coated with an enteric polymer (Eudragit® L 100) to supress release of CIN

in gastric conditions and enable release after dissolution of the coating polymer at pH>6. The in

vitro release profiles were determined in a two stage test: 2 h in SGF at pH 1.2 and then

continued in SIF at pH 6.8 for 3 h. For the uncoated granules, the release test was undertaken in

SIF. Figure 2.7b shows pH-dependent release of CIN from the coated granules which is absent

for the uncoated core granules. The enteric coating prevented CIN from release in the SGF

significantly with >70% of the loaded CIN being released in the SIF. Ultimately, after

dissolution of the coating material over 90% of the CIN was released.

2.3.5 Antimicrobial activity of CIN against E. coli K88: MIC and MBC determination

The effectiveness of antimicrobials against bacteria growth/viability can be generally expressed

in two forms: MIC90 and MBC. Inhibition of growth was determined for different concentrations

of CIN and its oxidation product cinnamic acid (CA). The MIC90 of CIN against E. coli K88

was found to be 150 µg/mL while CA showed much lower inhibitory activity (>1000 µg/mL).

Hence, no MBC determination for CA was carried out due to the low inhibitory activity against

the target strain as shown in Figure 2.8a. Figure 2.8a shows the inhibition of E. coli K88 growth

(initial 104 CFU/mL) by CIN at various concentrations in a liquid medium (TSB). The low

activity of CA suggested preventative measures against oxidation of CIN to CA using

antioxidant. From the curves in Figure 2.8b the MBC of free CIN oil and CIN encapsulated into

PA- and LA-granules against E. coli K88 were found to be 600, 550, and 450 μg/mL,

respectively.

63

a

b

Figure 2.8. (a) Inhibitory activity of CIN vs CA towards E. coli K88 grown in TSB culture

based on change in OD600 over 6 hrs @ 37° C with 200 rpm shaking. (b) Antibacterial

activity of CIN oil and granules against E. coli K88 in TSB medium. Data points represent

AVE±SD

64

2.4 Discussion

2.4.1 Importance of formulation on the properties of granules

The current results indicate that the melting point of the specific fatty acid used predominately

affected granule strength because granules were primarily held together by intermolecular van

der Waals forces between fatty acid molecules. As fatty acid chain length increases, its melting

temperature increases, and granules become harder at room temperature as compared to those

made from a shorter chain fatty acid. As such, palmitic acid (PA) yielded harder granules than

LA due to its higher melting temperature resulting from the longer fatty acid chain length (C16

vs. C12). Thus, the formulation with LA was substituted with PA, to obtain harder granules at

room temperature. Fatty acid blends were not used for making stronger granules because blends

did not have a well-defined melting transition temperature compared to pure fatty acids, possibly

due to less effective packing between the dissimilar chain length fatty acid molecules (e.g. LA

with SA). With regard to the amount of CIN that could be accommodated in the granules, the

independence of fatty acid type could be due to similar volume being occupied by these different

fatty acids in both molten and solid states. Structurally, the different fatty acids LA, MA, PA,

SA (12, 14, 16, and 18 C) and fatty alcohol CAl (C16) differ only by a few (-CH2-) groups.

When granules were made with two components, i.e. oil and fatty acid, the loading maximum

was about 28% w/w for LA, MA, PA, and SA. The CAl on the other hand, allowed loading up

to 40%, without excess oil being visible after cooling. Including Neusilin® US2 in the

formulation allowed increased oil loading levels up to 48% w/w, due to its high oil adsorbing

capacity.

Constructing the three-component phase diagram allowed us to define optimum ratios of the

mixture components for production of core granules with high oil loading, good dispersibility

65

(affected the size and shape of granules) at molten temperatures and good mechanical strength

(hardness) after solidification, which was necessary to endure subsequent processing steps. This

analysis led to an optimal weight ratio of each component of CIN:US2:PA at 48:10:42, and this

formulation was then used for scale up production. Granules were coated with Eudragit® L 100

to render pH dependent release behaviour with fast release at pH > 6 at the jejunum, ileum and

cecum sections of the intestines. For some applications, though, enteric coating may not be

required where a slow release into the environment is more desirable to reduce pathogens on the

feedlot floor surface. [143]

The behavior of the mixtures could be explained by the saturation point of the Neusilin® powder

component. As shown in Figure 2.2, Neusilin® powder became saturated around 78-80% w/w

liquid. Below this saturation point, the mixture still exists as a powder irrespective of the

temperature, hence cannot be dispersed well in an aqueous continuous phase without taking up

water in order to become dispersible (not processable). Thus, when at molten temperatures, the

three-component mixture simplifies to a two-phase, liquid (fatty acid and EO) and solid

(Neusilin US2) mixture and the level of powder saturation becomes dominant in affecting the

properties of the mixture. Figure 2.3 shows that at molten temperatures the fatty acid and CIN

are both miscible and they penetrated the powder as one liquid and depending on the extent of

US2 saturation, this mixture behaved as a liquid when near or above saturation of the powder

(~84% liquid:~16% powder).

It was observed that the granule formulation affected the size distribution primarily by the degree

of liquid saturation of the Neusilin® US2 powder (Figure 2.4). Granules became larger as the

amount of US2 was increased from 10 to 13% while the degree of powder saturation decreased.

As the level of available liquid (CIN + molten fatty acid) saturates the available powder (e.g.

66

>48% CIN, US2 ≤10%), smaller droplets were formed which led to smaller granules as the

molten mixture behaved as a low viscosity liquid. In contrast, as the saturation level of US2 is

decreased, or if the US2 > 10%, the powder exerts a viscosity enhancing effect on the molten

liquid mixture, leading to larger droplets and granules. This effect is one of the properties of

magnesium aluminum silicates (MAS) and other silicates widely used in cosmetics[144]. But if

the powder was oversaturated with oil, this effect appeared less influential in the present method.

After cooling to room temperature, the fatty acid re-solidifies while CIN remained liquid since

no chemical reactions among the components were expected or observed. Due to displacement

of CIN by molten FA during mixing with US2, the maximum loading in the cooled state was

about 48-50% or less than that without FA. Incorporating more than 50% CIN initially led to

excess superficial CIN appearing and a paste or liquid was obtained at room temperature instead

of granules. The loading limit was thus determined to be ≤ 50% CIN in the final formulation of

granules.

PA was chosen in formulating larger batches of granules for offering the best combination of

higher granules strength and fast, more complete release of the encapsulated CIN, in addition to

the lower temperature required for preparation vs SA. Thus, using a higher melting fatty acid like

PA yielded harder granules than LA and MA-based granules yet released the encapsulated oil

better than SA or CAl. However, fatty alcohol based granules resulted in decreased extent of

release possibly due to the less polar nature of the fatty alcohol compared with fatty acids.

Coated granules exhibited only partial resistance to release in SGF with around 27% of the CIN

being released from coated granules. The early release of the active in SGF could have been

caused by the migration of the volatile liquid oil from the cores to the coating layers during the

fluid bed process. The long coating times typically used (>6 h) as well as the complex processes

67

occurring during film formation[145] could have contributed to this, as well as loss of the active

as the constant fluidization by air could have drawn away some of the volatile CIN from the

cores. Further optimization of the fluid bed coating process or selecting an alternative coating

process[146] (e.g. dry coating) could be evaluated in future experiments.

2.4.2 Antibacterial activity of and storage stability of CIN granules

E. coli K88 is an entero-toxigenic strain (ETEC) of E. coli that commonly causes piglet diarrhea

during the post-weaning period when animals are most susceptible to infection.[147] This

particular strain of E. coli produces K88 pili that adheres at the jejunum and ileum epithelial

regions of the pig intestine, where specific receptor sites are expressed at the mucous

surface,[148] and then produce toxins that cause tissue damage and diarrhea. Furthermore,

compared to other strains of E. coli, K88 is resistant to a spectrum of antibiotics but was found to

be effectively inhibited by EOs like CIN.[41] Even though CA had little inhibitory activity

against E. coli K88, it has good inhibitory activity against a variety of fungi[149] and spoilage

bacteria[150] with MICs in the 1-5 mM (150-750 µg/mL) range. LA-based granules showed an

overall better killing effect with a lower MBC occurring at 450 µg/mL compared to the MBC of

600 µg/mL for free CIN oil (Figure 2.8b), while control granules containing only excipient (US2

and either LA or PA at the equivalent CIN concentrations ~450-1000 μg/mL) did not show

significant inhibitory activity against E. coli K88, suggesting that a synergistic activity of the

LA-CIN mixture might be happening. This result could be attributed to the antibacterial effect of

fatty acids have that were used in the granule formulation[151] in addition to the melting point

depression effect after mixing with the CIN oil, which allowed the granules to melt and

disintegrate at the incubation temperature of 37 °C (see Figure S1), allowing dispersal of LA and

CIN within the test medium. Further study into this effect could be performed, but due to

68

difficulty in coating LA granules, PA-based granules were chosen for subsequent development.

PA-based granules had the same antibacterial activity as the pure oil and did not exhibit an

enhanced effect as LA due to its higher melting temperature and lower solubility. The MBC of

CIN obtained in the current study was about 2-6 times higher than previous reports.[41, 126]

This difference could be attributed to variations in some minor essential oil components normally

found in EO products from different sources and furthermore, EO mixtures can exhibit stronger

antibacterial activity than the respective single EO component.[152] By using the current

strategy of encapsulating CIN (or other EOs) into granules and coating with an enteric polymer,

a sufficient concentration (MIC to MBC range) of EO can potentially be delivered to the lower

intestinal tract of pigs (site of disease) while minimizing or avoiding the early absorption of EO

compounds in the upper GI tract.[58] Such granule formulations could be an effective strategy,

without using antibiotics, to combat piglet post-weaning diarrhea caused by E. coli K88.

Since the antibacterial activity of CA was determined to be much less than CIN against the target

pathogen, it is necessary to prevent autoxidation of CIN in granules. This work has

demonstrated that with addition of antioxidant 1% eugenol to CIN, granules made with PA and

LA showed the same degree of stability, with or without coating, which can last for at least a

year at room temperature or refrigerator temperatures (Figure 2.5b). These positive results

prompted the exploration of in vivo performance of the coated CIN granules in a subsequent

study which will appear in another publication.[153]

2.5 Conclusions

Granular formulations of CIN, a model EO compound with good antimicrobial activity, were

successfully developed with loadings up to 50% w/w. The core granules were prepared by a

melt-dispersion-solidification method using rationally selected oil absorbing powder Neusilin®

69

US2 and lipids at an optimized ratio of CIN:US2:lipid. The enteric polymer coated granules

based on PA exhibited fast release of CIN at intestinal pH, suitable for site-specific delivery of

CIN in the lower intestines where E. coli reside most. The addition of antioxidant eugenol to the

granule formulation significantly increased CIN stability to at least one year. The LA-

encapsulated CIN was found more effective in vitro against E. coli K88 than free CIN with about

33% lower MBC but were unsuccessfully coated by fluid bed, while PA-based granules which

were coated successfully, exhibited a MBC similar to the free oil. The in vitro results obtained

here indicate the potential of CIN containing granules in combating E. coli K88 and other

susceptible pathogens. Further in-vivo studies are warranted to verify the effectiveness of CIN

granules. The formulations and method developed in this work for encapsulation of oils in solid

granules are relatively simple and economical and can potentially be used to encapsulate a

variety of pure essential oils, their mixtures or other lipophilic liquid actives.

2.6 Acknowledgements

The authors would like to thank the Natural Science and Engineering Research Council and

Agriculture Agri-food Canada, and the Canadian Poultry Research Council for funding the

research, and the Toronto Institute of Pharmaceutical Technology for carrying out the coating of

granules with enteric polymers, and the University of Toronto Open Scholarship to YM. E. coli

K88 strain JG280 was a gift from Dr. C. Gyles, University of Guelph.

70

Chapter 3. In vivo performance of essential oil granules for targeted delivery to the animal

GI tract

Yin-Hing Ma1,2, Marta Hernandez1, Qi Wang1, Joshua Gong1, Xiao Yu Wu2



2Graduate Department of Pharmaceutical Sciences, University of Toronto, 144 College St.,

Toronto, Ontario, Canada M5S 3M2

(All experimental work was carried out by YHM under the supervision of QW, JG and XYW.

The GC method development was performed by MH)

71

Abstract

The following study evaluated the in vivo release and distribution of an essential oil (EO)

compound trans-cinnamaldehyde (CIN) following oral administration in two forms: as CIN oil

and as coated granules loaded with 20%w/w CIN. The first study was performed in 2-week old

broiler chickens, given a dose of 2500 μg CIN/g feed by gavage. Coated CIN granules were able

to deliver higher average and peak concentrations to the lower gastrointestinal tract sections than

free oil at the same concentration. However, even after 3 hours significant portions of the dose

in some animals were still present in the upper alimentary tract of the chickens, namely the crop

and gizzard indicating incomplete passage of the dose within 3 hours. The second study was

performed in 3-week old pigs. After 7 days of acclimatization with the EO in the feed ad libitum,

the next day a dose was given at 1500 μg/g feed and pigs had access to the feed mixture and

water for 5 hours ad libitum. Quantification of the CIN content in the digesta showed that in

individual pigs, higher peak concentrations were observed when given the coated CIN granules

than when given CIN oil added to feed at the same concentration however, due to the high

variability between animals, group averages were not statistically different. The current study

provides evidence that formulation of essential oil compounds using a coated granule approach

can achieve more effective delivery of an EO compound closer to the distal intestinal tract

regions to target the reservoir of antibiotic resistant pathogens.

(keywords: trans-cinnamaldehyde, antimicrobials, antibiotic resistance, enteric coating,

antibiotic alternatives, pig, swine, chicken, poultry, livestock)

72

3.1 Introduction

Many plant-derived essential oils (EOs) have strong antimicrobial activity in-vitro [41, 43, 154].

However, direct addition of free EO to the feed has limited effectiveness due to several physical-

chemical factors such as high oil hydrophobicity, volatility, odour, oxygen sensitivity, and low

availability in the animal GI tract [58]. Trans-cinnamaldehyde (CIN) is a major component of

cinnamon essential oil and shows antimicrobial activity against Escherichia. coli K88, a

pathogen causing severe piglet diarrhea, in addition to other pathogens [63, 155]. CIN was

found to inhibit growth completely at 150 and bactericidal at 600 µg/mL in a liquid medium

[156]. Conventional direct addition of feed additives to feed of animals is possible for EOs but

they are volatile, strongly odorous and may be prone to degradation when unprotected [58, 157].

In addition, EO compounds were found to be rapidly absorbed in the upper, proximal GI tract

following administration [58]. Therefore, to overcome these limitations, a stable and easy-to-

handle solid granule formulation was developed to overcome these hurdles in EO application as

antibiotic alternatives for use in livestock farming. The objective of antimicrobial use is to

achieve the minimum inhibitory concentration (MIC) or higher at the location of pathogen

activity, and although higher concentrations are desired to maximize the antagonistic effects

against the target pathogen, too high concentrations could cause toxic effects to the host cells.

Due to the rapid absorption of the hydrophobic EO compounds in the upper GI tract, it is very

unlikely that free oil added to feed at moderate amounts will reach the lower GI tract in

sufficiently high concentrations [58]. A coated granular formulation of EO was described

previously [156] and the current paper continues on by evaluating the in vivo performance of the

coated granules to assesses the ability of coated CIN granules to target the delivery to the lower

regions of the GI tract of animals as compared to direct addition of EO compounds to the animal

feed.

73

3.2 Materials and Methods

Essential oil compounds trans-cinnamaldehyde (CIN) (99%) and eugenol (99%), palmitic acid

(C16), were all obtained from Sigma-Aldrich (Oakville, Ontario). Enteric polymers Eudragit

(Evonik Industries, Germany) L100 and S100 grades are methacrylic acid-methylmethacrylate

copolymers in the ratio 1:1 and 1:2, respectively, obtained in powder form from Almat

Pharmachem Inc., (Concord, Ontario, Canada). Neusilin® US2 (magnesium aluminum silicate)

was obtained from Fuji Chemical Industries as product samples (Fuji Health Science, Burlington,

New Jersey, USA). The animal studies involving chickens and pigs received prior experimental

design approval by the Animal Care Committee of the University of Guelph prior to being

carried out.

3.2.1 Preparation of core granules and enteric coating

Essential oils were encapsulated into spherical granules by melt-solidification in an aqueous

suspension with stirring as described previously [156]. Core granules were made of the palmitic

acid type and contained 46% w/w CIN prior to coating. After enteric coating by fluid bed

apparatus was carried out at the Toronto Institute of Pharmaceutical Technology, the loading of

CIN was 20% w/w in the final granules.

3.2.2 Oral administration to 2 week old chicks

Fifty newly-hatched broiler chicks were obtained from a local chicken hatchery (Maple Leaf

Poultry, New Hamburg, Ontario) and housed at the University of Guelph, Arkell Poultry

Research Station (Guelph, Ontario, Canada) in floor pens covered with wood shavings, and

given access to water and feed ad libitum. After reaching 2-weeks (14 days) of age, the chickens

74

were divided into two groups of 25, one group receiving free oil and the other group receiving

coated granules. In the free oil group, each animal received 0.5 g of feed containing (directly

added) 42.5 mg (or 40.5 µL) of CIN (w/1% eugenol as antioxidant) by gavage via 1 ml syringe at

8:30 am in the morning. In the coated granules group, each animal received 213 mg coated

granules (20% CIN loading) mixed with 0.25 g feed and administered by gavage using a 1 mL

syringe. The amount of dry feed estimated to be consumed during the experiment of 3 hours was

17 grams, therefore the CIN concentration based on this theoretical weight of feed was 2500

µg/g (ie. 425 mg/17 g feed).

After each time point was reached (1, 1.5, 2, 2.5, 3 hours) 5 chickens were randomly selected

from each group and euthanized by cervical dislocation. Immediately after, the chickens were

dissected and the intestinal contents collected into 5 ml screw cap polypropylene tubes and then

placed on ice. Once all samples were collected, they were brought back and stored at -20 °C

frozen until ready for sample analysis. The experimental summary is shown in Figure 3.1.

75

Figure 3.1 . Schematic of chicken trial

3.2.3 Pig trial

Twelve, 3-week old Yorkshire pure-bred pigs (average 12 kg weight) were divided into 6 pens of

two animals. Each pen housed one male and one female animal and was allowed access to water

ad libitum through nipple drinkers in each pen. Feed was placed in the troughs twice daily (8 am

and 4 pm) averaging 1 kg per meal, depending on how well the previous days’ feed was finished

(i.e. average daily feed intake). Acclimatization to the EO additive for 7 days consisted of

regular feed with added free oil or coated granules at a concentration of 1500 µg/g of dry feed.

For every 1 kg of dry feed (from Arkell feed mill), 1.5 g of CIN oil was mixed slowly by using a

76

dropper into the entire feed meal, followed by shaking in a closed glass jar for 30 seconds to help

distribute the oil throughout the feed. On the sampling day, the animals were fed once in the

morning (8 am) with ~1 kg feed containing free CIN or coated granules and then were

euthanized 5 hours afterwards (starting at 1 pm) by pentobarbital injection. Pigs were

immediately dissected and the digesta were collected into multiple 50 ml tubes from the stomach,

duodenum, jejunum, ileum, cecum, and colon of each animal and placed on ice for transport. All

individual digesta samples were stored frozen at -20 °C until ready for sample analysis.

3.2.4 CIN determination in digesta

To determine the CIN concentration in the digesta contents from the various GI sections, they

were first defrosted by placing in a warm water (30 °C) bath for 30-45 mins. Once defrosted,

digesta samples were vortexed and immediately weighed into 5 mL polypropylene screw cap

tubes (~1-4 g digesta). A weighed amount of ethyl acetate was added ( a 1:1 weight ratio) to the

sample to extract the CIN by shaking vigorously for 45 mins with a multi-tube vortex mixer,

followed by centrifugation (4000X g) for 15 mins to allow clean separation of the organic phase.

Aliquots from the top organic layer were transferred to individual glass gas chromatography (GC)

vials and further diluted with 10X dilutions with ethyl acetate, if needed, before being run on the

GC instrument.

3.2.5 Gas chromatography

The GC-FID instrument was an Agilent 6890 Series equipped with a Supelcowax 10 (Sigma-

Aldrich, Oakville, Ontario) fused silica capillary column (30m X 0.25mm X 0.25µm film

77

thickness) and running the software GC Chemstation Rev. A 10.01.1635 by Agilent

Technologies. The sample run program consisted of a ramp from 60-240 ºC at 8 ºC/min heating

rate with split-less injection and helium as carrier gas, the inlet temperature was 250ºC, with a

total run time of 24.5 min. To purge the column of contaminants a cleaning program was run in

between every 2-3 samples, by injection of solvents only while the temperature was held at 220

ºC for a 30 min run. Injection volumes were 1-5 uL and the retention time of the CIN peak

occurred at 15.8 mins under these conditions. Five standard solutions were prepared in ethyl

acetate and were run before and after samples regularly (intra-day) to verify the peak retention

time consistency. The calibration equation was linear within the concentration range 1.04-428.3

μg/mL with an R2 of 0.9996.

3.2.6 Antimicrobial activity in animal digesta

Pig digesta from the jejunum, ileum, and cecum sections were collected from previously

dissected pigs (Department of Animal Science, University of Guelph) and then frozen at -20 ºC

until used. In an incubation test, the digesta were defrosted overnight at 4ºC or by warm water

bath, vortexed and then divided up into 2 g of digesta per sterile tube. A fresh culture of E. coli

K88 grown the evening prior from a single colony, was diluted with sterile 0.1% peptone to

obtain an inoculum of (106 CFU/g digesta) when ~20-50 uL was added per 2 g digesta.

Treatments with addition of CIN were mixed a second time by either vortex or stirring with a

sterile wooden stick, depending on the viscosity of the digesta. Incubation was carried out for 5

hours at 37°C with shaking at 200 rpm followed by plating (20µL) on TSA agar plates incubated

overnight to identify the number of surviving colony forming units (CFU/mL).

78

3.2.7 In vitro release of CIN from granules

Simulated gastric fluid (SGF) and simulated intestinal fluid (SIF) were prepared at pH 1.2 and

6.8 respectively. Release of CIN was determined after 2 hours in SGF (750 mL) and then the pH

was adjusted to 6.8 by addition of sodium carbonate solution (~250 mL). The dissolution media

were maintained at 37 °C and the rotation speed of baskets was 100 rpm. The concentrations of

CIN were determined by UV at two wavelengths 291 and 272 nm. The % release was calculated

by converting concentrations into amounts and dividing by the CIN loading in granules.

