The Pennsylvania State University

The Graduate School

College of Agricultural Sciences

A STUDY ON THE EFFECT OF DELIVERY VEHICLES ON THE EFFICACY

OF BIFIDOBACTERIUM ANIMALIS SUBSP. LACTIS BB-12 IN HUMANS

A Dissertation in

Food Science

by

Zhaoyong Ba

© 2016 Zhaoyong Ba

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

May 2016 The dissertation of Zhaoyong Ba was reviewed and approved* by the following:

Robert F. Roberts Department Head of Food Science Professor of Food Science Dissertation Advisor Chair of Committee

Ryan J. Elias Frederik Sr. and Faith E. Rasmussen Professor of Food Science Associate Professor of Food Science

Rodolphe Barrangou Adjunct Professor of Food Science Associate Professor of Food, Bioprocessing and Nutrition Sciences North Carolina State University

Penny M. Kris-Etherton Distinguished Professor of Nutrition

James L. Rosenberger Director of SCC and Online Programs Professor of Statistics

*Signatures are on file in the Graduate School

Abstract

Bifidobacterium animalis subsp. lactis BB-12 (BB-12) is a probiotic strain that has been used as a food ingredient and food supplement worldwide since 1985 with well- documented health benefits. It has been commonly delivered in fermented dairy products for perceived health benefits including facilitating bowel transit, increasing gut short chain fatty acid (SCFA) production, and modifying gut microbiota (increase the ratio of potentially beneficial to harmful microbes). In addition to traditional fermented dairy- based probiotic-containing products, many new types of probiotic-containing products such as juice-based, chocolate-based, and capsules have been developed. While these provide more options for people to access probiotics, little research has been done on the effect of delivery matrix (dairy vs. non-dairy) on the efficacy of BB-12 in humans. In addition, it was not clear how yogurt fermentation might impact the survival of BB-12 in the product or on its performance in vivo. In order to address these questions, a randomized clinical trial was conducted using BB-12 delivered by capsule (CAP) or yogurt smoothies.

The first step in the project was to identify a selective medium for the enumeration of BB-12 in yogurt product where high concentration of live cultures

Lactobacillus delbrueckii subsp. bulgaricus (LB) and Streptococcus thermophilus (ST) are present. After evaluation of different selective media, the MRS-NNLP agar was found to effectively inhibit growth of ST and LB, and was subsequently chosen for BB-12 enumeration. Then three yogurt smoothies: 1) a control yogurt smoothie without BB-12

(YS), 2) a yogurt smoothie with BB-12 added pre-fermentation (PRE), or 3) post-

iii

fermentation (POST), were developed. Compositions (fat and total solids) of the yogurt smoothies were measured following each production. The concentration of BB-12, ST,

LB, and pH of the products were measured weekly during shelf life. Data from 27 batches indicated the BB-12 population in the yogurt smoothies was stable during shelf life and comparable within and across batches. The population of BB-12 declined in both PRE and POST throughout shelf life and was found to decrease faster in the POST drink as compared to the PRE drink. The BB-12 remained at the specified dose level (log10 10 ±

0.5 CFU per/serving) throughout shelf life of the products.

The second objective was to evaluate the effect of BB-12 interventions on the gut transit time and fecal SCFA production in healthy adults. A total of 36 healthy adults aged 18-40 years with a BMI of 20-35 kg/m2 were recruited in this controlled, 4-period crossover, free-living study. Participants received each of the 4 treatments in a random order. When consuming yogurt smoothies, participants consumed 240 g of yogurt smoothie per day or on the capsule treatment, they consumed a probiotic capsule per day, receiving a dose of log 10 ± 0.5 CFU of BB-12/day. Dietary intake of total calories and various nutrients were assessed via self-reported 3-day dietary recall. Gut transit times were measured at baseline using the blue dye and a telemetry device known as a

SmartPill®, and after each treatment using SmartPill® only. Stool and blood samples were collected at baseline and after each treatment. Fecal SCFAs were extracted using ethyl acetate and analyzed by gas chromatography (GC). A significant positive correlation

(Spearman rho = 0.67, P < 0.0001) was observed between blue dye and SmartPill® transit times, suggesting that the blue dye method remained a reliable cost-efficient approach for

iv

baseline screening for whole gut transit time (WGTT). No significant treatment effect was observed on either GTT or fecal SCFAs. However, this study is the first to demonstrate the possible relationships among regional gut transit times and fecal SCFAs in healthy adults, and the results from this study confirmed a number of correlations that have been reported previously. Notably, the predominant SCFAs negatively correlated with WGTT, colonic transit time (CTT), and gastric emptying time (GET), but had little to do with small bowel transit time (SBTT).