3.2.8 Statistics

To compare means between groups, the statistical software used was SAS version 9.0 for

windows. The general linear model procedure (proc GLM) was used for fitting data by analysis

of variance. Least squares means were computed and compared with significance level of 0.05

and the Tukey-Kramer adjustments were made for multiple comparisons. Comparison between

two means was performed with a normal student’s t-test.

3.3 Results

3.3.1 Chicken Study

Two groups of chickens were administered CIN as coated granules and as free oil together with

feed by gavage. Animals were dissected after 1, 1.5, 2, 2.5, and 3 hours post-gavage. Table 3.1

summarizes the CIN concentration (μg/g of digesta) found in the digesta as determined by GC

after extraction using ethyl acetate. Data from Table 3.1 represent means averaged from 3-5

chicks obtained at each time point from each treatment group, depending on the amount of

digesta obtained for that section of each animal. The mean concentrations were compared

between free oil and granule formulations within each section to evaluate the ability of the

79

granule formulation to deliver higher concentrations to the various sections of the chicken

intestine, while comparisons also across sampling time within the same rows assesses the effect

of time.

80

Table 3.1. Average CIN concentrations in the digesta of chickens following oral gavage of

two froms of CIN

Section formulation Time post-gavage (hours)

1 1.5 2 2.5 3

Crop free oil 598.92Aa 225.46Aa 445.03Aa 210.47Aa 23.53Aa

granule 1799.34Aa 194.89Ab 154.55Ab 1113.32Aab 499.94Ab

Gizzard free oil 737.85Aa 245.45Aa 157.77Aa 78.47Aa 59.23Aa

granule 1700.81Ba 938.46Bab 637.34Ab 706.75Bb 734.82Bb

duodenum free oil 0.81Aa 0.48Aa 0.37Aa 0.038Aa 0.05Aa

granule 11.14Bb 0.79Aa 0.86 Aa 0.36 Aa 5.54Aab

Jejunum free oil 2.33Aa 6.82Aa 1.87Aa 0.20Aa 1.27Aa

granule 53.38Aab 19.88Aa 18.92Aa 84.07Bb 16.33Aa

Ileum free oil 0.808 Aa 4.15 Aa 7.80 Aa 2.185Aa 3.13Aa

granule 73.99 Aac 9.28 Aa 41.48 Aac 102.25Bbc 18.02Aac

Ceca free oil 0.69Aa 0.77 Aa

0.54 Aa 0.24 Aa

1.07 Aa

granule 0.41Aa 0.80 Aa 9.47 Bb 0.25Aa 0.63Aa

Colon free oil 5.54Aa 3.30Aa 5.93Aa 2.67Aa 4.90Aa

granule 0.94Aa 3.74Aa 8.87Ab 7.56Ba 7.85Aa A, B, C indicates difference between formulation (vertically, within section row) a, b, c indicates difference between different times (between columns) at p <0.05

Table 3.2. Peak concentrations observed from individual chicks within each group

Section formulation Time post-gavage (hours)

1 1.5 2 2.5 3

Crop Free oil 1436.50 657.18 1133.54 445.86 50.80

Granule 2942.68 407.13 202.37 1385.93 878.00

Gizzard Free oil 944.86 309.22 237.89 84.28 79.36

Granule 2441.94 1274.57 1034.37 1309.56 1622.00

Duodenum Free oil 1.25 0.89 0.79 0.05 0.07

Granule 15.99 1.39 1.19 0.42 16.20

Jejunum Free oil 4.88 17.63 3.19 0.33 1.93

Granule 156.32 56.64 48.74 144.06 45.8

Ileum Free oil 1.24 8.71 10.89 2.86 4.46

Granule 290.07 12.82 91.50 171.38 59.40

Ceca Free oil 1.27 0.94 0.91 0.34 1.63

Granule 0.51 1.30 15.65 0.31 1.01

Colon Free oil 10.16 7.74 9.73 2.82 6.36

Granule 1.30 5.21 11.15 11.68 11.89

81

Table 3.2 shows the corresponding peak concentration or the highest concentration detected

within the group of replicate animals from the same time point and digesta section as shown in

Table 3.1. Observing the peak concentrations enables the upper range to be compared between

formulations and reflects the possibility that such levels could be attained in animals, but at

another time-point than observed caused by the differential rate of passage of the multi-

particulate dose between animals.

There was incomplete passage of the granules by 3 hours where a significant amount of granules

were still present in the crop and gizzard that were visibly intact. This delay of movement is

reflected in the peak concentrations in the crop and gizzard at 3 hours. On the other hand,

granules were broken down and not visible within digesta from the lower sections including

duodenum, jejunum, ileum, ceca and colon, although a green colour marker was visible, which

came from the dissolution of the coating material. Average concentrations produced by coated

granules were statistically higher in the jejunum, ileum, ceca and colon between 2-2.5 hours, but

other time points were not. Low CIN levels were found in the duodenum, ceca and colon,

possibly due to the smaller amount of digesta typically obtained (usually ~1 g or less) from those

sections relative to the other sections, and also absorption of CIN from the digesta likely

occurred. Since both groups of chickens received the same dose of CIN by gavage, this data

shows that granules were able to retain higher concentrations within the GIT for longer periods

of time (crop+ gizzard) and deliver higher concentrations of CIN than the free oil to the lower GI

tract sections. In the jejunum and ileum, free oil peak concentrations were <20 ppm, while

granules yielded up to 290 ppm. The peak concentrations listed in Table 3.2 provides some

evidence that coated granules were able to deliver higher concentrations to the lower parts of the

alimentary tract, which can be attributed to the delayed release of CIN from coated granules.

82

3.3.2 Pig Study

In a second study performed in pigs with the same goal of comparing the effect of using an EO

formulation vs free oil, two groups of pigs were administered either CIN in coated granule

formulation or as free oil together with feed at 1500 μg/g, but only one sampling time point was

taken, at 5 hours post-administration due to the higher experimental cost of using larger test

animals. Also the animals were allowed to consume the feed containing the EO additive

voluntarily to mimic in-feed conditions, in contrast to the chickens that received via gavage.

Table 3.3 summarizes the CIN concentrations observed in each digestive tract section for each

animal and also the averages for each section have been calculated. Comparisons between

formulation (within column) and at different GIT sections (within rows) are shown with

superscripts. Peak concentrations within the replicate animals of each GIT section are

highlighted in bold.

83

Table 3.3. CIN concentration in pig digesta after 5 hours post-feeding of feed containing

free oil or coated granules at 1500 µg/g (PPM) of feed

Pigs fed with coated CIN granules added to feed

pig # stomach duodenum jejunum ileum cecum colon

1 267.92 58.2 2.88 27.68 13.84 43.92

2 380.53 79.92 6.1 5.39 37.4 49.11

3 159.71 65.96 19.57 1.97 12.1 18.17

4 335.12 340.37 77.87 1.34 35.24 38.45

5 43.77 0.35 0.3 0.75 9.01 36.28

6 5.53 1.86 3.61 1.48 20.24 21.89

AVE 198.76Ab 91.11Aa 18.39Aa 6.43Aa 21.30Aa 34.64Aa

Pigs fed with free CIN oil added to feed

162.79 2.09 0.32 0.36 4.42 22.18

8 76.2 1.16 0.32 0.38 5.26 20.86

9 182.5 7.9 0.11 0.43 1.4 13.51

10 52.28 0.31 0.8 0.99 6.51 21.89

11 62.58 14.35 0.34 0.61 11.48 26.57

12 97.75 1.29 N/A 0.59 5.22 15.25

AVE 105.68Bb 4.52Ba 0.38Aa 0.56Aa 5.71Ba 20.04Ba

N/A = insufficient digesta material was obtainable for analysis.

Means with different superscript capital letters A, B indicates statistically significant

difference (P<0.05) between free oil and coated granules within the same digesta section.

Means with different superscript small cap letters a,b indicates statistically significant

difference (p<0.05) between columns

Table 3.3 shows that the average concentrations in stomach, duodenum, cecum and colon were

significantly higher in the group fed coated granules than the free oil group. During feeding of

some animals, the granules were observed to have segregated from feed due to their smaller size

and rounder shape compared with feed particles. The granules were gathered at the bottom of the

trough of pigs #3, 4 and 5, 6 (in 2 pens) and some leftover feed and granules were observed prior

to slaughter, thus the actual amount consumed was lower for these animals. The individual

variability within the group receiving granules was higher as a result. The peak concentrations

84

(upper range) indicate that higher concentrations could be attained, but likely occurring at other

time points.

3.3.3 Antimicrobial activity of CIN against E. coli K88 in animal digesta

In order to better simulate the antibacterial effect of CIN in vivo, digesta was obtained from

slaughtered pigs and used directly, as a semi-solid test medium. Table 5 shows the response of E.

coli K88 (initial 106 CFU/mL) to CIN at two concentrations in different sections of digesta, 600

and 1200 µg/ml which represent 1X and 2X the MBC of CIN for E. coli K88 determined

previously in a purely liquid medium (TSB).

The digesta exhibited an attenuating effect on the antimicrobial activity of CIN on E. coli K88 as

even at 2X the MBC there wasn’t complete eradication of the pathogen, and instead there was

dose dependent reduction observed in the ileum digesta. Secondly, the control bacteria grew to

an extent that was highly dependent on the type of digesta, with significantly less growth

occurring over the same period of time in the cecal digesta compared to jejunum and ileum

digesta. Digesta from the pig cecum appeared to neither promote nor inhibit growth, but rather

sustained the same concentration of pathogen that was added initially. But the dose of CIN

failed to exert significant antibacterial effect on the pathogen in this digesta type.

Granules containing CIN antimicrobial and resulted in a concentration dependent log reduction

in E. coli K88 after 5 hours incubation in the jejunum and ileum digesta same as the free oil.

Additionally, the bactericidal effect was lower as well compared with the effect in jejunum and

ileum digesta.

85

Table 3.4. Antibacterial activity of CIN against E.coli k88 in different digesta types (pig)

and formulation from 5 hour incubation, data presented as mean log10 CFU/mL

CIN

ppm

(μg/mL)

Digesta type

jejunum ileum ceca

Control 0 8.54Aa 8.46Aa 5.74Ab

Oil 600 4.08Ba 3.88Ba 5.72Aa

1200 3.90Ba 1.14Cb 5.30Aa

Granules 600 6.47Ba 6.02Ba 5.71Aa

1200 3.25Ba 1.42Cb 4.80Aa A, B, C indicate differences between means within the same column at the p < 0.05 level a, b, c

indicate differences between means within the same row at the p <0.05 level

3.4 Discussion

3.4.1 Chicken trial

Administering the granule formulation of CIN resulted in higher average concentrations in

several sections of the lower GIT of chickens as compared to the free oil. Over the 3 hours,

average concentration ranges of CIN in the jejunum (16-84 μg /g), ileum (9-102 μg/g), ceca (0.5-

9.5μg/g) and colon (0.9- 8.9 μg/g) sections were higher than the free oil groups where

concentrations were below 10 μg/g in all sections once past the gizzard. Peak concentrations

(from individual chickens) observed in the group administered coated granules were also about

10-30X higher, especially in the duodenum, jejunum, and ileum regions. These results provide

evidence that the granules can achieve delivery of CIN at higher concentrations and further down

the GIT. At the end of the experiment time-point (3 hours), however, highest concentrations

were still observed in the crop and gizzard, likely the due to delayed movement of the crop

contents. Thus, a longer experiment duration (>3 hours) would have been necessary to enable

more of the initial dose to pass further down the alimentary tract of the chickens, and obtain a

more complete picture of the released CIN in the intestine. The delayed emptying of the crop

86

even after 3 hours could be attributed to the normal storage function of the chicken crop which

may hold the ingested food for extended periods of time (>3 hours) until digestion is ready to

proceed as signalled by the gizzard and small intestine [158]. Importantly, the CIN

concentrations detected in the intestinal regions would have been high enough to reduce the total

anaerobic bacteria, for which CIN at a minimum concentration of 56 mg/L was reported to be

effective [63].

3.4.2 Pig trial

In the pig trial, following voluntary consumption of the feed containing 1500 μg/g CIN over 5

hour duration, the coated beads yielded higher digesta CIN concentrations in the stomach,

duodenum, cecum and colon sections but not in the ileum and jejunum sections. Several causes

could explain the discrepancies:

Segregation of beads - the spherical granules tended to separate from the feed and settle to the

bottom of the trough, resulting in incomplete uptake of the entire dose and/or non-uniform intake

by animals. To minimize this effect, a step to further granulate the beads with the feed particles

to prevent their after mixing, possibly with an inert binder fluid such as a starch solution could be

done. High inter-animal variability - concentrations determined within each group/section

displayed high variability between individual animals (Table 3.4), which could have been caused

by the divergent voluntary feeding behaviour among animals (i.e. faster or slower rates of

feed/water ingestion that could have affected the rate of movement of the dose along the GI tract,

and the location of the granules within the GIT over time). The larger volume of intestinal

digesta in the animals (>10X) as compared to smaller animals such as the chicken, could have

also played a role in limiting the even distribution of the dose. After oral administration of EO

87

compounds to pigs, high variations in the digesta concentrations among animals were observed

by others as well [65] and warranting the use of higher replicate animals provide stronger

evidence in future studies. Observing the peak concentrations suggests that coated granules of

CIN could deliver much higher concentrations further down the pigs’ GI tract as compared with

free oil in individual pigs. Our results were consistent with other studies [58, 65] of orally

administering EO compounds together with feed to pigs, which showed that very little of the

administered EO compounds (trans-cinnamaldehyde, carvacrol, thymol) reached the distal parts

of the intestine including the jejunum, ileum and cecum.

3.4.3 Antimicrobial activity

Previously CIN was reported to be inhibitory and bactericidal (MIC: 150 μg/mL, MBC: 600

μg/mL) toward pathogens like ECK88, in a purely liquid environment. When tested in digesta,

the current results indicated that CIN used at the MBC, several log10CFU unit reductions

occurred but total bactericidal action was not observed against this particular pathogen. In the

jejunum digesta: there was significant reduction of ECK88 compared to the control at 1X and

2X the MBC of CIN, but no difference between the two concentrations tested. In ileum digesta

there were dose dependent reductions occurring as a more significant reduction occurred at 2X

than at 1X the MBC. In cecal digesta, CIN at 1X or 2X the MBC failed to affect the levels of

ECK88 over the 5 hour incubation period. The CIN used as either free oil or as granules

produced the same results, indicating that the activity was preserved.

The main differences between a liquid medium (like tryptic soy broth) and actual pig digesta are:

viscosity, solid- and dry- matter content, and nutrient availability for bacterial growth, factors

which can significantly interfere with diffusion of EO compounds as compared to being in a

88

purely liquid environment where diffusion occurs from all angles to uniformly access the

bacterial surface [51, 64, 65]. Thus, the reduced antibacterial activity as compared to in vitro,

could be explained by two factors: 1) the EO accessibility in a semi-solid medium like digesta

blocking access to bacteria, and 2) the physiological state/growth rate of the bacteria could be

different in each digesta type. This second point is exemplified in the extent of growth of control

ECK88 in each type of digesta with rapid growth in jejunum and ileum digesta but no change in

cecal digesta. Differences in the digesta could be explained by the successive removal of water

and nutrients as material is undergoing digestion in the pig intestine. As the digesta arrives near

the cecum and colon sections, most of the nutrients and liquid have already been absorbed while

numerous other bacteria carrying out fermentation begin to dominate. Thus it appears the CIN

antibacterial activity relies upon physiological activity of the bacterium. ECK88 in the cecum

digesta could possibly be in more of a stationary phase or dormant state, unlike bacteria in the

jejunum and ileum where active growth occurred due to the availability of more nutrients and

higher water content. It is worth noting that CIN has retained good activity under the jejunum

and ileum digesta conditions as those are the main sites of infection of ECK88, where their

fimbriae mediates attachment to small intestinal brush border enterocytes of the jejunum and

ileum of pigs [148]. Although the effect of CIN on ECK88 in pig cecal digesta was negligible,

previous reports indicated that at concentrations of 100 μL/mL cinnamon oil in pig cecal digesta

had significant reductions (1-2 log10CFU) of E. coli O157:H7 as well as indigenous E. coli, and

coliform bacteria [41]. While a study performed in 20-day old broiler chickens showed that CIN

addition in feed was able to significantly affect Salmonella enteritidis counts in the ceca of

chickens but much higher inclusion concentrations were needed (0.5-1% v/w or 5000-

10000μg/g), while sub–inhibitory concentrations were found to affect bacterial motility and

invasiveness gene expression [67].

89

3.4.4 Release in vitro

Coated granules exhibited partial resistance to release when tested in simulated acidic conditions

which after 2 hours released around 27% of the CIN from coated granules, followed by release of

>90% of the remaining CIN under SIF conditions. The early release of the active in SGF could

have been caused by the migration of the volatile liquid oil from the cores to the coating layers

during the fluid bed process. The long coating times (>2 hours) used could have contributed to

this as well as loss of the active as the constant fluidization by air could have drawn away some

of the volatile EO from the cores. Further optimization of the fluid bed coating process or

selecting an alternative coating process (dry coating) would be needed in future experiments.

3.5 Conclusions

The current study has evaluated the in vivo performance of coated CIN granules fed orally to

chickens and pigs. Release of CIN was quantified in the various GIT sections over time.

Although the pig study yielded more variable results due to the uneven granule distribution in the

feed and non-uniform uptake of the dose by the time of sampling, there was some evidence of

the better performance of coated granules over free oil. The results from the chicken and pig

studies clearly showed that with the coated granules, higher local concentrations were delivered

to the lower GI tract that were above the MIC values determined in vitro for ECK88. Due to

high inter-animal variability in the current study, in order to obtain more reliable future data,

better distribution of granules within the feed to avoid separation form feed, allowing sufficient

time for all the animals to finish feeding and passage of the dose (extended for >3 hours for

chickens), and increasing the number of replicate animals may be possible future options. CIN

antimicrobial activity in digesta was diminished when compared with a purely liquid medium

due to marked differences between them, and also a source of considerably dissimilar growth

90

conditions of the bacterium in various digesta types. This warrants further investigation as it

may have implications against the acute effectiveness of other types of antimicrobials in digesta

or in animals. Formulation into coated granules enables conversion of EOs into a solid feed

additive that is stable during storage and achieves better delivery to lower GIT regions where

pathogens causing illness in livestock animals typically reside. Conventional addition of EO

compounds leads to minimal levels detected in the hind gut with questionable in vivo

effectiveness, but further research will be needed to test the antibacterial effectiveness of this

formulation in experimentally infected animals, furthering the possibility that essential oil

compounds can be used as antimicrobial drug alternatives in feed, shifting the use and resistance

development away from more medically important antibiotic drugs.


The authors would like to thank Guelph food research Centre, Agriculture Agri-Food Canada

and Graduate Department of Pharmaceutical Sciences, University of Toronto for funding of the

research. Also thanks to the Toronto Institute of Pharmaceutical Technology for carrying out the

coating of core granules.

91

Chapter 4. Temporal distribution of encapsulated bacteriophages during passage through

the chick gastro-intestinal tract

Yin-Hing Ma*,†, Golam S. Islam*,‡,Ying Wu*, Parviz M. Sabour*, James R. Chambers*, Qi

Wang*,1, X. Y. Wu†, Mansel Griffiths‡

*Guelph Research and Development Centre, Agriculture and Agri-Food Canada, 93 Stone Road

West, Guelph, Ontario N1G 5C9, Canada

†Graduate Department of Pharmaceutical Sciences, University of Toronto, 144 College Street

West, Toronto, Ontario M5S 3M2, Canada

‡Department of Food Science, University of Guelph, Guelph, Ontario N1G 2W1, Canada


Animal studies were carried out with the help of GI, YW, QW, JC. Data analysis carried out with

help of PS, JC, and MG)

Poultry Science, Online September 2016 doi: 10.3382/ps/pew260

Reprinted with permission of Oxford University Press

(See copyright acknowledgement/license)

92

Abstract

Encapsulation of bacteriophage (“phage”) protects phage against environmental deactivation and

provides a product that is easy to handle for storage and application with animal feed as an

antibiotic alternative. The objective of this study was to evaluate an orally administered,

encapsulated phage for efficient phage release in the gastrointestinal tract (GIT) of young chicks

receiving feed. An optimized formulation that consisted of 0.8% low molecular weight (MW)

alginate, 2 % ultra-low molecular weight alginate and 3% whey protein completely released the

encapsulated phage within 60 min under simulated intestinal conditions. This product was given

to broiler chicks to determine passage time and distribution of the viable phage within the GIT.

Following a single oral dose of 109 PFU/chick, the major portion (peak concentration) of the

encapsulated phage passed through the chick’s GIT and was detected in the feces within 4 hours,

with low levels being continuously excreted for up to 24 hours. In comparison, the passage of

free phage through the GIT occurred faster as indicated by a peak concentration in feces after 1.5

hours. In assessing the temporal phage distribution, both encapsulated and free phage treatments

showed no apparent difference, both having low levels of 102-106 PFU/g of contents along the

entire GIT after 1, 2 and 4 hours. These low concentrations recovered in vivo led us to examine

various exposure conditions (with feed, fecal material, and buffer solutions) that were suspected

to have affected phage viability/infectivity during oral delivery, sample recovery, and

enumeration by plaque assay. Results showed that the exposure conditions examined did not

significantly reduce phage viability and could not account for the observed low phage levels

following oral administration in chicks that are on feed. In conclusion, an oral encapsulated

phage dose can take more than 4 hours to completely move through the GIT of young chicks.

Thus, repeated or higher doses may be necessary to attain higher phage concentrations in the

GIT.

93

Key words: bacteriophage, oral phage therapy, broiler chicks, Salmonella Typhimurium DT104

4.1 Introduction

Many Salmonella enterica serovars, especially S. Typhimurium DT104 [159, 160], are prevalent

food-borne pathogens commonly associated with poultry and swine products. When present they

pose a high risk of antibiotic-resistant salmonellosis. With the impending ban on general

antibiotic use in food animal production, a safe and low-cost bio-control strategy is highly

desirable to reduce the pathogen in livestock and poultry at the pre-harvest production stage in

order to reduce not only the risk of pathogen spread to humans but also the potential for food-

borne disease outbreaks [161-163]. Recently, bacteriophage have re-emerged as a potential

alternative to antibiotics for control of pathogens in farm animal production, primarily due to

their specificity and potential role in the maintenance of normal gut microbiota [3, 93, 164, 165].