The third objective was to survey the gut microbiota of participants before and after different BB-12 interventions. Bacterial genomic DNA was isolated from fecal samples using the MOBIO PowerSoil DNA isolation kit according to the manufacturer’s protocol with modifications. The DNA samples were amplified for the V4 region of the

16S rRNA gene. Amplicons were then sequenced at the DNA Technologies Core Facility of the University of California, Davis on an Illumina® Genome Analyzer II sequencing platform. After removing samples with low quality and poor compliances, about 2.4 million sequences from 147 stool samples were analyzed using QIIME. Overall, 10 phyla and 109 genera were identified in the participants. Firmicutes, Bacteroidetes,

Actinobacteria, and Proteobacteria accounted for > 98% of the sequences at the phylum level. No significant treatment effect on the gut microbiota was detected due to the large interpersonal and intrapersonal variations observed except that yogurt interventions resulted in a higher relative abundance of Streptococcus than the capsule treatment. A significant gender effect was observed when comparing the gut microbiota of the present study cohort.

v

In an effort to understand the relationships between gut microbiota and host immune and metabolic responses, the beta-diversity results were grouped by parameters including sex, body mass index (BMI), blood pressure, glucose, high-density lipoprotein

(HDL), low-density lipoprotein (LDL), triglycerides (TGs), tumor necrosis factor alpha

(TNF-α), interferon gamma (IFN-γ), and etc., a number of statistically significant differences were found. However, further studies are needed to validate these relationships, because either only a small percentage of difference could be explained by the grouping, or only a few data points were in one of the two arms.

To further explore the effect of the interventions on the relative abundance of members of Bifidobacterium genus, DNA samples were analyzed using bifidobacteria specific terminal restriction fragment length polymorphism (Bif-TRFLP). Interestingly, the two BB-12-containing yogurt smoothies (PRE and POST) resulted in significantly higher percentage of B. animalis when compared to baseline, to the BB-12-free yogurt smoothie (YS), to the BB-12 containing capsule (CAP), and after final washout.

In conclusion, the BB-12 survived well in yogurt smoothies throughout shelf life

(< 1 log decrease) and survived better in the pre-added treatment than in the post-added product perhaps due to BB-12’s adaptation to the acidic environment during fermentation.

Little treatment effect was observed on either GTT or fecal SCFAs. The predominant

SCFAs negatively correlated with regional transit times except for SBTT. Small treatment effect on the gut microbiota was detected while gender effect was found to be significant in the present study cohort. BB-12 containing yogurt smoothies resulted in higher relative abundance of B. animalis than BB-12 containing capsule.

vi

Table of Contents

List of Figures ...... xi

List of Tables ...... xiii

List of Abbreviations ...... xiv

Acknowledgements ...... xix

Chapter 1 - Literature Review ...... 1

1.1. Probiotics ...... 1

1.2. The Genus Bifidobacterium ...... 2

1.3. Bifdobacteriam animalis subsp. lactis ...... 4

1.4. Bifidobacterium animalis ssp. lactis BB-12 ...... 5

1.5. Probiotic Effects of BB-12 ...... 6

1.5.1. BB-12 and GI Health ...... 7

1.5.2. BB-12 and Immune Health ...... 11

1.5.3. BB-12 and Other Health Effects ...... 17

1.5.4. BB-12 in Comparison with Other Bifidobacteria ...... 18

1.6. Influence of Probiotic Delivery Vehicles ...... 19

1.7. Production of BB-12 containing Yogurt ...... 21

1.8. Enumeration of BB-12 ...... 23

1.9. Fecal Transit Time ...... 24

1.10. Gut Microbiota ...... 30

1.10.1. Analysis of Microbiota ...... 31

1.10.1.1. 16S ribosomal DNA ...... 34

1.10.1.2. Operational Taxonomic Units (OTUs) ...... 34

1.10.1.3. Sequence Data Analysis ...... 35

vii

1.10.2. Is There a “Core Microbiota”? ...... 38

1.10.3. Gut Microbiota and Human Health ...... 40

1.10.3.1. Gut Microbiota and Probiotics ...... 47

1.10.3.2. Gut Microbiota and BB-12 ...... 51

1.11. Fecal Short Chain Fatty Acids (SCFAs) ...... 53

1.12. Conclusion ...... 59

1.13. Hypotheses and Specific Aims ...... 60

Chapter 2 - Characterization of Bifidobacterium animalis subsp. lactis BB-12 Containing Yogurt Smoothie Interventions ...... 61

2.1. Introduction ...... 61

2.2. Materials and Methods ...... 63

2.2.1. Bacterial cultures ...... 63

2.2.2. Production of yogurt smoothie interventions ...... 64

2.2.3. Shelf life and safety of the yogurt smoothies ...... 66

2.2.4. Selective enumeration of BAL ...... 66

2.2.5. Verification of BAL ...... 68

2.2.6. Composition of the yogurt smoothies ...... 68

2.2.7. Statistical analysis ...... 69

2.3. Results and Discussion ...... 69

2.3.1. Development of PRE-added yogurt smoothie ...... 69

2.3.2. S. thermophilus grows on modified MRS but not on MRS-NNLP ...... 70

2.3.3. BB-12 remained stable throughout shelf life ...... 72

2.3.4. S. thermophilus remained stable ...... 74

2.3.5. Yogurt smoothies had comparable composition ...... 75

2.4. Conclusions ...... 76

viii

Chapter 3 - The Effect of Bifidobacterium animalis subsp. lactis BB-12 and Different Delivery Vehicles on Gut Transit Time and Fecal Short Chain Fatty Acid Production in Healthy Adults ...... 78