However, apart from reported small experimental trials in various farm animals [6, 93, 166, 167],

their potential use as a bio-control agent in large-scale agricultural settings has yet to be fully

exploited. There are also reports that certain factors such as stomach acidity affect phage

survival during their passage through the GIT [90, 93, 119]. Encapsulation of phage to produce a

stable and solid form of protection could promote phage application in large scale food animal

production by enhancing phage delivery to the lower gut and providing long term storage

convenience and stability [118, 119].

94

Prior to initiating oral encapsulated phage therapy studies, several parameters require

investigation to gain a better understanding of the relationships in vivo between phage stability,

temporal distribution and potential therapeutic effect. In order to optimize conditions for oral

administration of encapsulated phage, a dosing study is necessary to determine the phage passage

time and phage concentrations achievable in various sections of the GIT over time. This

information is necessary for development of the encapsulation formulations and dosing

frequencies for optimal efficacy, i.e. to achieve a desired concentration of phage at the location

of high pathogen load to optimize the phage to bacteria ratio or multiplicity of infection (MOI).

There are few studies published [100, 103, 111] that have investigated the temporal distribution

of phages in animals and humans following oral delivery in the absence of identified host

bacteria in which the phage replicate. Previously we developed an encapsulated phage product

made of Ca2+-alginate-whey protein gel beads that insufficiently released the phage within the

body of young chicks after oral administration [168]. This could be due to the relatively short

GIT and immaturity, i.e. low enzymatic activities, of the young chick's digestive system [169,

170]. It is known that proteases contribute to the release of core substances encapsulated in the

alginate-whey protein microparticles; achieved mainly via pepsin and trypsin breakdown of the

protein component [171]. Therefore, there is a need to further improve the encapsulation

formulation so that release of encapsulated phage can occur rapidly in the absence of mature

digestive enzymes (mainly trypsin), and still provide adequate protection to the phage.

In the present study, we used newly hatched broiler chicks sourced from a Salmonella-free

commercial poultry producer. We aimed to identify any potential exposure conditions during

oral delivery that may have a detrimental effect on phage survival or during sample recovery for

enumeration by plaque assay, e.g. to feed and simulated digestive conditions, as phages are

95

known to be sensitive to many conditions [32, 90] [93]. Thus, the main objectives of the present

research were: a) to modify the calcium alginate-whey protein formulation to produce a faster

release rate, b) to identify any potential exposure conditions during oral delivery that may have a

detrimental effect on phage survival during sample recovery from the GIT for enumeration, and

finally c) to determine the passage and excretion times and in vivo distribution of orally

administered phage (single dose). Such information could provide invaluable guidance for

subsequent challenge trials to test the efficacy of phage therapy for bio-control of Salmonella in

broiler chicks.

4.2 Materials and methods

4.2.1 Chicks used

All animal maintenance, manipulations and experimental procedures were performed according

to the Animal Utilization Protocol (#09R117) approved by the University of Guelph Animal

Care Committee. Thirty six day-of-hatch broiler chicks were obtained from a commercial

poultry producer (Maple Leaf Farms, New Hamburg, Ontario, Canada) and transferred to

isolation units at the Central Animal Facility, University of Guelph. Chicks receiving the same

treatment were housed in individual isolation units (up to 12 chicks) and allowed access to non-

medicated starter feed crumbles (Floradale Feed Mill Ltd., Floradale, Ontario. Table A1) and

water ad libitum for the entire study length. The wire floors of the units with pans beneath

permitted feces collection and testing. Bacteria and phage manipulations and collection of GIT

contents were performed within a bio-safety cabinet to avoid contamination. Surgical

instruments and exterior of the euthanized chicks were disinfected with 70% ethanol and allowed

to dry before dissection of each animal.

96

4.2.2 Bacterial strain used for phage enumeration

Salmonella Typhimurium DT104 (ATCC # 700408) was grown and selected for resistance to

nalidixic acid sodium salt (STDT104 NalR) (Sigma-Aldrich) for studies involving digestive

material to enable visualization/isolation of the target strain colonies among the background

bacterial flora that could overgrow on non-selective media. This selection was done by first

streaking onto TSA (tryptic-soy agar, Becton-Dickinson) plates supplemented with 20-40 µg/mL

Nal and incubating at 37 °C overnight; following which cells were sub-cultured on TSA-Nal or

BGS-Nal plates. A single colony was picked for overnight propagation in liquid TSB when a

fresh culture was needed.

4.2.3 Bacteriophages used

Bacteriophage Felix O1 (FO1) was obtained from the Felix d’Herelle Reference Center

(Université Laval, Quebec, Canada). Phage FO1 was individually batch-propagated as

previously reported [118, 123] and was stored and diluted in sterile SM buffer made up of: 5.8

g/L NaCl, 2 g/L MgSO4·6H2O, 0.1 g/L gelatin (bovine type, Sigma-Aldrich), and adjusted to pH

7.5 with Tris-HCl (50mM), 1 M NaOH and Millipore filtered water.

4.2.4 Bacteriophage encapsulation into Ca2+-alginate-whey protein gel beads

Encapsulation materials used were: low molecular weight (MW) (4-12 cP of a 1% aqueous

solution @ 25ºC) sodium alginate (#A1112, Sigma-Aldrich, Canada), ultra-low molecular

weight (0.3-0.7 cP of a 10% aqueous solution @ 25ºC) sodium alginate (Manugel LBA, FMC

97

Biopolymer, Ireland), whey protein isolate (Bipro, Davisco Foods International Inc., USA) and

calcium chloride hexahydrate (Sigma-Aldrich, Canada). Solutions were prepared in distilled de-

ionized water (Millipore) and sterilized by autoclave or filtration using 0.45 µm membrane filters

(Millipore). The procedure for phage encapsulation was based on previous work and an

encapsulation formulation (1.5% low MW alginate + 3% whey protein) developed previously by

our group [118, 168]. Various modified formulations were prepared by introducing an ultra-low

MW alginate and modifying the low MW alginate and whey protein concentrations as shown in

Table 4.1. The first step involved making stock solutions of 3% low MW alginate, 10% ultra-

low MW alginate, and 10% denatured whey protein, respectively. The pre-encapsulation

solutions were made by combining the stock solutions with purified, concentrated bacteriophage

suspended in SM buffer to desired concentrations as listed in Table 1. The solutions were mixed

gently for 10 min with a magnetic stir bar and then extruded by air pressure on an Inotech IE-

50R encapsulator (Inotech Biosystems International) equipped with a 300 µm nozzle into the 0.1

M CaCl2 hardening bath (250-1000 mL) vessel below, rapidly forming cloudy-white gel beads.

The beads were further stirred for ½ hour to complete gelation and then were filtered, washed

lightly with distilled water and after excess water was dabbed away, they were weighed and

placed into closed 50 mL sterile polypropylene tubes and stored at 4 °C. Phage loading was

determined by weighing (200-250 mg) wet beads into a glass vial, then by adding sterile MBS

buffer solution (50mM of sodium citrate tribasic and Tris-HCl pH 7.5, 200 mM sodium

bicarbonate) to dissolve the beads into a final volume corresponding to a 2-10 initial dilution,

with gentle shaking at 37°C. The average phage loading was determined to be 2.93 1010 ± 2.02

109 PFU/g beads.

98

Table 4.1. Compositions of the different calcium alginate-whey protein formulations used

for phage encapsulation

Formulation Low MW

alginate

(w/v%)

Ultra-low MW

alginate

(w/v%)

Whey

protein

(w/v%)

F1 0.5 2 2

F2 0.5 2 3

F3 0.5 2 5

F4 0.8 2 2

F5 0.8 2 3

F6 0.8 2 5

4.2.5 Bacteriophage enumeration assay

Viable phages were enumerated by a plaque counting technique. After appropriate serial (10)

dilutions with SM buffer, samples (10 µL) were spotted in triplicate onto phage agar plates

supplemented with Nal (Fisherbrand square disposable petri dish with grid, Fisher Scientific,

Canada) and seeded on the surface with fresh S. Typhimurium lawns at a cell density of ~108

CFU/mL. Viable counts were enumerated after plaque formation on the bacterial lawn following

8-12 hrs aerobic incubation at 37°C.

4.2.6 In vitro release of encapsulated phage from Ca2+-alginate-whey protein gel beads

Initial phage concentrations used were between 6.0 105 - 1.4 106 PFU/mL. Phage viability

was assessed after 2 hours incubation in simulated gastric fluid (SGF) (34 mM NaCl adjusted to

pH 2.5 with 0.2 N HCl). Afterwards, the acid-treated beads were subjected to simulated

intestinal fluid (SIF) (50 mM monobasic potassium phosphate, adjusted to pH 6.8 with 0.2 N

HCl or NaOH) conditions to determine the release of viable phage after 30, 45, 60, 90, and 120

min to determine the % of phage released relative to the initial dose.

99

4.2.7 In vitro phage incubation with chick feed under simulated gastric and intestinal

conditions

Chick starter feed crumbles were incubated with phage FO1 for 2 hours in SGF (pH 2.5) with or

without porcine pepsin (3.2 g/L) then for another 2 hours in simulated intestinal fluid (SIF) (pH

6.8) with or without pancreatin (10g/L) added (Sigma-Aldrich, Oakville. Canada). Using 1 g of

feed, an initial inoculum of ~109 PFU phage in SM buffer (200 µL) or encapsulated form (200

mg) were added followed by 4 mL of SGF and subsequent incubation at 41 °C with gentle

mixing on a platform rocker (Roto-Shake Genie, Cole-Palmer Canada) placed within an

incubator. At the end of 2 hours incubation in SGF, 0.5 mL of supernatant was taken for phage

enumeration, then additionally 4 mL of SIF were added and incubation continued for another 2

hours. For the free phage (FP) incubation there was complete loss of viability after incubation in

SGF, so a second dose of FP was added after the addition of SIF for the second part of

incubation. One hundred micro litres were removed at each time point (30, 45, 60, 90 or 120

min) for viable phage enumeration.

4.2.8 In vivo phage distribution following oral administration to chicks

After one day of acclimatization, chicks were gavaged with FO1 in two forms: free phage in

liquid suspension using SM buffer (3 1010 PFU/chick, 12 per group) and in encapsulated form

(4.6-9 109 PFU/chick, 12 per group), using a 1mL syringe (Eppendorf Combitips), while

control chicks (n = 3) received SM buffer containing no phage. After the time points of 1, 2, or 4

hours, four chicks (as replicates) were dissected immediately following cervical dislocation and

the GIT contents collected and weighed individually for FO1 enumeration from the following

sections: crop (0.25-5.1g), gizzard (1.35-2.4g), ileum (0.28-1.1g), and cecum (0.13-0.84g) from

100

each animal. After obtaining the weight of individual contents, sterile suspension buffer (0.1%

peptone + 0.1% Tween 80) was added at a 1:2 (weight:volume) ratio to loosen and disperse the

contents following vortexing. Afterwards, ~500 uL of supernatant were filtered through 0.45 µm

filter centrifuge tubes (Costar Spin-X, Fisher-Scientific, Canada) by centrifuging at 11,000 g

for 5 min, and then were serially diluted using sterile dilution buffer (0.1% peptone water). The

dilutions were then spotted (10 µL) onto freshly seeded plates for phage enumeration after

plaque formation.

4.2.9 Fecal phage excretion profile following oral phage administration to chicks

The fecal excretion of phage from chicks was determined in a separate animal trial using 6

chicks per group serving as replicates. Two groups of chicks were given: 1) free phage in liquid

suspension and 2) encapsulated phage FO1 by gavage (109 PFU/chick), then all the feces after

each time point were collected (pooled) from the floor pan of the isolation unit and SM buffer

added in an amount equal to 1-9 the weight of collected fecal material, corresponding to a 2-

10 initial dilution. Lower dilutions were used when a low phage concentration was expected at

the earliest and latest time points. Control chicks were given 200 µL of SM buffer and feces were

collected at the same interval. Samples were mixed by vortexing and then centrifuged to settle

solid materials; the supernatant was used to spot in triplicate onto freshly seeded STDT104 NalR

lawns to determine plaque counts post incubation. Supernatant from control feces were spotted

onto STDT104 NalR agar plates to confirm the absence of endogenous plaque formation against

our selected strain over-night, prior to the fecal phage excretion study.

101

4.2.10 In vitro phage incubation with chick fecal material and buffer solutions

Fecal material excreted by control chicks was quickly collected into sterile screw-cap tubes for

use in incubation studies to determine its effects on phage FO1 viability. Free phage suspension

(0.1 mL) or encapsulated phage (0.1g beads) was mixed with feces (0.5 g) in a 10 mL screw cap

vial and a 10 volume (6 mL) of SM buffer or MBS solution was added to disintegrate the beads,

exposing phage to the mixture. The control did not contain added fecal material. Incubations

were carried out at 0.5, 1, 2, 4, 8 or 24 hours at 37 ⁰C with gentle shaking. The viable phages

were determined at the end of each incubation time point using the bacteriophage enumeration

assay previously described above to assess the % of viable phage compared to the initial titer.

The combinations of MBS or SM buffer solutions (10mL) individually and with fecal material

(0.5 g) were tested as well for any negative effects on phage survival.

4.2.11 Microscopic observation

To monitor the movement and state of degradation of the beads along the chicks’ GIT over time,

samples were collected during dissection to obtain the GIT contents. Pictures were taken prior to

collection of samples into tubes, to determine if any beads were present. Beads were made

visible by incorporation of trypan blue (0.1% w/v) prior to extrusion of the alginate-whey protein

mixture into a calcium bath.

4.2.12 Statistical Analysis

Concentrations (PFU/mL) of phage were converted to log10 (PFU/mL) and then averaged. Data

are presented as means ± standard deviation (STDEV)/standard error (SEM). Differences

between means were evaluated using SAS 9.0 software for Windows with a P-value <0.05

102

considered statistically significant. Multiple comparisons between means were done with PROC

GLM with LSMEANS and Tukey’s correction to perform two-way ANOVA.

4.3 Results

Phage release from different formulations was tested in the absence of added digestive enzymes.

The results indicate that substitution of low MW alginate partially with ultra-low MW alginate

substantially accelerated the release of phage from the beads compared to a formulation without

ultra-low MW alginate, as shown in Figure 4.1. After exposure to SIF, there was close to 80%

release by 30 min and around 100% after 60 min from all formulations. There was little

difference between the six formulations in terms of the phage release rate. As such, formulation

F5 was chosen for further study because the beads from this formulation were more uniform in

size distribution and contained more high-MW alginate for better mechanical strength.

Figure 4.1. Release profiles of the six encapsulation formulations compared to the 1.5%

Alginate-3% whey protein formulation, in simulated intestinal fluid without digestive

0

20

40

60

80

100

0 30 60 90 120

Ph

age

rele

ased

fro

m b

ead

s,%

Time (min)

F1

F2

F3

F4

F5

F6

1.5%A-3%WP

103

enzymes. F1-F6 refer to the different encapsulation formulations found in Table 4.1.

(AVE±STDEV, n=3)

4.3.1 In vitro incubation of phage FO1 with chick feed and digestive enzymes

Incubation of viable phage FO1 was performed with chick starter feed to evaluate any negative

effects under simulated gastric (low pH) and intestinal (neutral pH) conditions. Table 4.2

summarizes the effects of incubating free and encapsulated phage FO1 with chick feed and

digestive enzymes over time. The results showed that there was no time-dependent detrimental

effect on phage viability (i.e. > 1 log10 reduction) observed during incubation with the chick feed.

In addition, the presence of added digestive enzymes for 2 hours in SGF (pepsin) followed by 2

hours in SIF (pancreatin) did not affect phage viability over the time course tested. Free phage

lost viability completely in the simulated acidic environment in the absence of feed, but was

resilient to the acidic conditions in the presence of feed. This was caused by the observed

buffering effect of feed, which raised the pH of the feed-phage mixtures to between 4.5-5.2 after

mixing with SGF of pH 2.5. Phage FO1 survived in the encapsulated form without feed due to

the Ca2+-alginate-whey protein matrix preventing direct exposure to acid. Although there were

some statistically significant differences observed against some controls at the initial time points

(within rows and within columns) in Table 4.2, they were likely caused by the use of a more

concentrated phage stock during the preparation of controls for each treatment time point. These

differences were minor (< 0.7 log) and did not occur against all controls and the level of phage

remained steady across the incubation times (within column).

104

Table 4.2. Incubation of phage FO1 with and without chick feed and digestive enzymes in

simulated gastro-intestinal conditions. Values are presented in log10 (PFU/mL) (AVE ± SE,

n=3). FP = free phage, EP = encapsulated phage. Means with significant difference (at P <

0.05 level) as determined by a two-way ANOVA and Tukey’s correction, are indicated by

different letters (a,b,c) across row and by different subscript numbers (1,2) within column.

(Feed, enzymes) (Feed, no enzymes) (No feed, enzymes)

Time

(min)

FP EP FP EP FP EP

Initial

log10(PFU/mL)

0 9.59 9.57 9.41 9.37 9.51 9.37

SGF (pH 2.5) 120 9.171𝑎±0.22 8.791

𝑏±0.08 9.181𝑎±0.04 8.951

𝑎,𝑏±0.13 03

𝑐 9.161𝑎±0.17

SIF (pH 6.8)

30 9.502𝑎±0.09 9.051,2

𝑎,𝑏±0.22 9.311

𝑎,𝑏±0.09 9.211

𝑎,𝑏±0.38 9.531

𝑎±0.07 8.621𝑏±0.63

45 9.802𝑎±0.08 9.131,2

𝑎,𝑏±0.09 9.121,2

𝑎,𝑏±0.13 9.381

𝑎,𝑏±0.41 9.381,2

𝑎,𝑏±0.02 8.721

𝑏±0.62

60 9.702𝑎±0.09 9.181,2

𝑎,𝑏±0.25 8.912

𝑎,𝑏±0.14 9.331

𝑎,𝑏±0.26 9.352

𝑎,𝑏±0.05 8.721

𝑏±0.66

90 9.622𝑎±0.08 9.161,2

𝑎 ±0.14 9.161,2𝑎 ±0.04 9.321

𝑎±0.37 9.511,2𝑎 ±0.02 9.081

𝑎±0.57

120 9.662𝑎±0.06 9.452

𝑎±0.35 9.101,2𝑎 ±0.04 9.551

𝑎±0.43 9.521𝑎±0.11 8.951

𝑎±0.47

4.3.2 Phage distribution within the chicken digestive tract following single oral dose in

absence of host bacteria

The distribution of phage FO1 within the GIT contents of chicks administered as a single oral

dose of FP or EP was determined after 1, 2 or 4 hours (Figures 4.2a, b, and c, respectively), in

the absence of host Salmonella. Figure 4.2a shows that phage from EP and FP were present in

the gizzard, ileum and cecum by 1 hour but the highest concentrations remained in the crop.

After 2 hours, both encapsulated and free phage were present in the ileum and cecum at between

3-4 log10 PFU/g contents, which was about ~3 log10 PFU lower than in the crop. After 4 hours,

numerous beads were still present in digesta of the crop and gizzard as seen in Figure 4.3;

105

seemingly functioning as a reservoir for many hours post-gavage. Secondary treatment of these

crop contents with MBS solution yielded phage levels between 6-8 log10 PFU/g, which indicated

that high titres of phage were protected within the beads still in the crop but were disintegrated in

the other regions as comparable levels were found pre- vs post-treatment with MBS (data not

shown). The distribution of FP in the GIT was similar to that for EP in that significantly higher

phage concentrations were found in the crop contents even after 4 hours, compared to the levels

observed in the gizzard, ileum and cecum. There were no significant differences between the

formulations within each time point. Nonetheless, phage FO1 was able to reach all parts of the

chick digestive tract within 1 hour, appearing at concentrations similar to the beads of around 3-5

log10 PFU/g, which was fairly low relative to the initial dose of 9.77 log10 PFU/chick. These

distribution patterns strongly suggested that a single phage dose can take much longer than 4

hours for complete passage through the GIT.

107

Figure 4.2. The in vivo distribution of free/released FO1 after single oral doses of

encapsulated phage (EP) and free phage (FP) in SM buffer after (a) 1 h, (b) 2 h, and (c) 4 h.

(AVE ± ST DEV, n =4). *indicates significant difference between means at p < 0.05 level

after two-way ANOVA and Tukey’s correction. Chicks were given a dose of (FP) 3 1010

PFU/chick and (EP) 4.6-9 109 PFU/chick.

108

Figure 4.3. Images of chicken GI tract contents after oral gavage of encapsulated phages. In

crop contents (a) some blue beads (encapsulated phage) are observed after 2 hours and (b) beads

are still visible after 4 hours. Numerous blue beads are seen among ingested feed in gizzard

contents after (c) 2 hours and (d) 4 hours following gavage.

4.3.3 Fecal excretion of an oral phage dose over time

Another study was carried out to monitor the passage time of phage FO1 through the chick GIT

and excretion into the feces following oral doses of EP and FP (5.86 109 or 9.77 log10

PFU/chick) with the results illustrated in Figures 4.4 and 4.5, respectively. As shown in Figure

4.4, it took only 0.5 hour for viable phage to be released from encapsulated beads and detected in

the chick’s feces. This suggests that a portion of the dose moved rapidly through the chick’s GIT,

possibly driven by water consumption and feed intake immediately following gavage; however,

the rapid movement could also be due to the immature GIT of young chicks. A peak

concentration (~4.5 log10 PFU/g) was found excreted in feces by 4 hours, followed by lower

levels (~3.5 log10 PFU/g) being continuously detected for up to 24 hours post gavage.

109

Comparing Figure 4.5 and Figure 4.4, the movement of the liquid phage dose, FP, through the

GIT of the chicks was more rapid, indicated by the higher fecal concentration after 1.5 hours

compared to 4 hours for EP, then forming a plateau with sustained phage excretion at levels of

between 3-4 log10 PFU/g for more than 6 hours following the initial dose. Data were not

collected past 6 hours because it was initially estimated that all phage from a liquid dose will

have passed out within this timeframe. However, these results suggest phage excretion occurred

for a longer period. The concentrations observed in the feces were comparable to EP levels and

this meant that the free phage in liquid form survived the journey through the chick’s GIT. The

results obtained thus far reveal that phage levels found in the GIT and feces were consistently

lower by many orders of magnitude than the high initial dose given.

Figure 4.4. Fecal excretion of FO1 over time following gavage of a single oral dose of 200

mg beads delivering 6x109 PFU/chick (AVE ± SEM, n=6).

110

Figure 4.5. Fecal excretion levels of phage FO1 given orally in liquid suspension containing

~6x109 PFU/chick (AVE ± SEM, n=6).

4.3.4 In vitro incubation of FO1 in chick feces, MBS and SM buffer

In order to rule out the possibility that short term storage of phage in the sample matrices had an

effect on phage viability, phages were incubated in the presence of chick fecal material in

combination with MBS or SM buffer for up to 24 hours. Table 4.3 illustrates there were no

detrimental effects on phage viability during the incubation (storage) with the chick fecal

material in combination with the encapsulation materials and the buffer solutions used to

dissolve the beads.

111

Table 4.3. Incubation of encapsulated and free phage with buffer in chick feces and its

effect on phage survival over time. Data shown are in log10 PFU, (AVE ± SE, n=3)

Time

(h)

EP + MBS

(control)

EP + MBS

+ feces

FP + SM buffer

+ feces

0.5 9.43±0.03 8.97±0.04 8.90±0.01

1 9.41±0.08 8.92±0.05 8.92±0.06

2 9.39±0.04 9.00±0.10 8.89±0.03

4 9.39±0.02 8.92±0.04 8.68±0.17

8 9.35±0.02 8.92±0.04 8.28±0.57

24 9.28±0.03 8.94±0.03 7.14*±2.28

* indicates significant difference between means (P < 0.05) within row.

4.4 Discussion

The main objectives of the current work were to assess specific concerns surrounding the oral

delivery of encapsulated phage and free phage to young chicks for eventual bio-control of

pathogenic Salmonella, preferably during the pre-slaughter stage of production. Firstly, we aim

to achieve more rapid phage release under intestinal pH conditions considering the specific

characteristics of the GI of young chicks. The current result (Figure 4.4) indeed demonstrated

that encapsulated phage appeared in the feces in less than one hour after oral administration.

Therefore, a fast release formulation would minimize the likelihood of phage being excreted

from the chicks within the beads. Based on a phage encapsulation formulation from earlier

work [118, 123], rapid release was achieved by incorporation of an ultra-low MW alginate into

the formulation and with only slight modifications to the other components, low-MW alginate

and whey protein, the six formulations showed similar fast phage release kinetics resulting in

close to 100% release by 45 minutes (Figure 4.1). In addition, the improved formulation did not

compromise its protective effect in simulated gastric conditions (Table 4.2).

112

Phages are known to have widely differing sensitivities to environmental conditions such as

chemicals, pH, heat, UV light and other agents (Adams, 1959; Ly-Chatain, 2014). In oral

application, phages are directly exposed to the feed, digestive tract contents and enzyme

secretions, most notably, the acidic pH, which can deactivate phages. To combat this, some

authors have added acid-neutralizing agents (calcium carbonate) before oral administration of

phages to animals [32, 102]. Therefore, we investigated the effect of incubating phage together

with feed and digestive enzymes on FO1 survival. We found that incubation of encapsulated

phage FO1 with feed in simulated acidic and intestinal conditions showed no consistent negative

effects on phage viability. The high initial titer of phage was maintained to the end of incubation

in SGF in the case of EP. The chick feed was found to buffer FP against deactivation by acid,

whereas FO1 was easily deactivated at pH 2.5 in absence of feed [118]. Although other studies

have found more significant reductions (~1-2 log10 units) in viable phage following addition to

feed [110], this was not evident with our tested feed. Incubations in the current study were

performed with liquid surrounding the feed particles. Had phage been added directly to a dry

substrate (e.g. dry feed pellets), it would have lost viability from desiccation [90]. Similarly the

incubation of FO1 (both FP and EP) with chick feces showed no negative effects on phage

viability (Table 3). Samples incubated in chick feces appeared slightly lower compared to the

control (no feces) and this was consistent over all time points, suggesting that it was likely due to

surface adsorption of phage [110] onto solid particulates with no significant loss in viability.

Overall, our results indicated that phage FO1 was not deactivated under these simulated

exposures and corroborates phage being found in diverse environments [114] [106] such as

sewage and fecal material [172] and being able to pass through the human digestive tract as well

113

[100]. Some deactivation of the free phage occurred after 24 h in the fecal material, which

suggests that an assay for viable phage is best performed before this storage time is reached.

We assessed the temporal distribution of phage in young chicks after oral phage administration

in encapsulated and liquid forms, and with access to feed and water ad libitum, as our goal was

for phage to be delivered along with animal feed. In the current study, host bacteria were

excluded in order to avoid phage amplification, which could confound phage counts yielded by

passive, oral delivery. Overall, the FP had a similar distribution profile as EP following gavage

of a single dose (~9.8 log10 PFU) of FO1 per chick. Results obtained from fecal phage excretion

(Figures 4.4 and 4.5) and in vivo distribution measurements (Figure 4.2) agreed well. There were

around 3 log10 PFU of phage present shortly after administration and for a sustained period in the

lower intestines, which suggested they had potential to attack pathogenic bacteria. The presence

of significantly higher phage concentrations retained in the crop after 4 hours suggests longer

time is needed for complete passage of a gavaged dose. On the other hand, due to the potential

of the crop to act as a reservoir and its neutral pH, the prolonged retention of phage in this site

could provide sustained exposure to any new susceptible pathogen being ingested orally,

allowing phage to interact and/or propagate to continuously supply the gut with more phage.

The fecal phage excretion profiles showed that maximum concentrations were observed after 1.5

and 4 hours for free and encapsulated phage, respectively, which indicated that a faster moving

portion of the dose passed through the chicks into feces, yet after 4 hours the highest

concentrations (between 5-8 log PFU) still persisted in the upper GIT (crop and gizzard)

following both forms of phage administration, suggesting that non-uniform distribution

(separation) of the dose occurred. This was possible since EP allowed portions of the dose to

move and distribute separately from the feed along the GIT, similar to a liquid mixed with feed,

114

but because of their size, were thus retained longer as evidenced in the slower drop in phage

concentration in the crop. Passage of the FP dose was faster than EP, evidenced by earlier

appearance of peak phage concentration in the feces as well as a more rapid drop in phage

concentration in the crop between 1 and 4 hours. The apparent dissimilar fecal excretion profiles

between EP and FP was most likely due to the fact that EP beads required extra time to be

broken down and release phage from the gizzard, whereas FP behaved more like a liquid,

distributing among the GIT digesta while being carried at the rate of feed passage. FP survived

the journey through the acidic proventriculus and gizzard, which can be explained by the

buffering effect of feed, in addition to the young age of chicks, which do not yet have mature

digestive and enzyme secretion capabilities [170]. These results suggested that upon

administration of a phage FO1 dose, the observed low concentrations of viable phage were not

due to phage deactivation. Rather, the high initial phage dose experienced segmented movement

along the GIT from peristalsis in addition to dilution by continued feeding/drinking, and one

cannot rule out absorption of phage into the chick’s organs or other compartments, since phage

have been observed to pass into the blood stream following oral administration in chicks [111],

mice and humans [98], with the spleen and liver being the primary organs responsible for

removing phage from the circulation. Further studies should be done to examine intestinal phage

distribution after longer periods of time and in the organs in order to observe more complete

passage of the dose to the lower GIT. Although neutralizing antibodies to phage have been

observed by some authors they require longer times to develop and depended strongly on the

route of phage administration [166] [102].

An uneven temporal distribution pattern of ingested material was similarly reported by other

researchers who placed soluble and insoluble markers within feed to study the passage rate in 3-6

115

week old chickens [158, 173]. Svihus et al. showed that during feeding the ingested material

could be retained in the crop for extended periods of time (after 3 hours) before being gradually

passed further down the GIT (to the proventriculus) at an average rate of 5 g contents per hour,

depending on the chicks’ feed intake level and activity of the gizzard. Yet after 30 minutes, a

notable amount of marker was already distributed into all parts of the small intestine. Vergara et

al.[158] reported that the typical half-emptying time of the crop of chickens was between 2.3 -

3.1 hours and insoluble materials/particulates were retained by the gizzard longer than soluble

ones (e.g. Cr-EDTA). The crop and gizzard together regulate downward movement of feed to

allow for continuous digestion and absorption, with the crop able to store food for longer periods

of time in order to sustain long-term energy needs of the animal. The gizzard regulates passage

by withholding material only until particles are reduced to a homogeneously small size before

being passed to the intestines for further digestion and absorption. Considering these digestive

processes work to regulate movement of material, the complete movement of a phage dose

would take much longer than 4 hours, consistent with the observed slow and partial movement of

both the EP and FP doses. Furthermore, the ceca of chicks only permits very small particles

(<0.2 mm) and typically the liquid fraction of the passing digesta to be taken up (where up to

50% of a soluble marker was found accumulated) and can be held for up to 7 hours [158].

Together, these slow physiological processes along with the absence of host bacteria for phage

amplification could explain the low concentrations that we observed in the lower GIT, which can

potentially limit high phage concentrations reaching the cecum from an oral (passive) delivery,

unless the phage loading density can be increased or multiple dosing of phage is applied to

maintain a high ratio of phage to bacteria in vivo.

116

The studies carried out with the model bacteriophage Felix O1 revealed the temporal distribution

of phage in the GI lumen of young broiler chicks consuming feed and water and following oral

delivery of a dose of bacteriophage. This information is expected to guide future studies

evaluating the ability of bacteriophages to influence pathogen colonization levels in young

broiler chicks. Our results showed that the encapsulation matrix successfully allowed delivery of

phage to the entire GIT. Viable phage was found distributed along the small intestine by 1 hour

and appears to be able to persist in the chicks for much longer than 4 hours due to gradual

emptying of crop contents. Limited phage levels were achieved in the ileum and cecum due to

the segmented movement of contents in the digestive system, suggesting that in young chicks

multiple dosing may be required with intervals longer than 4 hours, or the dosage can be

increased. Continuous in-feed phage delivery could present a viable strategy for farmers to

protect flocks against incoming Salmonella by maintaining a steady phage supply along the GIT.

Further study is necessary to examine the dissemination of phage to various organs.


The authors would like to thank OMAFRA Food Safety Program (FS 070724), Poultry Industry

Council, and Agriculture and Agri-Food Canada for funding of this research and Victoria Nowell

for providing technical assistance in carrying out phage screening and feed incubation studies.

117

Chapter 5. Encapsulation of bacteriophages for enhanced delivery to the chicken intestine

in vivo performance

Yin-Hing Ma1,2, Xianhua Yin1, Golam Islam1, *Qi Wang1, Parviz Sabour1, Jim Chambers1,

Keith Warriner3, Mansel Griffiths3, Xiao Yu Wu2



2Graduate Department of Pharmaceutical Sciences, University of Toronto, 144 College St.,

Toronto, Ontario, Canada M5S 3M2

3Department of Food Science, University of Guelph, Guelph, Ontario N1G 2W1, Canada

(Keywords: bacteriophage, phage therapy, encapsulation, broiler chickens, Salmonella

Typhimurium DT104)

*Corresponding Author:

Qi Wang

Guelph Research and Development Centre

Agriculture & Agri-Food Canada

93 Stone Road West

Guelph, Ontario, N1G 5C9, Canada

Telephone: + 1 226 217 8077

Fax: +1 2256 217 8181

E-mail: [email protected]


Phage isolation, purification, screening, and animal studies performed with the help of XY, GI,

KW, MG. Data analysis performed with the help of PS and JC)

118

Abstract

Salmonella enterica serovar Typhimurium STDT104, a pathogen closely linked with poultry,

swine and cattle production and end-products, is a significant economic burden on human health

and food safety. Antibiotic drug-resistance limits further prophylactic use of antibiotics in

livestock, thus, effective alternatives such as bacteriophages are desired. Our studies evaluated

two phage cocktails (CT1, CT2) for their ability to control STDT104. We used optical density

(OD600) to screen phage cocktails that were effective at eliminating STDT104 in broth over 24

hours. Incubation with phage cocktail CT1 in four types of chicken intestinal digesta showed

reductions of 0.5 – 6 log10 units over 24 hours. Two animal Trials assessed colonization levels

after 2 and 7 days of continuous phage treatment with either free phage or encapsulated phage to

simulate continuous in-feed administration. In Trial #1 phage cocktail CT1 was found

ineffective at reducing colonization levels in spleen, liver, ileum and cecal contents of broiler

chicks. Trial #2 expanded sample sizes to re-evaluate CT1 along with an improved cocktail CT2

in oral encapsulated form. CT1 caused a slight increase in STDT104 levels in the livers and ceca

after 2 days, but then a minor reduction in the ileum after 7 days. Treatment with CT2 showed

no effect after 2 days, but after 7 days treatment there were significantly reduced STDT104 in

terminal sections by between 1.3-2.1 log10 CFU/g digesta in the ileum, cecum and colon.

Optimized encapsulated phage cocktails used continuously can potentially be used to decrease

Salmonella colonization and shedding in broiler chicks.

119

5.1 Introduction

Salmonella enterica serotype Typhimurium DT104 (STDT104) is a common, virulent food-

borne pathogen of high risk to food safety commonly associated with contaminated poultry meat,

eggs, fresh produce and related products, which after consumption leads to outbreaks of human

salmonellosis and amounts to a costly annual healthcare problem worldwide [91, 159, 160]. The

originating source of contamination is usually the colonized intestinal tract of chickens, when

through fecal shedding and post-slaughtering procedures exposes the pathogen to poultry meat

and by-products, processing equipment and environment eventually disseminating to humans

causing salmonellosis [26].

Use of a safe, low-cost strategy to reduce the pathogen load pre-slaughter without the need for

medically important antibiotics, is highly desirable currently in times of rapidly proliferating,

multi-antibiotic resistant pathogens worldwide. It has been proposed that pre-slaughter

interventions aiming to reduce pathogen levels in the digestive tract can significantly reduce

downstream carcass contamination [162] and reduce the incidence of outbreaks in humans [27].

Using lytic bacteriophages as antibacterial agents has several advantages [85]. They are highly

specific, targeting only specific strains of bacteria possessing the required surface

antigen/receptors for phage adsorption and initiation of the processes leading to bacterial lysis

and viral replication. Therefore, non-susceptible gut bacterial flora will not be directly affected.

Salmonella phages have been readily isolated from the environment due to their natural

abundance (over 170 have been identified) and are easy to mass produce in a short time at

relatively low cost [91]. In addition, bacteriophages are non-toxic to animals and humans

(eukaryotes) and have been shown to be able to replicate or decrease in number over time in the

body or organs depending on the host population present [86, 98, 106, 110]. To increase their

effectiveness, phages can also be used in various cocktail combinations without constraint to

120

target different surface receptors and to lower the chances of resistance to phage infection of a

single or multiple strains of pathogens [110, 113, 174, 175]. There are several factors that could

limit or reduce efficacy of a phage in vivo. For example, as biological agents they may lose their

viability when exposed to external environmental factors [90, 176] such as low pH, moisture

level, ionic activity, temperature and deactivation by antibodies. Together these factors may

limit efficacy of phages for in vivo applications which requires viability to be maintained from

storage to delivery to the site(s) of pathogen residence. This is why buffered antacid solutions,

sprays, or injectable solutions have been previously tested to apply viable phage for treating

animals [113, 115, 177]. Phage therapy using protected phage by encapsulation can help

overcome some of these limitations [118, 178]. Advantages of phage encapsulation include

enhanced delivery and controlled release, production of a stable, solid form product that

facilitates ease of storage, handling, and application in the farm setting, a more desirable and

simple method to use compared to a liquid form which requires more strict handling and storage

requirements in order to preserve phage viability. Here we report a series of studies carried out

to evaluate the efficacy of orally delivered encapsulated phages in reducing STDT104

colonization in young broiler chicks beginning from 2 days of age.

5.2 Materials and methods

5.2.1 Bacterial strains used

Salmonella enterica serovar Typhimurium DT104 (ATCC# 700408) (STDT104) is resistant to

multiple antibiotics and pathogenic towards humans and animals. STDT104 was further selected

for nalidixic acid sodium salt (Nal) resistance by streaking a fresh culture on tryptic soy agar

(TSA) plates containing 20-40 µg/mL Nal and incubating at 37 °C overnight (aerobic) forming

121

white colonies. For longer term storage, stocks were maintained at -80º C in 20% glycerol. A

single colony was picked to inoculate into 3 mL of TSB containing 30µg/mL Nal then grown

overnight aerobically (shaken at 150 rpm, 37º C) to obtain a fresh culture. Cultures were

maintained at 4 ºC on BGS agar plates containing 30µg/mL Nal and appeared as round, pink-

white colonies.

5.2.2 Enumeration from samples containing STDT104

All digesta and tissue samples collected were placed in pre-weighed 15 mL sterile falcon tubes

and immediately stored in an ice cooler for transport back to the lab for processing. Digesta

samples were initially diluted based on weight with sterile buffer (2-5X weight with 0.1%

peptone water + 0.1 % tween 80) then serially diluted for same day plating. To weighed liver

and spleen samples were added 1X sterile buffer and were homogenized using a low speed bead

beater for 30 seconds then serially diluted with 0.1% peptone. The plating method used 40-100

µL from at least 2 dilutions from a single sample and was performed in duplicate using a spiral

plating instrument (IUL Eddy Jet spiral plater) or by hand using a sterilized cell spreader onto

BG sulfa agar (BGS) plates containing 40 µg/mL Nal on the same day of sample collection.

Spiral plates containing colonies were counted using the IUL Flash & Go Automatic colony

counter machine and software and verified by visual inspection, to obtain counts in CFU/mL

which were converted for dilution factors into CFU/g of sample material.

Enrichment for Salmonella was also performed on the same day of sample collection. 0.5-1 mL

was taken from each sample and added to sterile culture tubes containing 3 mL of selenite

cystine broth (Difco/BD) and incubated at 37ºC for 12 hours with shaking at 100 rpm

122

aerobically. Positive samples were subsequently plated (100 µL) onto BGS plates containing

Nal (40 ug/mL) then incubated overnight again to confirm presence/absence of pink-white

colonies of STDT104 NalR. STDT104 NalR was used in phage lytic activity screening and in

vivo animal challenge Trials while STDT104 was used for phage propagation and enumeration

procedures.

5.2.3 Bacteriophages used

Bacteriophage Felix O1 was obtained from the Felix d’Herelle reference center (Université

Laval, Quebec, Canada) and formed characteristic pinhole sized plaques after 12-16 hrs

incubation. Bacteriophages F24, F3, F4, F32 were lytic phages obtained from the collection of

Prof. Keith Warriner (University of Guelph, Ontario, Canada) that were isolated from local pig

farms and they characteristically formed similarly large plaques after ~6-8 hours incubation. A

phage cocktail (YJ104) was isolated from untreated municipal sewage water from a local water

treatment plant in the City of Guelph, Ontario, Canada. All phages used formed plaques on

lawns of S. Typhimurium DT104 NalR after 4-12 hours incubation. YJ104 was repeatedly

observed to consist of 2 phages: one forming pinhole (<0.5 mm) and one forming larger (>1 mm)

plaques (after 4-6 hr incubation) which would expand to a halo on further incubation of plates

and could be counted after ~4-5 hrs incubation. Distinct phages contained in YJ104 are being

further characterized by DNA restriction fragment length polymorphism analysis (data not

shown).

123

5.2.4 Phage propagation and purification

Preparation of phage stocks for experiments: a procedure similar to the one given in [179] was

used. Briefly, 4 liter flasks containing 200-1000 mL of sterile trypticase soy broth (TSB) growth

medium were used. Phage used for propagation was pre-mixed with a fresh O/N culture of

Salmonella Typhimurium DT104 for 10 mins and then added to the flask and shaken @37 ºC for

8- 16 hrs until the OD started decreasing. Phage lysate was obtained by chloroform lysis of

bacteria and centrifugation to obtain crude phage in the supernatant. The clear phage

preparations were pooled and then treated with 1 µg/mL pancreatic DNase I and RNase A

(Sigma-Aldrich, Canada). Phage was precipitated overnight by addition of NaCl (1 M) and PEG

8000 (10%w/v) and then pellets were obtained by centrifugation at 11,000 g at 4 °C. Pellets

were washed and re-suspended with sterile SM buffer (NaCl 0.58%, MgSO4 0.2%, bovine

gelatin 0.01%w/v, buffered to pH 7.5 with 0.2M Tris), then extracted 3-4 times with chloroform

to remove PEG and then further purified by CsCl gradient ultra-centrifugation for 4 hours at 4°C

using a swinging-bucket rotor type SW28 (Beckman). Phage was collected at the 1.5 g/mL band

then dialyzed three times overnight with sterile-filtered Hank’s balanced buffer solution and then

sterile filtered (0.45 µm) lastly. Phage titre of the produced stocks was determined using plating

method and were then stored at 4 ºC until used and were generally around 1011-12 PFU/mL at the

end of purification. All phages were propagated separately (FO1, F3, F4, F24, F32) with the

exception of phage cocktail YJ104 which yielded two plaque types at nearly identical PFU /mL

in the crude lysate (present at same log10 PFU ~9-10), appearing to replicate at the same rate

during repeat (n>6) large-scale propagations. YJ104 was also purified to ~1010-11 PFU/mL using

the same procedures for other phages as mentioned above.

124

5.2.5 Phage enumeration

Phage agar (PA) plates (Nutrient Broth 2%, NaCl 0.85%, Bacto Agar 1.25% w/v) were prepared

in square, 36 box grid plates (Fisher Scientific) and allowed to solidify. Nal was added just

before pouring plates to a final concentration of 20µg/mL. To prepare for phage titration, a fresh

culture of STDT104 was diluted to ~108 CFU/mL (1/10) and spread onto cooled plates followed

by drawing off excess liquid by pipette and air drying of plates for 15 mins. Dilutions of phage

samples were then dropped in 10ul spots in duplicate then incubated 5-16 hrs at 37ºC until clear

plaques were visible against the growing bacterial lawn. The limit of detection was ~500

PFU/mL.

5.2.6 Bacteriophage in vitro activity screening assay

To study the in vitro lytic activity of phages, screening was performed in liquid bacterial growth

media (TSB). Briefly ~106 CFU/mL of STDT104, appropriately diluted from a fresh ~16 h

culture, was added into a sterile 96-well plate (up to 300 µL volume per well) with each

individual well serving as sample replicates, while repeat experiments were carried out on

another day using a new plate. The OD600 was recorded every 20 mins with shaking at 37 ºC for

up to 24 hours under aerobic conditions in a Powerwave XS multi-plate reader. Experiments

were repeated at least twice. The ability of phages to reduce bacterial growth was examined

against control bacterial growth at the same initial CFU/ml. Effect of phage types and

multiplicity of infection (MOI, the ratio of phage to bacteria) were investigated at 10-fold

intervals while keeping the initial CFU constant.

125

5.2.7 Bacteriophage encapsulation into alginate-whey protein gel beads

Encapsulation materials used included: low viscosity sodium alginate (Sigma-Aldrich), ultra-low

viscosity sodium alginate (Manugel LBA, FMC International, Ireland), whey protein isolate

(Bipro, Davisco Foods Int. Inc.), and calcium chloride hexahydrate (Sigma-Aldrich). The

procedures for encapsulating phages were based on previous work in our group [118]. The

formulation of beads was further modified to optimise the phage release rate. Encapsulation of

phage involved first making a pre-encapsulation mixture containing, a final concentration of,

sodium alginate (0.8% w/v), denatured whey protein (3% w/v, a 10% solution was heated to 80

ºC for 30 mins then cooled), and bacteriophage (~1011-12 PFU/mL) in SM buffer. This was

mixed evenly for 10 min at room temperature with a magnetic stir bar at slow speed and then

extruded into an agitated 0.1M CaCl2 gelation bath at room temperature using an Inotech IE-50

encapsulator (operated with 1000 Hz, 1.2 kV, 300 µm nozzle). The gel beads were stirred slowly

for ½ h to complete gelation and then filtered, washed briefly with distilled water and after

excess water was removed, weighed and placed into capped 50 mL sterile polypropylene tubes.

Phage loading was determined by having weighed (200-250 mg) wet beads into a vial, then

added MBS (50 mM sodium citrate, 0.2 M sodium bicarbonate, 50 mM Tris-HCl adjusted to pH

7.5) solution to dissolve the beads into a volume corresponding to a 5-10X initial dilution

followed by successive decimal dilutions. After dilutions with SM buffer they were spotted onto

STDT104 lawns and viable counts were obtained after plaque formation (5-16 hrs). The final

beads contained up to ~1x1011 PFU/(g of wet beads).

126

5.2.8 In vitro phage-bacteria incubations in animal digesta

Digesta (intestinal contents) was obtained from the duodenum, jejunum, ileum and cecum of un-

medicated chickens (2 weeks old), pooled and frozen at -80 °C and was defrosted prior to use.

Incubations of phages with bacteria were done in digesta to determine if the efficacy was

comparable to being in liquid medium. In a typical experiment 1-2 g of digesta was weighed

into 2.5 mL capped sterile glass vials, then a fresh culture of STDT104 was used to inoculate

~106 CFU/mL into the digesta by adding a volume (20-50 uL) of appropriately diluted culture in

TSB. Then phage(s) diluted in SM buffer were added next at an MOI of 10 followed by

vortexing for 30 sec to distribute phage and bacteria throughout the digesta. Incubation was

carried out at 37 °C with periodic mixing every ½ hr by vortexing. Three sample replicates were

used per time point. At desired times small samples of 100 uL were taken and diluted

appropriately for plating and enumeration. At the end of incubation, a final vortexing was done

and 100 uL samples were taken and serially diluted then plated for enumeration.

5.2.9 In vivo Trials of oral phage therapy effectiveness in broiler chicks

All animal maintenance, manipulations and experimentations were performed according to the

Animal Utilization Protocol (# 09R117) that was approved by the University of Guelph Animal

Care Committee prior to carrying out experiments.

5.2.10 Animals used for experiment and care

Day-of-hatch (day 0) broiler chicks (Gallus gallus) were purchased from Maple Leaf Poultry

farms (New Hamburg, Ontario, Canada), a commercial hatchery and transferred to animal

127

isolation units at the Department of Animal Science, University of Guelph which are operated

under strict biosecurity measures. Each unit (box within room) housed a maximum of 13 chicks

and provided distilled water and autoclaved chick starter feed (Arkell Feed Mill, Guelph,

Ontario, Canada) ad libitum during the entire study. The isolation units were box shaped (0.6m

X 0.6m X 0.9 m, height x width x depth, and floor space was about 0.6m x 0.9 m and were fed

clean air and had infrared lamps positioned on the outside as a heat source. On sampling days

animals were sacrificed by using cervical dislocation or CO2 asphyxiation protocol for small

animals. All animals were dissected immediately after euthanization in a sterile operating room

and equipment sterilized by (70%) ethanol between individual chicks. Samples collected from

animals included digesta from ileum, cecum, and colon and whole spleen and a section of the

liver were all placed into individually labelled sterile containers and immediately kept on ice.

5.2.10.1 Challenge Trial # 1

Seventy six broiler chicks were randomly divided into groups of 12-13 chicks then placed into

isolation units, one unit for each of the 6 treatment groups: (A) negative control (no phage or

bacteria), (B) positive control (bacteria only), (C) free phage CT1, (D) encapsulated phage CT1,

(E) bacteria + free phage CT1, (F) bacteria + encapsulated phage CT1. Cocktail 1 consisted of

F24+FO1 at a 1:1 ratio. The number of chicks per treatment group differed by 1-2 due to some

extra chicks present at start of the Trial or because of mortality on a non-sampling date. After 1

day of acclimatization to the isolation box, the chicks were infected once during the entire study

by gavage using 1 ml tuberculin syringes containing 0.2 mL of 0.1% peptone with 105 CFU of

STDT104 NalR on day 2 (at 2 days of age). Groups not receiving bacteria were gavaged with

sterile 0.1% peptone water as negative control. Two hours post-gavage, oral phage

128

administration was started and groups receiving phage treatment (free phage or phage

encapsulated beads) were gavaged twice daily, in the morning (~9 am) and evening (~9 pm) for

the remainder of the entire study. Chicks were sacrificed after 2 days (at 4 days of age) and 7

days (at 9 days of age) continuous phage treatment to obtain samples for STDT104 NalR and

phage enumeration.

5.2.10.2 Challenge Trial # 2

For the Trial #2, a total of 125 chicks from the same source as Trial #1 were obtained and

divided into 5 groups of 25 chicks for each treatment. Each treatment group was divided into 2

adjacent isolation units so the maximum number of chicks per isolation box was limited to 13.

The sample collection procedures were same as Trial 1, but with 11-13 animals being sampled

from each treatment group and day, and with the supplementary collection of the colon digesta

from all remaining birds on the last day of sampling. This Trial compared CT1 with CT2, which

was composed of CT1+YJ104 in a 1:1 ratio.

5.2.11 Fecal phage excretion after a single oral dose

To obtain an idea of the phage transit time through the alimentary tract of young chicks after

single gavage of phage formulations, the feces from groups of chicks administered the same

phage dose were collected, pooled and determined for the presence of phage over 24 h at ½ h

intervals. Chicks were placed into groups of 6 per treatment with 3 replicate groups. The feces

were collected by scraping all material from the floor bedding and pooled into a 15mL

polypropylene capped tube per replicate group per time point using a metal spatula then

immediately diluted 2X-10X in SM buffer by weight and stored on ice. Serial dilutions were

129

performed with SM buffer and then spotted onto STDT104 NalR phage agar plates for phage

enumeration after incubation.

5.2.12 Data presentation and statistical analysis.

Bacteria counts are presented after conversion into log10 (CFU/g or mL) and phage counts into

log10 (PFU/g or mL), and means are presented as average ± standard error. Statistical analyses

were performed using GraphPad v. 5.0 for performing a 2-way ANOVA by entering log10 CFU

or log10 PFU values for individual bacteria and phage counts, respectively, and then compared

with control groups for phage treatment effect and for the effect of treatment duration (2 and 7

days), followed with a Bonferroni-adjusted multiple comparisons of means of treatment groups

to the control group. Each sample type (spleen, liver, ileum digesta, cecal digesta, colon digesta)

was analyzed independently. For these tests, the initial hypothesis (H0) was that phage

treatments produced no difference in mean bacterial counts in chicks compared to the control

group (no phage treatment). In Trial #2 the colon digesta counts, which were only recovered at a

single time point (7 days), were analyzed using 1-way ANOVA for detecting a treatment effect

versus control.

5.3 Results

Preliminary studies. Preliminary studies evaluated the lytic activity of phage cocktails in vitro

and the effect on STDT104 NalR growth in raw chicken digesta. The passage times through the

chick alimentary tract from an orally administered dose of free phage in buffer solution and in an

130

encapsulated form, were evaluated by monitoring fecal excretion of the phage up to 24 hours

post administration.

In vitro lytic activity of individual phages and in combination was compared using the optical

density method to reflect bacterial viability over time [99, 177]. Starting with an initial CFU/ml

of ~106 of STDT104 NalR the phage cocktails with best lytic activity were selected based on 24 h

study. Cocktail 1 (CT1, a 1:1 mixture of FO1 and F24) showed much improved lytic activity

over single and other 2-phage mixtures against STDT104 NalR at lower MOI, as shown in Figure

5.1 and Figure 5.2.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 5 10 15 20 25

OD

60

0

h

STDT104 controlFO1-0.1FO1-1FO1-10FO1-100

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 5 10 15 20 25

OD

60

0

h

STDT104 controlV6-0.1V6-1V6-10V6-100

131

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 5 10 15 20 25

OD

60

0

h

STDT104 control

V16-0.1

V16-1

V16-10

V16-100

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 5 10 15 20 25

OD

60

0

h

STDT104 controlF3-1F3-10F3-100F3-1000

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 5 10 15 20 25

OD

60

0

h

STDT104 control

F4-1

F4-10

F4-100

F4-1000

132

Figure 5.1. Effect of single bacteriophages: FO1, V6, V16, F3, F4, F24, F32 on STDT104

NalR growth at various initial MOI in TSB growth medium and initial cell density of 104

CFU/mL

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 5 10 15 20 25

OD

60

0

h

STDT104 control

F24-1

F24-10

F24-100

F24-1000

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

0 5 10 15 20 25

OD

60

0

h

STDT104 controlF32-1F32-10F32-100F32-1000

133

Figure 5.2. Optical density (@ 600 nm) readings showing in vitro lytic activity of phages

CT1, YJ104 and CT2 at MOI: 1, compared with a control bacterial growth curve (initial

106 CFU/mL of STDT104). Shown are AVE+SE values from experiments repeated twice

with n≥6 per experiment.

Figure 5.3. Optical density (OD600) reading of CT1, YJ104 and CT2 at MOI of 10 with an

initial 106 CFU/mL of STDT104. Note difference in Y-axis scale compared with Figure 5.2.

Shown are AVE+SE readings of experiments replicates ≥6 per experiment.

Figure 5.2 shows that at a multiplicity of infection (MOI) of 1 and initial CFU/mL of 106, CT1

was more effective than the newly isolated YJ104 phage mixture, which was not effective at

such a low MOI. CT2 however, was more effective than either CT1 or YJ104 at this MOI, but

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0 5 10 15 20 25

OD

60

0

h

YJ104-1

CT1-1

CT2-1

STDT104 control

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0 5 10 15 20 25

OD

60

0

h

YJ104-10

CT1 -10

CT2-10

134

did not fully eliminate STDT104 NalR. Figure 5.3 shows that at a higher MOI of 10, both YJ104

and CT2 prevented re-growth of STDT104 NalR, but CT1 did allow some surviving bacteria to

re-emerge at low amounts. The result suggested a synergistic effect of combining CT1 and

YJ104 at MOI 1 and 10. Subsequently, CT2 was selected for evaluation in animal Trial #2.

Following in vitro liquid broth activity assays, raw chicken digesta was used as a medium for

evaluating phage cocktail effectiveness in a simulated intestinal environment. Digesta from

various regions of the alimentary tract were collected and used directly, undiluted to incubate

phage cocktail CT1 with 106 CFU of STDT104 NalR and then bacteria was enumerated after 5,

16 and 24 hours of incubation. Table 5.1 shows that CT1 at MOI 10 resulted in reductions of 3-5

log10 in the duodenum/jejunum digesta, but only 0.5-1 log10 reductions occurred in ileum and

cecal digesta. There appears to be time dependence for the reduction where a greater reduction

was observed with longer incubation times, but total elimination of STDT104 NalR was not

attained, in contrast to what was seen in a purely liquid medium. These experiments were

performed prior to the animal Trials and the results suggested that phage cocktail CT1 could be

effective for achieving pathogen reduction in vivo.

Table 5.1. Effect of phage cocktail CT1 (phage FO1+F24) at MOI 10 on Salmonella growth

in raw chicken digesta isolated from different region of GI tract over time. Data presented

as AVE+ SE of log CFU/g digesta.

Source of Digesta

hr

duodenum jejunum ileum ceca

5 MOI 10 4.71±0.09 5.68±0.43 6.37±0.35 5.64±0.17

control 8.23±0.01 8.66±0.03 8.46±0.04 6.00±0.05

16 MOI 10 5.14±0.22 7.31±0.18 7.33±0.65 5.40±0.03

control 9.06±0.15 9.04±0.10 8.83±0.07 5.88±0.11

24 MOI 10 2.67±0.47 4.61±0.24 4.42±0.08 5.35±0.09

control 8.91±0.10 8.22±1.01 5.35±0.17 5.98±0.04

135

In another pilot experiment we evaluated the transit time of a single oral dose of CT1, in the

absence of host bacteria, through the chicken alimentary tract. We compared the free phage (FP,

in SM buffer) and encapsulated phage (EP, in alginate beads) at two different chick ages

corresponding to the day of sampling of chicks in the forthcoming animal Trials. Figure 5.4

shows the passage of a single FP dose (~5x109 PFU/chick) given to each chick in the same group

(n=10) at 4 days of age and then again 5 days later at 9 days of age, using the same group of

chicks. The chicks received feed and water ad libitum. Free phage given in a solution form

passes through chicks and appears in the feces fairly quickly, appearing after 1 hour and then

reaching a maximum concentration at between 2.5-3.5 hours post administration. There was no

significant difference in phage passage time between 4 day old and 9 day old chicks shown in

Figure 5.4.

Figure 5.4. Excretion profile of phage in feces after a single oral dose of free phage cocktail (~5 x

109 PFU/chick of CT1 in 0.2 mL SM buffer) in 4 day and 9 day-old broiler chicks (AVE±SE, N=3)

0

1

2

3

4

5

6

0 2 4 6 8 10 12 14 16 18 20 22 24

log 1

0(P

FU/g

fe

ces)

Hr

FP- 9 days old

FP-4 days old

136

Figure 5.5 shows the phage excretion patterns in the feces following gavage of a single dose of

encapsulated phage beads of ~2 x109 PFU/chick delivered in 0.2 g of beads. The peak phage

concentrations appeared between 2.5-3.5 h then dropped and tapered off after 4 h and low levels

were detected even after 24 h. Together, these results (Figures 5.4 and 5.5) showed that the

majority of the administrated free or encapsulated phage dose (as CT1) has already passed

through the chicks in less than 3 hours. It is noteworthy that the excreted phage concentrations

from FP were 1-1.5 logs higher than encapsulated phage.

Figure 5.5. Excretion profile of phage in feces after a single oral dose of encapsulated phage

cocktail (~109 PFU/chick of CT1 in 0.2 g wet beads) in 4 and 9 day-old broiler chicks (AVE±SE, N=3)

5.3.1 Animal Trial #1

After determining the in-vitro efficacy of selected phages we looked at their effectiveness in

reducing STDT104 NalR levels in different intestinal regions of interest in broiler chicks, by

carrying out two randomized control Trials. The first Trial evaluated the effectiveness of

0

1

2

3

4

5

6

0 2 4 6 8 10 12 14 16 18 20 22 24

log1

0 (

PFU

/g f

ece

s)

Hours

EP-B-9 days

EP-B 4 DAYS

137

continuous treatment with CT1 (in free and encapsulated forms) to reduce colonization. The

chick gut sampling time points chosen were to be 2 and 7 days post infection/phage

administration to assess any changes in STDT104 and phage levels during these periods of daily

phage administration. The treatment groups with or without bacteria and phage, and animal

sample size (n) per group per sampling day are summarized in Table 5.2 for Trial #1.

Table 5.2. Treatment groups and sample sizes used in trial #1

Treatment Group Bacteria Phage (CT1) n (day 4) n (day 9)

A - Negative control - - 6 6

B - Bacteria control + - 5 6

C - Free phage control - + (free phage) 6 5

D - Encapsulated phage control - + (beads) 6 6

E - Free phage therapy + + 5 6

F - Encapsulated phage therapy + + 6 5

For Trial #1, 76 day-of-hatch broilers were randomly assigned to 6 groups of 12-13, each housed

within an individual isolation box (day 0). Chicks were cloaca-swabbed and determined to be

Salmonella free after streaking on BGS-Nal plates incubated overnight (chicks were obtained

from a specific pathogen free farm). After 48 hrs (day 2) of feed and housing acclimatization,

chicks in groups receiving bacteria were inoculated by gavage with STDT104 NalR (105 -

CFU/chick). Then, after 2 hours of rest phage treatment or buffer (negative control) was initiated

for all groups, receiving twice a day dosing (morning and evening) for the remainder of the Trial,

except on the day of sampling. Animals were euthanized after 2 days of phage treatment (chicks

at 4 days of age) where 5-6 chicks from all treatment groups were randomly selected to obtain

individual whole spleen, section of liver (~0.2 g ), ileal and cecal digesta by extruding contents

into individual sterile containers then placed on ice for further processing. The last sampling day

was after 7 days of phage treatment (chicks at 9 days of age) and the remaining chicks in each

138

group were euthanized and the same sample types were obtained for bacteria and phage

enumeration.

The average log10 transformed STDT104 NalR counts for each treatment group and sample type

are summarized in Table 5.3, while the corresponding phage counts are found in Table 5.4, with

statistically significant differences in means noted (p < 0.05).

Table 5.3. Salmonella Typhimurium DT104 NalR recovery after 2 and 7 days continuous phage

treatment in trial #1 (AVE ± SE, N=5-6) log10 (CFU/g of tissue or digesta)

Treatment group

Sample day STDT104 control FP-CT1 EP-CT1

spleen (tissue) 2 0.62 ± 0.62 0.60 ±0.60 0.73a ± 0.73

7 1.83 ± 0.48 2.45 ± 0.65 3.40b ± 0.57

liver (tissue) 2 1.45 ± 0.64 1.90 ± 0.52 1.27 ± 0.83

7 2.53 ± 0.51 2.13 ± 0.23 3.35 ± 0.73

ileum (digesta) 2 3.52 ± 0.37 4.15 ± 0.32 3.61 ± 0.64

7 4.03 ± 0.78 3.25 ± 0.34 3.28 ± 0.78

ceca (digesta) 2 7.76 ± 0.33 8.61 ± 0.14 7.46a ± 0.50

7 7.86 ± 0.25 8.25 ± 0.09 8.67b ± 0.09

(Note: different subscript letters indicate significant differences (P <0.05) when compared vertically

within treatment group and sample type after analysis using 2-way ANOVA and Bonferroni adjusted

post-tests separately for each sample type)

Results from Trial #1 (Table 5.3) shows that STDT104 NalR was found predominately in the

cecal and ileal contents/region, consistent with the location of Salmonella colonization in chicks.

Furthermore, it can be readily seen that the strain we used was virulent for the chicks since

bacteria was found in the liver and spleen of all 3 bacteria treated groups. The spleen of chicks

treated with EP-CT1 were colonized more after 7 days treatment than after 2 days phage

treatment (p<0.05). Although on day 7 the STDT104 NalR levels appear higher in the spleen and

liver of phage treated groups there were no significant differences found compared to control

chicks. In the CT1-EP group levels in the ceca were found to be higher after 7 days than 2 days

139

treatment. Overall, neither free nor encapsulated CT1 phage significantly reduced pathogen

population. CT1 was not effective in reducing the colonization level in chicks after 2 and 7 days

of continuous phage administration.

5.3.1.1 Phage counts in the absence of host bacteria

Bacteriophage counts in absence of host bacteria from Trial #1 (Table 5.4, left two columns)

show that phage concentrations were found to be highest in the ceca of chicks, and shows that

phage could be delivered to this terminal region of the chick where Salmonella is expected to

reside. After 2 days of administration, both FP and EP control groups had phage detectable in

the ileum and cecum of chicks at levels of between 1.9-5.8 log10 PFU/g, while little to no phage

was found in the spleens and livers of chicks. After 7 days of phage administration, the phage

levels were below the detectable limit in spleen, liver and ileum and were significantly lower in

the ceca than at 2 days.

Table 5.4. Bacteriophage recovery after phage treatment in trial #1 (AVE ± SE, N=5-6) log10 (PFU/g

of tissue or digesta)

Treatment group

sample day FP control EP control FP + STDT104 EP + STDT104

spleen (tissue) 2 0.00 ± 0.00 0.00 ± 0.00 0.00 ± 0.00 2.47a1 ± 1.14

7 0.67 ± 0.67 0.00 ± 0.00 0.00 ± 0.00 0.00b ± 0.00

liver (tissue) 2 1.37 ± 0.62 0.00 ± 0.00 1.05 ± 0.65 1.83 ± 1.30

7 0.00 ± 0.00 0.00 ± 0.00 0.00 ± 0.00 0.00 ± 0.00

ileum (digesta) 2 3.01a ± 0.64 1.92 ± 0.87 2.50a ± 0.65 3.06a ± 1.17

7 0.00b ± 0.00 0.00 ± 0.00 0.00b ± 0.00 0.00b ± 0.00

ceca (digesta) 2 5.76 ± 0.53 5.64a ± 0.39 6.57 ± 0.89 6.84a ± 0.45

7 4.041,2 ± 0.17 2.65b

2 ± 0.85 5.541 ± 0.29 2.92b2 ± 1.25

(Note: subscript letters indicate significant differences (P <0.05) when compared vertically while

superscript numbers indicate differences when compared within row, within same treatment group and

sample type. Data analyzed using 2-way ANOVA and Bonferroni adjusted post-tests separately for each

sample type).

140

5.3.1.2 Phage counts in presence of host bacteria

When host bacteria was present (Table 5.4, right two columns), phage levels were not

significantly different from host being absent and the two forms of phage (FP and EP)

administration also yielded similar levels of phage in the chicks. Looking at the results across

rows (Table 5.4), there were no marked differences in phage levels within each tissue and digesta

type either in the absence or presence of host bacteria. This suggests that amplification of the

phage cocktail CT1 was not occurring or observable within our chosen sampling times.

Furthermore, there was little to no phage found after 7 treatment days than at 2 treatment days in

the spleen, ileum and ceca when bacteria were present. However, the drop in phage levels after

7 days does not appear to depend on the host being present, as this occurred in controls as well.

This observation of dropping levels of phage after a longer administration time (even though the

same dose was given daily) could be caused by the action of the chicken immune system because

during the first week of life the immune system is still immature but is in fact quite active in

producing pro-inflammatory cytokines and gallinacins in response to colonization of the gut by

various microbes [109, 180]. Phage was found to be higher in the spleens of the EP treated

group after 2 days which correlated with a lower STDT104 NalR count (Table 5.3), but after 7

days, phage levels were below detection while host levels increased. After 7 days, phage levels

in ceca appear to be dependent on the form of administration, i.e. lower in EP than in FP treated

groups. The results from Trial #1 show that when phages are continuously administered to

young chicks with feed the organs and digestive tracts do not accumulate high concentrations of

viable phage and the highest concentrations were found in the ceca.

141

5.3.2 Animal Trial #2

The second Trial was designed to use a greater number of animals in each treatment group to re-

test the effectiveness of CT1 in addition to an improved phage cocktail CT2 a 1:1 mixture of

CT1 + YJ104. As there was no significant difference in performance found between FP and EP

administration in Trial #1 we proceeded directly with the encapsulated form for CT2. The Trial

followed the same timeline and sampling days as in Trial #1 while the treatment groups and sizes

are summarized in Table 5.5.

Table 5.5. Treatment groups and sample sizes used in trial #2

Treatment Group Bacteria Phage (CT1) Phage (CT2) n (day 4) n (day 9)

1 - Negative control - - - 11 11

2 - Bacteria control + - - 11 11

3 - FP1 free phage 1 + + (free phage) - 11 13

4 - EP1 encapsulated phage 1 + + (beads) - 11 13

5 - EP2 encapsulated phage 2 + - + (beads) 11 12

The results of Trial #2 for bacteria and phage counts after 2 and 7 days of continuous twice-daily

oral phage administration are shown in Tables 5.6 and 5.7, respectively. Consistent with Trial

#1, a single inoculation with 105 CFU/chick of STDT104 NalR produced sustained colonization

of chicks over a one week period and with ceca being the most highly colonized region at more

than 7 log10 CFU/g digesta as presented in Table 5.6. There was no statistically significant

decrease in bacterial counts after 2 days of treatment in any of the three phage treated groups

when compared with the control group. In fact, STDT104 NalR levels were higher in the livers

and ceca of the EP-CT1 group, but after 5 days of further treatment there was no difference

compared to the bacteria control group.

142

Table 5.6. Trial #2, Salmonella Typhimurium DT104 NalR counts recovered from digesta

and tissue samples after 2 and 7 days continuous phage treatment (AVE ± SE, log10 (CFU/g

of tissue or digesta)

Treatment group

Sample days bacteria control FP-CT1 EP-CT1 EP-CT2

spleen (tissue) 2 2.94±0.14 2.01±0.52 3.28±0.29 2.00±0.50

7 3.27±0.25 2.40±0.37 3.00±0.24 2.89±0.32

liver (tissue) 2 1.92±0.25 2.42±0.27 2.75*±0.25 2.00±0.30

7 2.44±0.19 2.05±0.29 2.84±0.14 2.15±0.20

ileum (digesta) 2 3.26±0.48 2.34±0.62 3.89a±0.38 2.60±0.56

7 3.16±0.32 2.22±0.32 1.61b* ±0.31 1.10**±0.34

ceca (digesta) 2 7.14±0.42 6.85±0.44 8.29a*±0.11 7.33a±0.17

7 7.14±0.17 7.05±0.21 6.64b±0.28 5.89b** ±0.21

colon (digesta) 7 5.18±0.33 3.99±0.49 4.06±0.41 3.19**±0.43

(Note: subscript letters when not same indicate significant differences (P <0.05) when compared

vertically, within treatment group and sample type. Data were analysed using a 2-way ANOVA and

Bonferroni post-tests separately for each sample type. Asterisks indicate significance difference found

between treatments and control group within the same row, * p<0.05, **p<0.01)

Table 5.7. Trial #2, Bacteriophage counts recovered from treatment groups treated with

two phage cocktails (AVE ± SE, N=11-13) log10(PFU/g of tissue or digesta) Treatment group

sample days FP-CT1 EP-CT1 EP-CT2

spleen (tissue) 2 3.02a1±0.15 3.982±0.21 4.28a

2±0.20

7 4.19b1±0.10 3.172±0.30 3.25b

2±0.13

liver (tissue) 2 3.951±0.15 3.551,2±0.21 3.202±0.20

7 3.471,2±0.16 3.511±0.08 3.012±0.10

ileum (digesta) 2 3.91±0.17 4.69a±0.33 4.58a±0.15

7 3.251±0.21 2.35b2±0.20 3.07b

1,2 ±.0.41

cecum (digesta) 2 6.41a±0.51 6.24a±0.61 5.97a±0.20

7 3.99b±0.36 3.70b±0.28 4.53b±0.25

colon (digesta) 7 2.58±0.52 2.72±0.44 2.99±0.53

(Note: subscript letters not the same indicate significant differences (P <0.05) when compared vertically

while superscript numbers indicate differences when compared within same row, within same treatment

group and sample type. No numbers or letters indicate no significant differences found. Data were

analysed using a 2-way ANOVA and Bonferroni post-tests separately for each sample type).

It was only after 7 days of continuous phage administration that phage treated groups yielded

significant decreases in bacterial colonization. In the EP-CT2 treated group, significant

143

reductions were found in the ileum and cecum of 2.06 and 1.25 log10 (CFU/g), respectively.

Colon digesta was collected on the last day only, when sufficient material was obtained for

analysis. A 1-way ANOVA comparison of the colon counts of treatment groups revealed a

statistically significant 1.99 log10 reduction with EP-CT2. Treatment with EP-CT1 resulted in a

significant reduction only in the ileum digesta of 1.54 log10 (CFU/g) but no differences were

found in the tissues, cecum or colon when compared to control. The results obtained for

treatment with FP-CT1 and EP-CT1 are consistent with results found for this cocktail in Trial #1,

where there was a general lack of effectiveness of CT1.

Table 5.7 shows that phage counts tended to drop over time in some of the sampled regions in

the treatment groups, such as the spleen, ileum and cecum after 7 days treatment. This trend was

also observed in Trial #1. Table 5.7 shows that the phage counts were again highest in the ceca,

where bacteria density is also highest, and levels were fairly constant across the treatment

groups. Although CT2 was more effective than CT1 in reducing colonization, phage levels were

not found to be significantly different.

5.4 Discussion

Previous studies in our group have shown that bacteriophage can be successfully encapsulated

into Ca+-alginate-whey protein microspheres/beads at high efficiency and high phage titers and

protects from external environmental deactivation. However, since the level of viability

maintained post-drying has not yet been fully achieved in recent studies [118, 181, 182], for the

present studies wet beads were used to maintain the high PFU/g of 1010 PFU/g wet beads so the

144

dose given to animals was easily measured by weighing of beads into individual gavage syringes.

In chapter 4, we optimized the formulation of beads for faster release and in this study we

showed that a two-phage cocktail yielded much improved in vitro lysis of STDT104 compared

with single phages, producing clear plaques even after 24 hrs of incubation. Furthermore,

preliminary experiments helped identify potentially useful phage cocktails in vitro and then we

tested their effectiveness in digesta to more closely simulate the intestinal environment. Phages

in liquid form or encapsulated form were compared for their transit time through chicks gavaged

with a single dose, while animal challenge Trials were carried out to test the effectiveness of two

oral encapsulated phage cocktails to reduce STDT104 NalR colonization in newly-hatched broiler

chicks.

Preliminary studies looked at the in vitro effectiveness of various phage cocktails against

STDT104 NalR viability over time using OD600 as an indicator in broth culture and then testing

their effectiveness to lyse STDT104 NalR in chicken digesta. Figure 5.2 shows that with an MOI

of 1 and initial CFU/mL of 106, CT1 was more effective than YJ104, which was least effective,

while CT2 (CT1+YJ104) was more effective than either one alone for up to 24 hrs. Figure 5.3

shows that at higher MOI of 10, growth was absent for YJ104 and CT2. Based on this synergistic

effect of CT1 with YJ104, they were used together in CT2. The next experiment was to use

chicken digesta, a semi-solid medium to test phage cocktail effectiveness. Table 5.1 summarized

the effectiveness of CT1 vs STDT104 NalR at an MOI of 10 in different types of chicken digesta.

Over a 24 hour incubation period, there were between 3 to 6 log10 reductions of STDT104 NalR

found in the duodenum/jejunum digesta over the 24 hr period, while only 0.5-2 log10 reductions

were found in ileum and cecal digesta. The results in Table 5.1 suggest that the viscosity of

digesta affected the ability of the phages to reduce STDT104 NalR . Digesta obtained from the

145

duodenum contained most liquid while further towards the ceca the digesta approached a semi-

solid with lack of liquid and had higher viscosity, likely hampering the phage-bacteria interaction

resulting in lower lytic ability [183].

5.4.1 Fecal excretion trial

Two forms of administration by phage in buffer and by encapsulated phage beads were

compared by giving a single dose of each to groups of 10 chicks at 4 and 9 days of age. The fecal

excretion of the phage was consequently monitored for 24 hours post-gavage. Figures 5.4 and

5.5 showed that no significant differences in passage time were found, with the majority of the

phage dose passing out of the chicks in under 3 hours. This rapid rate of passage of a single dose

was similar in 4 and 9 day-old chicks and suggests that orally delivered phage would have

reached locations of bacterial colonization inside the chick intestine within this time frame. The

excretion pattern of viable phage in the encapsulated phage group indicated that phage was

released successfully and rapidly inside the chicken. Although the FP group showed that average

PFU/g feces was higher than EP by about 1-1.5 log10, further treatment of samples from the EP

group with MBS solution (used to dissolve beads) did not yield higher phage count and no beads

were visible in the collected feces leading us to believe that beads were disintegrated and phage

released in the body.

5.4.2 Trial #1 discussion

CT1 was identified previously to be strongly lytic against STDT104 NalR in vitro and so was

anticipated to be highly effective in vivo. CT1 was tested in EP and in FP forms gavaged to

chicks experimentally infected with STDT104 NalR over a period of 7 days. The results of Trial

#1 (Table 5.2) showed that CT1 failed to reduce colonization in the chick intestine contents and

146

tissue over the course of the study period with twice-daily oral phage administration. Lower

phage levels were found in spleen, ileum and ceca after 7 days of treatment compared with 2

days of treatment. Since in vitro results failed to translate to in vivo, it was hypothesised that

young chicks may have had anti-phage antibodies present from the yolk sac transferred from the

hen. Yolk sacs were collected from sacrificed chicks during Trial #1 and the material was

incubated with phage FO1 and F24 for 1 hour followed by dilution and plaque counting. Rather

than inhibiting plaque formation though, yolk incubation was found to enhance plaque formation

of FO1 on bacterial lawns grown on phage agar plates. This is possibly due to the presence of

rich minerals and other nutrients in yolk sac contents and no phage-inactivation activity was

detected from incubating FO1 with yolk sac contents of chicks, and instead a protective effect

was observed [97]. However, it was observed that while using distilled water (as control) as a

medium to incubate FO1with STDT104 that plaques became barely visible on the lawns, which

suggested that FO1 was missing a required co-factor for lytic activity/plaque formation normally

found in buffer solutions and in egg yolk. Phage F24 formed plaques readily even in distilled

water. FO1 adsorption to Salmonella was reported to be dependent on specific concentrations of

cations such as Na+ and Ca2+ [184] whereas other phages require ionic and/or organic cofactors

for optimum adsorption [90]. These requirements might not have been met in vivo, hence only

phage F24 of CT1 was active, reducing the effectiveness of the cocktail. The requirement of sea

salts (used in vitro) of certain vibrio phages for adsorption and lytic activity has been reported to

explain for the lack of efficacy in vivo [104]. In light of these requirements of FO1 adsorption

and lytic activity, it will be necessary to identify a replacement for FO1 in future phage cocktails

for in vivo use, whereas in other applications such as for diagnostic purposes [91] or food surface

treatment [99], the surrounding environment may contain sufficient concentrations of cofactors

for proper phage adsorption.

147

Trial #1 was also designed to follow the fate of phage administration in liquid and encapsulated

forms after days of continued administration to determine if phage would accumulate in the

tissues in the absence of susceptible host bacteria. Results of FP and EP controls (first 2 columns

in Table 5.5) showed that there was little to no viable phage accumulation in liver and spleen.

Fairly low levels remained in the ceca of between 1.9-5.8 log10 PFU/g digesta. It appears that the

high initial doses of phage given (~1010) were not fully recoverable in the examined areas of

chicken and suggests viable phages are lost by some other unknown mechanism or possibly the

chicken immune system was involved in clearing of phage from the gut resulting in little to no

phage being detected after 7 days of phage administration.

After Trial #1 was concluded, our group isolated a new phage mixture YJ104 from a local

municipal waste water treatment plant that was lytic to STDT104. Initial in vitro screening

(Figure 5.2 and 5.3) showed that YJ104 was effective on its own, but when combined with CT1

yielded improved in vitro lytic activity due to the increased variety of phages that were lytic to

the same bacterium being present. Other studies have shown that inclusion of multiple lytic

phages against one bacterial strain can be more effective and at the same time reduce the

possibility of resistance [112, 114, 117, 174]. Hence, YJ104 was combined with CT1 to make a

potentially more effective cocktail (CT2) and both cocktails were evaluated in Trial #2.

5.4.3 Trial #2 discussion

Trial #2 was performed with increased number replicates (11-13) per treatment group to test the

effects of phage treatment (FP-CT1, EP-CT1, and EP-CT2) and number of days of continuous

148

treatment (2 and 7 days) on STDT104 NalR colonization in the same type of broiler chicks. The

increased number of replicates also allowed better statistical power to detect significant

differences resulting from the phage treatments. Chicks were infected individually with the same

strain used in Trial #1 of STDT104 NalR at 2 days of age and phage treatment was started 2 hrs

post-infection. Chicks were sacrificed after 2 and 7 days of daily phage administration. Results

after 2 days showed no significant reductions of STDT104 NalR took place yet. After 7 days of

treatment, the improved cocktail EP-CT2 was able to reduce the STDT104 NalR levels by 1.3-2.1

logs within three intestinal regions: ileum, cecum and colon, when compared to the control

group. By reducing colonization in these distal regions of the chicken digestive tract there would

likely be decreased shedding to the environment via the feces. On the other hand, treatment with

EP-CT1 was effective in lowering STDT104 NalR in the ileum region only and shedding likely is

not significantly affected by this phage cocktail. The range of reduction found with EP-CT2 in

our study is comparable with what others have achieved by phage therapy against Salmonella

Typhimurium in vivo of 0.3-1.3 [110] and ~2.2 [177] log10 units in the ceca of young broiler

chicks.

5.4.4 Effect of treatment length of time and multiple dosing, timing of administration

Another important result of the 2-way ANOVA of Trial #2 was the presence of significant

interaction (p <0.05) between days of treatment and phage treatment group meaning both the

phage cocktail used and the treatment days were important in producing the reduction or

“difference in means”. This is reflected in what the data in Table 6 shows, that is the longer

treatment duration for 7 days with CT2 was more effective than 2 days treatment with CT2

149

suggesting phage lytic activity takes longer to achieve reductions in vivo than in vitro where the

damage towards bacterial growth is typically seen within hours. This difference could be

because more rapidly dividing bacteria are present in growth media whereas less active and high

density bacteria colonize the animals’ gut. Similar to our results the study by Bardina et al. [174]

showed that multiple dosing for longer periods resulted in significant reductions in the ceca and

their use of a cocktail of phages likely prevented resistance development. In contrast, the

findings by Filho et al. [114] were that only short-lived reductions (less than 2 days) occurred in

the ceca with their phage and the subsequent loss of further lytic activity was thought to be

caused by resistance to phage.

While the less effective CT1 contained two phages (FO1 and F24) lytic towards STDT104 NalR,

the more effective CT2 contained four lytic phages, each with distinct plaque morphologies

(YJ104+FO1+F24), providing further evidence that increasing the number of distinct phages

specific to one bacterial strain can be more effective for lowering levels of a specific pathogen in

vivo. Even though broad host range phages can be used to target more than one strain of

Salmonella [177], rapid resistance can develop, which necessitates the use of a cocktail of

strongly lytic phages. Phage cocktails should be optimized for strong lytic activity before being

used and is one of the most important steps in any successful phage therapy for colonization

reduction [85, 91, 117].

5.4.5 Phage concentration in vivo MOI

Average bacteriophage concentrations in vivo have been observed to naturally decrease over time

in other studies [110] or as a result of bacterial host dropping below a threshold (around 104- 106

CFU per mL) [97, 106]. However, our results from two separate Trials show that the phage

150

concentration decreased with time in both the absence (Trial #1) and presence of host (Trial #1

and #2) after 7 days. This suggests that phage amplification did not occur in vivo to produce

significantly higher concentrations of phage even in the presence of high host density (e.g. in the

ceca). Although the MOI is easily controlled in in vitro experiments it appears difficult to

achieve a MOI of even 10 inside the intestinal contents of chickens, an optimal ratio to achieve

substantial bacterial killing as discussed by Abedon [95]. Even with the chicks receiving close to

~1010 PFU, twice daily, the concentrations found inside the chicks’ tissues and digestive tract

were several orders of magnitude lower. Therefore, phage effectiveness in vivo and specifically

in the alimentary tract could be limited by MOIs slightly above and below 1 and did not exceed

10, evidenced from our results. Some studies performed in mice [117] have demonstrated that

phage amplification occurred in vivo when host bacteria were present while a study in humans

[98] given low oral doses (~105 PFU/mL) showed that T4 phage was not substantially amplified

in the human gut even in presence of host bacteria, while other modes of administration like

intracranial, intramuscular injection have been observed to yield phage amplification in the blood

[97].

5.4.6 Future outlook of phage therapy

In recent reviews and studies [6, 117, 185] experimenting with phage for pathogen reduction

(such as E. coli, Campylobacter jejuni and Salmonella) in food animals, the common outcome

was that pathogen reduction can occur to some extent reducing mortality and/or bacterial load

depending on specific phage(s) used, but total elimination of pathogen from animals was rarely

reported and remains to be achieved reproducibly with phage therapy. Several common factors

identified from successful trials include: careful selection of strongly lytic phages, use of

151

cocktails to reduce resistance, a mode of administration that preserves phage viability, and

continuous frequent dosing for extended times at sufficiently high titers. While the main factors

that were found to reduce the success of phage treatments were: development of resistance to

phage attack leading to continued bacterial growth in the presence of phage, inability of phage to

replicate to high MOI levels or attack bacteria in vivo (due to physical barriers such as mucosal

membrane, intracellular invasion, etc.), and phage inactivation pathways may be at work causing

non-specific binding to digesta or host immune system producing phage-neutralizing antibodies

[97, 109]. Additionally, there is a potential “side-effect” that phage treatment can lead to which

is the loss of virulence factors by host bacteria through mutation/adaptation to become resistant

to phage attack. For example, after phage treatment some E. coli pathogens were found to be

much less invasive towards the host animals after losing specific surface antigens that were

targeted by the phage [101], but this type of outcome needs to be clearly demonstrated and

studied for specific Salmonella Typhimurium DT104 and phage combinations [91, 117]. If the

virulence of pathogens do sufficiently drop after treating with phages then total elimination from

the animal host may not be absolutely necessary.

Our results show a potential future for phage therapy to control Salmonella in chickens. In the

future it may be more effective to use oral phage therapy in conjunction with another antibiotic

alternative, such as essential oils, organic acids or competitive exclusion bacteria [186] to

improve the pathogen reduction since they are agents that affect bacteria via different

mechanisms. Encapsulation can allow simple combination of phage cocktails with these

alternatives nowadays as the methods can be used for many types of chemical and biological

materials. They can be co-encapsulated with phage or encapsulated separately and mixed

afterwards in desired proportions. Dry powdered encapsulated phages are a low cost, low

152

maintenance in-feed delivery system since phages in this form are more easily handled and

stored in farm settings as compared to a wet phage solution. Phage viability is maintained as well

through encapsulation and use of a suitable protectant after drying whereas a phage solution

added directly to feed is unstable and around 2 log10 PFU is lost either through adsorption to feed

[110] or denatured by exposure to drying, temperature, and other potentially phage-deactivating

conditions [176]. A dry, powder product can be more easily handled and distributed within

animal feed and when phage is stabilized in the dry form it becomes a stable phage formulation

during storage and administration [97].

Our results show that bacteriophages encapsulated into Ca-alginate-whey protein gel beads can

be effectively delivered to the lower intestine of chickens to significantly reduce Salmonella

levels after receiving continuous oral administration. The reductions were dependent on the

length of treatment and the specific phages used (CT2). Once further work to preserve phages

after drying with a suitable sugar protectant (e.g. maltose or trehalose) is achieved, the

encapsulated phage we have developed can be a convenient way to continuously administer

phage to livestock animals to help reduce pathogen levels and reduce shedding. The cocktail of

phages can be further improved upon by isolation, characterization and addition of new phages

lytic to STDT104 NalR.


The authors would like to thank OMAFRA Food Safety Program (FS 070724), Poultry Industry

Council, and Agriculture and Agri-Food Canada for funding of this research and Victoria Nowell

for providing technical assistance in carrying out phage screening and feed incubation studies.

153

Funding information

The authors would like to thank Agriculture and Agri-Food Canada, Poultry Industry Council,

and Ontario Ministry of Agriculture and Rural Affairs (Project #FS070724) for funding the

research.

154

Chapter 6. Conclusions and Future Research Discussion

6.1 Original Contributions and Conclusions from this Thesis

1) Developed formulation and method to prepare EO loaded granules. For the first

time, a model liquid EO compound trans-cinnamaldehyde was processed into core

granules with high loading using a novel formulation and simple preparation method with

pharmaceutical ingredients. The oil loading was high (~48%) and had long term stability

of up to 1 year, and the formulation is easily amenable to encapsulating other types of

similar EO’s/compounds.

2) Prepared coated granules with pH dependent release. EO granules were coated

successfully by fluid bed with an enteric polymer (Eudragit L) to impart pH dependent

release at above pH 6, while the stability of the active compound was maintained by in

the coated form upon storage. Final coated granules were suitable for oral in-feed

delivery to livestock animals, including chickens and pigs for evaluating the delivery

performance.

3) Evaluated the coated granule formulation in animal studies. When tested in both

chicken and pig trials, higher average and peak concentrations of CIN were delivered to

the lower sections of the animal GIT by the coated granule formulation as compared to

free oil in feed which undergoes rapid absorption in the upper GIT. Concentrations

achieved by coated granules in the jejunum and ileum of chickens were >100X the free

oil at various time points post-gavage.

4) Determined the temporal distribution of oral encapsulated and free phage in chicks.

Model phage Felix O1 was encapsulated and gavaged to chicks to investigate the transit

time and temporal distribution in the GIT and appearance in the feces, to better guide the

155

dosing interval for animal trials. Conditions thought to have affected phage survival were

investigated but they did not significantly degrade the ability of phage to infect the host

bacteria. Passage of the dose was found to be delayed and prolonged beyond 4 hours,

which suggested that multiple dosing was required to maintain high concentrations at the

lower GIT.

5) Optimized phage cocktail and encapsulation formulation in vitro. In vitro studies

with individual phage and combinations enabled improved antibacterial activity by

combinations of phages as opposed to single phage against the target bacteria strain. This

led to selection of phage cocktails with strong antibacterial activity against STDT104NalR

for encapsulation into microbeads. The alginate-whey protein formulation was improved

to allow faster phage release in the intestine of chicks, through modifying the MW of the

alginate component and whey protein content used.

6) Demonstrated efficacy of oral administration of phage cocktail in broiler chicks.

The optimized phage cocktail when administered orally, twice daily was effective at

lowering the pathogen (Salmonella Typhimurium DT104) levels in chicks after 7 days of

continuous administration with feed. Levels of STDT104 NalR were lowered by 2.06,

1.25, and 1.99 log10CFU units in the ileum, ceca, and colon of treated chicks as

compared with control, respectively. The results showed that Salmonella fecal shedding

can be significantly reduced by oral encapsulated phage cocktail treatment.

7) Studied the antibacterial effectiveness of CIN and bacteriophage in vitro and in

digesta. The antibacterial concentrations (MBC) of EO’s needed in the digestive tract

were higher than in liquid media and dependent on the digesta type, caused by the

presence of a complex food matrix undergoing digestion. Bacteria grew less actively in

(cecal) digesta than under liquid broth conditions. This effect was also observed during

156

testing of the antibacterial activity of phages against STDT104 NalR in digesta, which

demonstrated that digesta from different sections have disparate effects on the growth of

the bacteria that could affect activity of other antimicrobial agents in vivo as well and

warrants further investigation.

Overall, the work in this thesis has shown that through proper selection of excipients and

formulation optimization, the enhanced delivery of antibiotic alternative compounds was

achieved to potentially target gut pathogens in livestock. Furthermore, due to their distinct

mechanism of action, these antibiotic alternatives can potentially co-administered to improve the

antibacterial activity in vivo.

6.2 Discussion of Limitations of Thesis Work and Future Research Directions

1) Preparation of coated granules

Although fluid-bed spray coating is widely used in coating solid dosage forms, the current fatty

acid-based granules and the liquid nature of the active likely allowed some escape and migration

to the coating layers during the long processing time to completion. A lower temperature could

evaluated to prevent softening during coating, or more rapid coating methods such as pan coating

with a powder based enteric coating material may be more suitable for as well. The barrier

properties of the sub-coating need improvement to prevent escape of the volatile EO during

enteric coating and long-term storage as well as preventing release under acidic conditions. The

loading of CIN in granules need be tested from different particle sizes before and after coating to

observe if there is any difference in relation to particle size and drug loading, in addition to the

effectiveness of the barrier properties of the coatings.

157

2) Preventing de-mixing of granules with feed

During multi-particulate delivery through inclusion with animal feed (pig feed) there was

separation of the granules from the feed, which collected at the bottom of the feed trough. A

method to ensure a consistent concentration and uniform distribution among the feed particles is

needed to obtain more reliable data for future studies of delivery to different parts of the GIT at

various time points, which can reduce the variability of results between animals when sampled at

desired time points post feed intake. Over time, the steady-state concentrations can also be

observed. Better mixing can be achieved by using a binder solution or material that can allow

coated granules to stick to the larger feed particles during mixing with the animal feed.

3) Phage as oral bio-control agents in broiler chicks

Although our experimental trial showed that phage cocktail was effective in broiler chicks under

small scale, isolated conditions, the efficacy of the encapsulated phage cocktail needs to be

assessed at farm/production-scale to assess effectiveness of phage cocktail CT2 in the outdoor

environment to increase the confidence and reliability of results. Also the effect of the phage

cocktail against the Salmonella should be assessed at a later age (closer to slaughter), when the

chicken immune system is more mature which could aid in controlling the Salmonella levels in

the animals.

4) Oral administration to livestock and length of study for passage kinetics and distribution

in broiler chicks

In the gavaging of young broiler chicks with the formulation, the initial expectation that the

gavaged dose would have completely passed through the alimentary tract within 4 hours was

underestimated, and longer time points (6-8 hours) should be taken in future experiments. As the

158

experimental results have shown in the EO and phage trials that even after 3 hours, a large

proportion of the dose still remained in the upper alimentary tract (crop, gizzard). Indeed,

literature reviews have reported the common occurrence of the delayed movement of crop

contents further down the tract, which was dependent on the level of satiety, feeding rate, and

digestion of the broiler chickens which were not taken into account. These natural processes

appeared to have delayed passage of the given dose as chicks had access to feed ad libitum. The

crop of chickens normally acts as a reservoir for long term storage of food in order to provide

longer term energy needs through segmented passage and digestion, in contrast to mammalian

stomachs which immediately begin to digest the contents. This difference can result in division

of a multi-particulate dose into smaller portions thus limiting the potential peak concentrations

attainable from a given dose.

5) Understanding the mmechanism of action of phage cocktail

There is potential to control specific pathogens with phage cocktails, but more research is

necessary to elucidate the basic biology of individual phages used in cocktails and their

effects/interactions at the cellular level of host bacteria. Even though a limited number of phages

have been whole-genome sequenced, much remains unknown about the normal functions of the

genes. Researchers are only beginning to elucidate the function of some of the sequences and

their gene products, while large parts still remain undetermined. The next step would be

sequencing of the phages to better understand the genetic differences between the phages used in

the cocktail. It will also be prudent to study the potential for development of host resistance to

infection by the phages contained in the cocktail.

6) Formulation of encapsulated gel beads

159

The next step in the phage encapsulation formulation is to obtain dried alginate-whey protein

beads to improve the preservation of phage viability through inclusion of specific types of sugars

as protectants (e.g. maltose, trehalose) prior to drying of the gel beads which will enable better

storage and handling convenience.

160

References

1. Zhanel, G.G., et al., Prevalence of antimicrobial-resistant pathogens in Canadian

hospitals: results of the Canadian Ward Surveillance Study (CANWARD 2008).

Antimicrob Agents Chemother, 2010. 54(11): p. 4684-93.

2. Laxminarayan, R., et al., Antibiotic resistance-the need for global solutions. Lancet Infect

Dis, 2013. 13(12): p. 1057-98.

3. Maura, D. and L. Debarbieux, Bacteriophages as twenty-first century antibacterial tools

for food and medicine. Appl Microbiol Biotechnol, 2011. 90(3): p. 851-9.

4. Glynn, M.K., et al., Emergence of Multidrug-Resistant Salmonella enterica Serotype

Typhimurium DT104 Infections in the United States. The New England Journal of

Medicine, 1998. 338(19): p. 1333-1339.

5. Bell, B.P., et al., A multistate outbreak of Escherichia coli O157:H7-associated bloody

diarrhea and hemolytic uremic syndrome from hamburgers. The Washington experience.

JAMA: The Journal of the American Medical Association, 1994. 272(17): p. 1349-1353.

6. Goodridge, L.D., Application of bacteriophages to control pathogens in food animal

production, in Bacteriophages in the control of food-and waterborne pathogens, P. Sabour

and M. Griffiths, Editors. 2010. p. 61-77.

7. Rangel, J.M., et al., Epidemiology of Escherichia coli O157:H7 outbreaks, United States,

1982-2002. Emerging Infectious Diseases, 2005. 11(4): p. 603-609.

8. Amezcua, R., et al. Presentation of postweaning Escherichia coli diarrhea in southern

Ontario, prevalence of hemolytic E. coli serogroups involved, and their antimicrobial

resistance patterns. Can J Vet Res. 2002. 66, 73.

161

9. Blood, D. and J.A. Henderson, Diseases caused by Bacteria (CH. 17), in Veterinary

Medicine 1968, Williams & Wilkins Company: Baltimore, USA. p. 343-388.

10. Muirhead, M.R. and T.J.L. Alexander, Managing pig health and the treatment of disease:

A reference for the farm, 1997, 5M Enterprises Ltd: Sheffield, UK. p. Ch. 1,2,9-Ch. 1,2,9.

11. Threlfall, E.J., et al., The emergence and spread of antibiotic resistance in food-borne

bacteria. International Journal of Food Microbiology, 2000. 62(1-2): p. 1-5.

12. Mølbak, K., et al., An Outbreak of Multidrug-Resistant, Quinolone-Resistant Salmonella

enterica Serotype Typhimurium DT104. The New England Journal of Medicine, 1999.

341(19): p. 1420-1425.

13. Cornick, N.A. and A.F. Helgerson, Transmission and Infectious Dose of Escherichia coli

O157:H7 in Swine. Applied and Environmental Microbiology, 2004. 70(9): p. 5331-5335.

14. Acar, J.F. and G. Moulin Antimicrobial resistance at farm level. Revue Scientifique et

Technique. 2006. 25, 775.

15. van den Bogaard, A.E. and E.E. Stobberingh, Epidemiology of resistance to antibiotics:

Links between animals and humans. International Journal of Antimicrobial Agents, 2000.

14(4): p. 327-335.

16. McEwen, S.A. and P.J. Fedorka-Cray, Antimicrobial Use and Resistance in Animals.

Clinical Infectious Diseases, 2002. 34(Journal Article): p. S93-S106.

17. Doyle, M.E., Alternatives to antibiotic use for growth promotion in animal husbandry,

2001, Food Research Institute: USA. p. 1-17.

18. Steiner, T., Managing Gut Health: Natural Growth Promoters as a Key to Animal

Performance. 2006: Nottingham University Press.

162

19. Lallès, J.-P., et al., Weaning : A challenge to gut physiologists. 10th International

Symposium on Digestive Physiology in Pigs, Denmark 2006, Part 1, 2007. 108(1-3): p.

82-93.

20. Pond, W.G. and K.A. Houpt, The biology of the pig. 1978, London, UK: Comstock

Publishing Associates. 371.

21. Scott, L., et al., Antimicrobial resistance in fecal generic Escherichia coli and Salmonella

spp. obtained from Ontario sheep flocks and associations between antimicrobial use and

resistance. Canadian Journal of Veterinary Research, 2012. 76(2): p. 109-119.

22. WHO, Antimicrobial resistance: global report on surveillance, 2014, World Health

Organization.

23. Nolen, R.S., Lawmakers seek to curb antimicrobial use in livestock. Journal of the

American veterinary medical association (online News Archives), 2003 (September 15).

24. Rugbjerg, H., et al., Estimating the number of undetected multi-resistant Salmonella

Typhimurium DT104 infected pig herds in Denmark. Preventive Veterinary Medicine,

2004. 65(3-4): p. 147-171.

25. Khachatourians, G.G., Agricultural use of antibiotics and the evolution and transfer of

antibiotic-resistant bacteria. Canadian Medical Association Journal, 1998. 159(9): p.

1129-1136.

26. Dantas, G. and M.O.A. Sommer, How to Fight Back Against Antibiotic Resistance.

American Scientist, 2014. 102(1): p. 42-51.

27. Rosenquist, H., et al., Quantitative risk assessment of human campylobacteriosis

associated with thermophilic Campylobacter species in chickens. Int J Food Microbiol,

2003. 83(1): p. 87-103.

163

28. Huang, J., et al., New Delhi Metallo-β-Lactamase-1 in Carbapenem-Resistant Salmonella

Strain, China. Emerging Infectious Diseases, 2013. 19(12): p. 2049.

29. McArthur, A.G., et al., The Comprehensive Antibiotic Resistance Database.

Antimicrobial Agents and Chemotherapy, 2013. 57(7): p. 3348-3357.

30. Hooper, D.C., Efflux pumps and nosocomial antibiotic resistance: a primer for hospital

epidemiologists. Clin Infect Dis, 2005. 40(12): p. 1811-7.

31. Spellberg, B., et al., The epidemic of antibiotic-resistant infections: a call to action for the

medical community from the Infectious Diseases Society of America. Clin Infect Dis,

2008. 46(2): p. 155-64.

32. Atterbury, R.J., et al., Bacteriophage therapy to reduce Salmonella colonization of broiler

chickens. Appl Environ Microbiol, 2007. 73(14): p. 4543-9.

33. Ferber, D., Antibiotic resistance. WHO advises kicking the livestock antibiotic habit.

Science, 2003. 301(5636): p. 1027.

34. Kahler, S.C., Eliminate nontherapeutic antimicrobial use in agriculture: AMA. Journal of

the American veterinary medical association (online News Archives), 2001. August 1

(Journal Article).

35. Namkung, H., et al., Impact of feeding blends of organic acids and herbal extracts on

growth performance, gut microbiota and digestive function in newly weaned pigs.

Canadian Journal of Animal Science, 2004. 84(4): p. 697-704.

36. Roselli, M., et al., Alternatives to in-feed antibiotics in pigs: evaluation of probiotics, zinc

or organic acids as protective agents for the intestinal mucosa. A comparison of in vitro

and in vivo results. Animal Research, 2005. 54(3): p. 203-218.

164

37. Fairbrother, J.M., E. Nadeau, and C.L. Gyles, Escherichia coli in postweaning diarrhea in

pigs: an update on bacterial types, pathogenesis, and prevention strategies. Anim Health

Res Rev. 2005. 6, 17.

38. Snoeck, V., et al., Influence of porcine intestinal pH and gastric digestion on antigenicity

of F4 fimbriae for oral immunisation. Veterinary Microbiology, 2004. 98(1): p. 45-53.

39. Burt, S., Essential oils: their antibacterial properties and potential applications in foods--a

review. Int J Food Microbiol, 2004. 94(3): p. 223-53.

40. Carvalho, C.M., et al., The in vivo efficacy of two administration routes of a phage

cocktail to reduce numbers of Campylobacter coli and Campylobacter jejuni in chickens.

BMC Microbiol, 2010. 10: p. 232.

41. Si, W., et al., Antimicrobial activity of essential oils and structurally related synthetic

food additives towards selected pathogenic and beneficial gut bacteria. Journal of

Applied Microbiology, 2006. 100(2): p. 296-305.

42. Chorianopoulos, N., et al., Essential Oils of Satureja, Origanum, and Thymus Species:

Chemical Composition and Antibacterial Activities Against Foodborne Pathogens.

Journal of Agricultural and Food Chemistry, 2004. 52(26): p. 8261-8267.

43. Angienda, P.O., D.M. Onyango, and D.J. Hill, Potential application of plant essential oils

at sub-lethal concentrations under extrinsic conditions that enhance their antimicrobial

effectiveness against pathogenic bacteria. African Journal of Microbiology Research,

2010. 4(16): p. 1678-1684.

44. Bakkali, F., et al., Biological effects of essential oils - A review. Food and Chemical

Toxicology, 2008. 46(2): p. 446-475.

165

45. Baser, K.H.C. and C. Franz, Essential oils used in veterinary medicine, in Handbook of

essential oils: science, technology, and applications, K.H.C. Baser and G. Buchbauer,

Editors. 2009, CRC Press. p. 881-894.

46. Vaughn, S.F., Plant volatiles, in: Encyclopedia of life sciences, 2001, Wiley Interscience

- John Wiley & Sons, Ltd: London ; New York.

47. Lambert, R.J.W., et al., A study of the minimum inhibitory concentration and mode of

action of oregano essential oil, thymol and carvacrol. Journal of Applied Microbiology,

2001. 91(3): p. 453-462.

48. Burt, S.A. and R.D. Reinders, Antibacterial activity of selected plant essential oils against

Escherichia coli O157:H7. Letters in Applied Microbiology, 2003. 36(3): p. 162-167.

49. Zhou, F., et al., The antibacterial effect of cinnamaldehyde, thymol, carvacrol, and their

combinations against the foodborne pathogen Salmonella Typhimurium. Journal of Food

Safety, 2007. 27(2): p. 124-133.

50. Roller, S., Natural antimicrobials for the minimal processing of foods, 2003: Elsevier.

51. Brul, S. and P. Coote, Preservative agents in foods: Mode of action and microbial

resistance mechanisms. International Journal of Food Microbiology, 1999. 50(1-2): p. 1-

17.

52. Davidson, P.M., J.N. Sofos, and A.L. Branen, Antimicrobials in food, 2010: CRC Press.

53. Helander, I.M., et al., Characterization of the Action of Selected Essential Oil

Components on Gram-Negative Bacteria. Journal of Agricultural and Food Chemistry,

1998. 46(9): p. 3590-3595.

54. Dorman, H.J.D. and S.G. Deans, Antimicrobial agents from plants: antibacterial activity

of plant volatile oils. Journal of Applied Microbiology, 2000. 88(2): p. 308-316.

166

55. Gill, A.O. and R.A. Holley, Mechanisms of Bactericidal Action of Cinnamaldehyde

against Listeria monocytogenes and of Eugenol against L. monocytogenes and

Lactobacillus sakei. Applied and Environmental Microbiology, 2004. 70(10): p. 5750-

5755.

56. Smith-Palmer, A., J. Stewart, and L. Fyfe, Antimicrobial properties of plant essential oils

and essences against five important food-borne pathogens. Lett Appl Microbiol, 1998.

26(2): p. 118-22.

57. Ooi, L.S., et al., Antimicrobial activities of cinnamon oil and cinnamaldehyde from the

Chinese medicinal herb Cinnamomum cassia Blume. Am J Chin Med, 2006. 34(3): p.

511-22.

58. Michiels, J., et al., In vitro degradation and in vivo passage kinetics of carvacrol, thymol,

eugenol and trans-cinnamaldehyde along the gastrointestinal tract of piglets. Journal of

the Science of Food and Agriculture, 2008. 88(13): p. 2371.

59. Knight, K.P. and R.C. McKellar, Influence of cinnamon and clove essential oils on the D-

and z-values of Escherichia coli O157:H7 in apple cider. J Food Prot, 2007. 70(9): p.

2089-94.

60. Valero, M. and M.J. Giner, Effects of antimicrobial components of essential oils on

growth of Bacillus cereus INRA L2104 in and the sensory qualities of carrot broth.

International Journal of Food Microbiology, 2006. 106(1): p. 90-94.

61. Chang, S.T., P.F. Chen, and S.C. Chang, Antibacterial activity of leaf essential oils and

their constituents from Cinnamomum osmophloeum. Journal of Ethnopharmacology,

2001. 77(1): p. 123-127.

62. Si, W., et al., In vitro assessment of antimicrobial activity of carvacrol, thymol and

cinnamaldehyde towards Salmonella serotype Typhimurium DT104: effects of pig diets

167

and emulsification in hydrocolloids. Journal of Applied Microbiology, 2006. 101(6): p.

1282-1291.

63. Michiels, J., et al., In vitro dose-response of carvacrol, thymol, eugenol and trans-

cinnamaldehyde and interaction of combinations for the antimicrobial activity against the

pig gut flora. 10th International Symposium on Digestive Physiology in Pigs, Denmark

2006, Part 2, 2007. 109(1-3): p. 157-160.

64. Shelef, L.A., Antimicrobial effects of spices. Journal of Food Safety, 1984. 6(1): p. 29.

65. Michiels, J., et al., Effects of dose and formulation of carvacrol and thymol on bacteria

and some functional traits of the gut in piglets after weaning. Arch Anim Nutr, 2010.

64(2): p. 136-54.

66. Petrujkic, B.T., et al., Ex vivo absorption of thymol and thymol-beta-D-glucopyranoside

in piglet everted jejunal segments. J Agric Food Chem, 2013. 61(15): p. 3757-62.

67. Kollanoor-Johny, A., et al., Reduction of Salmonella enterica Serovar Enteritidis

Colonization in 20-Day-Old Broiler Chickens by the Plant-Derived Compounds trans-

Cinnamaldehyde and Eugenol. Applied and Environmental Microbiology, 2012. 78(8): p.

2981-2987.

68. Dunkley, K.D., et al., Foodborne Salmonella ecology in the avian gastrointestinal tract.

Anaerobe, 2009. 15(1-2): p. 26-35.

69. Venkitanarayanan, K., et al., Use of plant-derived antimicrobials for improving the safety

of poultry products. Poultry Science, 2013. 92(2): p. 493-501.

70. Biester, H.E., Salmonella (CH. 23), in Diseases of swine, H.W. Dunne, Editor 1964, Iowa

State University Press: Ames. p. 897-897.

168

71. Jacob, J. and T. Pescatore. Chicken Anatomy and Physiology: Digestive System. [cited

2013 January ]; Available from:

http://www2.ca.uky.edu/agc/pubs/ASC/ASC203/ASC203.pdf.

72. Gauthier, R., Organic acids and essential oils, a realistic alternative to antibiotic growth

promoters in poultry. Animal world, 2005: p. 148-157.

73. Kararli, T.T., Comparison of the gastrointestinal anatomy, physiology, and biochemistry

of humans and commonly used laboratory animals. Biopharmaceutics & Drug

Disposition, 1995. 16(5): p. 351-380.

74. Ibekwe, V.C., et al., An investigation into the in vivo performance variability of pH

responsive polymers for ileo-colonic drug delivery using gamma scintigraphy in humans.

J Pharm Sci, 2006. 95(12): p. 2760-6.

75. Aqil, M.. et al., Status of terpenes as skin penetration enhancers. Drug Discovery Today.

2007. 12:p. 1061-1067.

76. Greenblatt, H.C., et al., Encapsulation and controlled release of flavours and fragrances,

in Encapsulation and controlled release, D.R. Karsa and R.A. Stephenson, Editors. 1993,

Royal Society of Chemistry: Cambridge. p. 148.

77. Madene, A., et al., Flavour encapsulation and controlled release - a review. International

Journal of Food Science & Technology, 2006. 41(1): p. 1-21.

78. Thies, C., A survey of microencapsulation processes, in Microencapsulation : methods

and industrial applications, S. Benita, Editor 1996, Marcel Dekker: New York. p. 1-20-

640.

79. Moretti, M.D., et al., Essential oil formulations useful as a new tool for insect pest control.

AAPS PharmSciTech, 2002. 3(2): p. E13.

http://www2.ca.uky.edu/agc/pubs/ASC/ASC203/ASC203.pdf

169

80. Ubrich, N. and P. Maincent, Recent advances in heparin delivery, in Microencapsulation:

Methods and Industrial Applications, S. Benita, Editor 2006, Taylor & Francis: New

York. p. 481-520.

81. Mehnert, W. and K. Mader, Solid lipid nanoparticles: production, characterization and

applications. Adv Drug Deliv Rev, 2001. 47(2-3): p. 165-96.

82. Ukita, K. and T. Murakami, Preparation of essential oils loaded granule by melt

granulation. Drug Development and Industrial Pharmacy, 1994. 20(6): p. 981-992.

83. Meunier, J.P., et al., Use of rotary fluidized-bed technology for development of

sustained-release plant extracts pellets: Potential application for feed additive delivery.

Journal of Animal Science, 2006. 84(7): p. 1850-1859.

84. Chourasia, M.K. and S.K. Jain, Pharmaceutical approaches to colon targeted drug

delivery systems. Journal of Pharmacy and Pharmaceutical Sciences, 2003. 6(1): p. 33-66.

85. Sulakvelidze, A., Z. Alavidze, and J.G. Morris, Bacteriophage Therapy. Antimicrobial

Agents and Chemotherapy, 2001. 45(3): p. 649-659.

86. Dabrowska, K., et al., Bacteriophage penetration in vertebrates. Journal of applied

microbiology, 2005. 98(1): p. 7-13.

87. Brüssow, H. and E. Kutter, Phage ecology. in Bacteriophages: biology and applications,

2005: p. 129-163.

88. Moineau, S. and C. Lévesque, Control of bacteriophages in industrial fermentations. in

Bacteriophages: biology and applications, 2005: p. 285-296.

89. Ryan, K.J., C.G. Ray, and J.C. Sherris, Sherris Medical Microbiology: an Introduction to

Infectious Diseases. 4th ed. McGraw-Hill, USA. 2004.

90. Adams, M.H., Bacteriophages, 1959, Interscience Publishers. p. 49-62, 137-160.

170

91. Kropinski, A.M., et al., Salmonella phages and prophages—genomics and practical

aspects, in Salmonella. 2007, Springer. p. 133-175.

92. Wagner, P.L. and M.K. Waldor, Bacteriophage control of bacterial virulence. Infect

Immun, 2002. 70(8): p. 3985-93.

93. Ly-Chatain, M.H., The factors affecting effectiveness of treatment in phages therapy.

Front Microbiol, 2014. 5: p. 51.

94. Payne, R.J., D. Phil, and V.A. Jansen, Phage therapy: the peculiar kinetics of self-

replicating pharmaceuticals. Clin Pharmacol Ther, 2000. 68(3): p. 225-30.

95. Abedon, S.T., Kinetics of phage-mediated biocontrol of bacteria. Foodborne Pathogens

and Disease, 2009. 6(7): p. 807-815.

96. Whichard, J.M., et al., Complete genomic sequence of bacteriophage felix o1. Viruses,

2010. 2(3): p. 710-30.

97. Brüssow, H., Phage therapy: the Escherichia coli experience. Microbiology, 2005. 151(7):

p. 2133-2140.

98. Bruttin, A. and H. Brussow, Human volunteers receiving Escherichia coli phage T4

orally: a safety test of phage therapy. Antimicrobial Agents and Chemotherapy, 2005.

49(7): p. 2874-2878.

99. Whichard, J.M., N. Sriranganathan, and F.W. Pierson, Suppression of Salmonella growth

by wild-type and large-plaque variants of bacteriophage Felix O1 in liquid culture and on

chicken frankfurters. Journal of Food Protection, 2003. 66(2): p. 220-225.

100. Sarker, S.A., et al., Oral T4-like phage cocktail application to healthy adult volunteers

from Bangladesh. Virology, 2012. 434(2): p. 222-32.

171

101. Smith, H.W. and M.B. Huggins, Effectiveness of phages in treating experimental

Escherichia coli diarrhoea in calves, piglets and lambs. J Gen Microbiol, 1983. 129(8): p.

2659-75.

102. Smith, H.W., M.B. Huggins, and K.M. Shaw, Factors influencing the survival and

multiplication of bacteriophages in calves and in their environment. J Gen Microbiol,

1987. 133(5): p. 1127-35.

103. Jaiswal, A., et al., Efficacy of cocktail phage therapy in treating Vibrio cholerae infection

in rabbit model. Microbes Infect, 2013. 15(2): p. 152-6.

104. Cerveny, K.E., et al., Phage therapy of local and systemic disease caused by Vibrio

vulnificus in iron-dextran-treated mice. Infection and Immunity, 2002. 70(11): p. 6251-

6262.

105. Callaway, T.R., et al., Evaluation of phage treatment as a strategy to reduce Salmonella

populations in growing swine. Foodborne Pathog Dis, 2011. 8(2): p. 261-6.

106. Berchieri, A., M. Lovell, and P. Barrow, The activity in the chicken alimentary tract of

bacteriophages lytic for Salmonella Typhimurium. Research in Microbiology, 1991.

142(5): p. 541-549.

107. Huff, W.E., et al., Alternatives to antibiotics: utilization of bacteriophage to treat

colibacillosis and prevent foodborne pathogens. Poult Sci, 2005. 84(4): p. 655-9.

108. Huff, W.E., et al., Evaluation of the influence of bacteriophage titer on the treatment of

colibacillosis in broiler chickens. Poult Sci, 2006. 85(8): p. 1373-7.

109. Huff, W.E., et al., Immune interference of bacteriophage efficacy when treating

colibacillosis in poultry. Poult Sci, 2010. 89(5): p. 895-900.

172

110. Sklar, I.B. and R.D. Joerger, Attempts to utilize bacteriophage to combat Salmonella

enterica serovar Enteritidis infection in chickens. Journal of Food Safety, 2001. 21(1): p.

15-29.

111. Oliveira, A., et al., The influence of the mode of administration in the dissemination of

three coliphages in chickens. Poult Sci, 2009. 88(4): p. 728-33.

112. Fiorentin, L., N.D. Vieira, and W. Barioni, Jr., Oral treatment with bacteriophages

reduces the concentration of Salmonella Enteritidis PT4 in caecal contents of broilers.

Avian Pathol, 2005. 34(3): p. 258-63.

113. Borie, C., et al., Bacteriophage treatment reduces Salmonella colonization of infected

chickens. Avian Diseases, 2008. 52(1): p. 64-67.

114. Andreatti Filho, R., et al., Ability of bacteriophages isolated from different sources to

reduce Salmonella enterica serovar Enteritidis in vitro and in vivo. Poultry Science, 2007.

86(9): p. 1904-1909.

115. Loc Carrillo, C., et al., Bacteriophage Therapy To Reduce Campylobacter jejuni

Colonization of Broiler Chickens. Applied and Environmental Microbiology, 2005.

71(11): p. 6554-6563.

116. Harris, D.L. and N. Lee, Compositions and methods for reducing the amount of

Salmonella in livestock, US Patent #6656463. 2003, USPTO.

117. Johnson, R.P., et al., Bacteriophages for prophylaxis and therapy in cattle, poultry and

pigs. Animal Health Research Reviews, 2008. 9(Special Issue 02): p. 201-215.

118. Ma, Y., et al., Microencapsulation of Bacteriophage Felix O1 into Chitosan-Alginate

Microspheres for Oral Delivery. Applied and Environmental Microbiology, 2008. 74(15):

p. 4799-4805.

173

119. Dini, C., et al., Novel biopolymer matrices for microencapsulation of phages: enhanced

protection against acidity and protease activity. Macromol Biosci, 2012. 12(9): p. 1200-8.

120. Gombotz, W.R. and S.F. Wee, Protein release from alginate matrices. Advanced Drug

Delivery Reviews, 2012. 64, Supplement(0): p. 194-205.

121. Huguet, M.L., et al., Hemolglobin encapsulation in chitosan/calcium alginate beads.

Journal of Applied Polymer Science, 1994. 51(8): p. 1427-1432.

122. Chen, L. and M. Subirade, Alginate–whey protein granular microspheres as oral delivery

vehicles for bioactive compounds. Biomaterials, 2006. 27(26): p. 4646-4654.

123. Tang, Z., et al., Whey protein improves survival and release characteristics of

bacteriophage Felix O1 encapsulated in alginate microspheres. Food Research

International, 2013. 52(2): p. 460-466.

124. Maron, D.F., T.J. Smith, and K.E. Nachman, Restrictions on antimicrobial use in food

animal production: an international regulatory and economic survey. Global Health, 2013.

9: p. 48.

125. Ferber, D., WHO Advises Kicking the Livestock Antibiotic Habit. Science, 2003.

301(5636): p. 1027.

126. Michiels, J., et al., In vitro characterisation of the antimicrobial activity of selected

essential oil components and binary combinations against the pig gut flora. Animal Feed

Science and Technology, 2009. 151(1): p. 111-127.

127. Busquet, M., et al., Effects of cinnamaldehyde and garlic oil on rumen microbial

fermentation in a dual flow continuous culture. J Dairy Sci, 2005. 88(7): p. 2508-16.

128. Müller, R.H., K. Mäder, and S. Gohla, Solid lipid nanoparticles (SLN) for controlled drug

delivery - a review of the state of the art. European Journal of Pharmaceutics and

Biopharmaceutics, 2000. 50(1): p. 161-177.

174

129. Benita, S., Microencapsulation : methods and industrial applications. Drugs and the

Pharmaceutical Sciences ; v. 158. Vol. 2nd. 2006, New York: Taylor & Francis. 756-756.

130. Wong, H.L., et al., Chemotherapy with anticancer drugs encapsulated in solid lipid

nanoparticles. Adv Drug Deliv Rev, 2007. 59(6): p. 491-504.

131. Wong, H.L., et al., Development of solid lipid nanoparticles containing ionically

complexed chemotherapeutic drugs and chemosensitizers. J Pharm Sci, 2004. 93(8): p.

1993-2008.

132. Chattopadhyay, N., et al., Solid lipid nanoparticles enhance the delivery of the HIV

protease inhibitor, atazanavir, by a human brain endothelial cell line. Pharm Res, 2008.

25(10): p. 2262-71.

133. Voinovich, D., et al., Melt pelletization in high shear mixer using a hydrophobic melt

binder: influence of some apparatus and process variables. European Journal of

Pharmaceutics and Biopharmaceutics, 2001. 52(3): p. 305-313.

134. Hamdani, J., A.J. Moes, and K. Amighi, Development and evaluation of prolonged

release pellets obtained by the melt pelletization process. Int J Pharm, 2002. 245(1-2): p.

167-77.

135. Esposita, E., R. Cortesi, and C. Nastruzzi, Production of lipospheres for bioactive

compound delivery. Lipospheres in Drug Targets and Delivery: Approaches, Methods

and Applications. New York, NY: CRC, 2005: p. 23-40.

136. Amezcua, R., et al., Presentation of postweaning Escherichia coli diarrhea in southern

Ontario, prevalence of hemolytic E. coli serogroups involved, and their antimicrobial

resistance patterns. Can J Vet Res, 2002. 66(2): p. 73-8.

137. Sawyer, D.T., et al., Chemistry experiments for instrumental methods. Wiley, 1984.

175

138. Hili, P., C.S. Evans, and R.G. Veness, Antimicrobial action of essential oils : the effect of

dimethylsulphoxide on the activity of cinnamon oil. Letters in Applied Microbiology,

1997. 24(4): p. 269-275.

139. Westermarck, S., et al., Microcrystalline cellulose and its microstructure in

pharmaceutical processing. Eur J Pharm Biopharm, 1999. 48(3): p. 199-206.

140. Hentzschel, C.M., A. Sakmann, and C.S. Leopold, Suitability of various excipients as

carrier and coating materials for liquisolid compacts. Drug Development & Industrial

Pharmacy, 2011. 37(10): p. 1200-1207.

141. Friedman, M., N. Kozukue, and L.A. Harden, Cinnamaldehyde Content in Foods

Determined by Gas Chromatography-Mass Spectrometry. Journal of Agricultural and

Food Chemistry, 2000. 48(11): p. 5702-5709.

142. Cheng, S.S., et al., Chemical composition and mosquito larvicidal activity of essential

oils from leaves of different Cinnamomum osmophloeum provenances. J.Agric.Food

Chem, 2004. 52(14): p. 4395-4400.

143. Varel, V.H., D.N. Miller, and E.D. Berry, Incorporation of thymol into corncob granules

for reduction of odor and pathogens in feedlot cattle waste. Journal of Animal Science,

2006. 84(2): p. 481-487.

144. Elmore, A., Final report on the safety assessment of aluminum silicate, calcium silicate,

magnesium aluminum silicate, magnesium silicate, magnesium trisilicate, sodium

magnesium silicate, zirconium silicate, attapulgite, bentonite, Fuller's earth, hectorite,

kaolin, lithium magnesium silicate, lithium magnesium sodium silicate, montmorillonite,

pyrophyllite, and zeolite. International Journal of Toxicology, 2003. 22: p. 37.

145. Porter, S.C. and L.A. Felton, Techniques to assess film coatings and evaluate film-coated

products. Drug Development and Industrial Pharmacy, 2010. 36(2): p. 128-142.

176

146. Obara, S., et al., Dry coating: an innovative enteric coating method using a cellulose

derivative. European Journal of Pharmaceutics and Biopharmaceutics, 1999. 47(1): p. 51-

59.

147. Fairbrother, J.M., E. Nadeau, and C.L. Gyles, Escherichia coli in postweaning diarrhea in

pigs: an update on bacterial types, pathogenesis, and prevention strategies. Anim Health

Res Rev, 2005. 6(1): p. 17-39.

148. Jin, L. and X. Zhao, Intestinal receptors for adhesive fimbriae of enterotoxigenic

Escherichia coli (ETEC) K88 in swine–a review. Applied Microbiology and

Biotechnology, 2000. 54(3): p. 311-318.

149. Faria, N.C.G., et al., Enhanced activity of antifungal drugs using natural phenolics against

yeast strains of Candida and Cryptococcus. Letters in Applied Microbiology, 2011. 52(5):

p. 506-513.

150. Roller, S. and P. Seedhar, Carvacrol and cinnamic acid inhibit microbial growth in fresh-

cut melon and kiwifruit at 4° and 8°C. Letters in Applied Microbiology, 2002. 35(5): p.

390-394.

151. Kabara, J.J., et al., Fatty Acids and Derivatives as Antimicrobial Agents. Antimicrob.

Agents Chemother., 1972. 2(1): p. 23-28.

152. Fratini, F., et al., Antibacterial activity of essential oils, their blends and mixtures of their

main constituents against some strains supporting livestock mastitis. Fitoterapia, 2014. 96:

p. 1-7.

153. Ma, Y.H., et al., In vivo performance of essential oil granules for targeted delivery to the

animal GI tract. (unpublished results).

177

154. Yossa, N., et al., Antibacterial Activity of Cinnamaldehyde and Sporan against

Escherichia coli O157:H7 and Salmonella. Journal of Food Processing and Preservation,

2012. 38:749-757.

155. Friedman, M., P.R. Henika, and R.E. Mandrell, Bactericidal activities of plant essential

oils and some of their isolated constituents against Campylobacter jejuni, Escherichia

coli, Listeria monocytogenes, and Salmonella enterica. Journal of food protection, 2002.

65(10): p. 1545-1560.

156. Ma, Y., et al., Formulation of granules for site-specific delivery of an antimicrobial

essential oil to the animal intestinal tract. Journal of Pharmaceutical Sciences, 2016.

105(3):1124-1133.

157. Hovorka, S.W. and C. Schöneich, Oxidative degradation of pharmaceuticals: Theory,

mechanisms and inhibition. Journal of Pharmaceutical Sciences, 2001. 90(3): p. 253-269.

158. Vergara, P., et al., Factors determining gastrointestinal transit time of several markers in

the domestic fowl. Q J Exp Physiol, 1989. 74(6): p. 867-74.

159. Fedorka-Cray, P.J., et al., Colonization of broiler chicks by Salmonella Typhimurium

definitive phage type 104. J Food Prot, 2001. 64(11): p. 1698-704.

160. Hogue, A., et al. Situation Assessment: Salmonella Typhimurium DT104. 1997 [cited

2013 June 12]; Available from: http://www.fsis.usda.gov/OPHS/stdt104.htm.

161. Berge, A.C. and M. Wierup, Nutritional strategies to combat Salmonella in mono-gastric

food animal production. Animal, 2012. 6(04): p. 557-564.

162. Jordan, D., et al., Pre-slaughter control of Escherichia coli O157 in beef cattle: a

simulation study. Preventive Veterinary Medicine, 1999. 41(1): p. 55-74.

http://www.fsis.usda.gov/OPHS/stdt104.htm

178

163. Soon, J.M., S.A. Chadd, and R.N. Baines, Escherichia coli O157:H7 in beef cattle: on

farm contamination and pre-slaughter control methods. Animal Health Research Reviews,

2011. 12(02): p. 197-211.

164. Billington, C., J.A. Hudson, and E. D'Sa, Prevention of bacterial foodborne disease using

nanobiotechnology. Nanotechnol Sci Appl, 2014. 7: p. 73-83.

165. Kim, K.H., et al., Evaluation of Anti-SE Bacteriophage as Feed Additives to Prevent

Salmonella Enteritidis (SE) in Broiler. Asian-Australas J Anim Sci, 2013. 26(3): p. 386-

93.

166. Capparelli, R., et al., Bacteriophage therapy of Salmonella enterica: a fresh appraisal of

bacteriophage therapy. J Infect Dis, 2010. 201(1): p. 52-61.

167. Saez, A.C., et al., Direct feeding of microencapsulated bacteriophages to reduce

Salmonella colonization in pigs. Foodborne Pathog Dis, 2011. 8(12): p. 1269-74.

168. Wang, Q., et al., Development of encapsulated bacteriophage for effectively delivery to

chicken intestine for the control of food-borne Salmonella spp. Project Number SF6073,

2010, Annual Report for OMAFRA Competitive Research Programs. Ontario Ministry of

Agriculture, Food and Rural Affairs, Ontario, Canada.

169. Jin, S.-H., A. Corless, and J.L. Sell, Digestive system development in post-hatch poultry.

World's Poultry Science Journal, 1998. 54(04): p. 335-345.

170. Noy, Y. and D. Sklan, Digestion and absorption in the young chick. Poult Sci, 1995.

74(2): p. 366-73.

171. Chen, L. and M. Subirade, Effect of preparation conditions on the nutrient release

properties of alginate–whey protein granular microspheres. European Journal of

Pharmaceutics and Biopharmaceutics, 2007. 65(3): p. 354-362.

179

172. Anany, H., et al., A Shigella boydii bacteriophage which resembles Salmonella phage ViI.

Virol J, 2011. 8: p. 242.

173. Svihus, B., et al., Passage rate through the anterior digestive tract of broiler chickens fed

on diets with ground and whole wheat. Br Poult Sci, 2002. 43(5 Suppl): p. 662-8.

174. Bardina, C., et al., Significance of the bacteriophage treatment schedule in reducing

Salmonella colonization of poultry. Appl Environ Microbiol, 2012. 78(18): p. 6600-7.

175. Nale, J.Y., et al., Bacteriophage Combinations Significantly Reduce Clostridium difficile

Growth in vitro and Proliferation in vivo. Antimicrobial Agents and Chemotherapy, 2016.

60(2): p. 968-981.

176. Jończyk, E., et al., The influence of external factors on bacteriophages—review. Folia

Microbiologica, 2011. 56(3): p. 191-200.

177. Atterbury, R., et al., Bacteriophage therapy to reduce Salmonella colonization of broiler

chickens. Applied and Environmental Microbiology, 2007. 73(14): p. 4543-4549.

178. Wang, Q. and P.M. Sabour, Encapsulation and controlled release of bacteriophages for

food animal production, in Bacteriophages in the control of food-and waterborne

pathogens, P.M. Sabour and M. Griffiths, Editors. 2010, ASM Press. p. 237-255.

179. Maniatis, T., E.F. Fritsch, and J. Sambrook, Molecular cloning: a laboratory manual. Vol.

545. 1982: Cold Spring Harbor Laboratory Cold Spring Harbor, NY.

180. Crhanova, M., et al., Immune response of chicken gut to natural colonization by gut

microflora and to Salmonella enterica serovar enteritidis infection. Infection and

Immunity, 2011. 79(7): p. 2755-2763.

181. Sabour, P.M. and M.W. Griffiths, Bacteriophages in the control of food-and waterborne

pathogens, 2010: ASM Press.

180

182. Ma, Y., et al., Enhanced alginate microspheres as means of oral delivery of bacteriophage

for reducing Staphylococcus aureus intestinal carriage. Food Hydrocolloids, 2012. 26(2):

p. 434-440.

183. Weld, R.J., C. Butts, and J.A. Heinemann, Models of phage growth and their applicability

to phage therapy. J Theor Biol, 2004. 227(1): p. 1-11.

184. Lindberg, A.A. and T. Holme, Influence of O Side Chains on the Attachment of the Felix

O-1 Bacteriophage to Salmonella Bacteria. Journal of Bacteriology, 1969. 99(2): p. 513-

519.

185. Sheng, H., et al., Application of bacteriophages to control intestinal Escherichia coli

O157:H7 levels in ruminants. Appl Environ Microbiol, 2006. 72(8): p. 5359-66.

186. Hume, M.E., et al., Early Salmonella challenge time and reduction in chick cecal

colonization following treatment with a characterized competitive exclusion culture. J

Food Prot, 1998. 61(6): p. 673-6.

New Formulation and Evaluation of Multi-particulate Antibiotic … · 2019. 2. 9. · for Oral...

Documents

Transcript of New Formulation and Evaluation of Multi-particulate Antibiotic … · 2019. 2. 9. · for Oral...