3.1. Introduction ...... 78

3.2. Materials and Methods ...... 81

3.2.1. Participants ...... 81

3.2.2. Recruitment and screening ...... 82

3.2.3. Design and intervention ...... 83

3.2.4. Diet and physical activity assessment ...... 85

3.2.5. Gut transit time measurement ...... 86

3.2.6. Fecal SCFAs measurement ...... 87

3.2.7. Statistical analysis ...... 90

3.3. Results and Discussion ...... 90

3.3.1. Participant characteristics ...... 90

3.3.2. Blue dye results correlate with SmartPill® measurements ...... 93

3.3.3. Assessment of gut transit time ...... 95

3.3.4. Fecal SCFA concentrations ...... 97

3.3.5. Gut transit times correlate with predominant fecal SCFAs ...... 100

3.4. Conclusions ...... 103

Chapter 4 - The Effect of Bifidobacterium animalis ssp. lactis BB-12 and Different Delivery Vehicles on the Gut Microbiota of Healthy Adults ...... 114

4.1. Introduction ...... 114

4.2. Materials and Methods ...... 117

4.2.1. Design and intervention ...... 117

4.2.2. Participants ...... 118

ix

4.2.3. Recruitment and screening ...... 118

4.2.4. Diet and physical activity assessment ...... 120

4.2.5. Stool sample collection ...... 120

4.2.6. Isolation of stool DNA ...... 121

4.2.7. Compliance check using PCR ...... 122

4.2.8. Illumina® sequencing ...... 123

4.2.9. Bif-TRFLP ...... 124

4.2.10. Sequence data analysis ...... 125

4.3. Results and Discussion ...... 127

4.3.1. Participant characteristics ...... 127

4.3.2. Compliance ...... 129

4.3.3. Compositional characteristics of the fecal microbiota of the participants ..... 131

4.3.4. Diversity of the participants’ fecal microbiota ...... 135

4.3.5. Fecal bifidobacterial distribution ...... 141

4.4. Conclusions ...... 144

Chapter 5 – Conclusions and Future Directions ...... 153

References ...... 157

Appendix I: The Procedure of Yogurt and Yogurt Smoothie Preparation ...... 187

Appendix II: Treatment Sequences ...... 191

Appendix III: Stool Collection Instructions for Participant ...... 192

Appendix IV: Gut transit times of all participants ...... 193

x

List of Figures

Figure 1.1. A complex interplay between diet, the gut microbiota, and GI transit. .... 27

Figure 2.S1. An example of PCR verification of BAL...... 77

Figure 3.1. Schematic diagram for randomization design...... 85

Figure 3.3. Diagram showing the extraction of SCFAs...... 89

Figure 3.4. Baseline whole gut transit time as measured by blue dye and SmartPill®...... 94

Figure 3.S1. Recruitment strategy used in the study...... 104

Figure 3.S2. Changes in regional gut transit time of the participants’ after each treatment from baseline...... 104

Figure 4.1. Taxonomic cladogram of LDA effect size comparing relative abundance of taxa between male and female...... 134

Figure 4.2. Boxplot showing the comparison of alpha diversity indices of the fecal microbiota of participants before and after each treatment...... 136

Figure 4.3. UPGMA tree based on weighted UniFrac distance (beta diversity) demonstrating the hierarchical relationships between the fecal samples ...... 138

Figure 4.4. Weighted UniFrac distance PCoA of bacterial communities with jackknife support grouped by treatment and gender...... 139

Figure 4.5. Relative proportion of Bifidobacterium species in the stool DNA samples at baseline, after each treatment, and after final washout as determined by Bif-TRFLP (AluI) ...... 143

Figure 4.S1. Schematic diagram for randomization design...... 145

Figure 4.S2. Recruitment strategy used in the study...... 145

Figure 4.S3. Alpha rarefaction curves of the sequences...... 145

Figure 4.S4. Histogram of the LDA scores computed for features differentially abundant between intervention or final washout and baseline...... 145

Figure 4.S5. Histogram of the LDA scores computed for features differentially abundant between intervention or final washout and baseline in FEMALE participants...... 145

xi

Figure 4.S6. Histogram of the LDA scores computed for features differentially abundant between intervention or final washout and baseline in MALE participants...... 145

Figure I-1. Processing flow diagram ...... 189

xii

List of Tables

Table 2.1. Target yogurt composition ...... 65

Table 2.2. Composition of strawberry yogurt smoothies ...... 65

Table 2.3. BB-12 population in different samples as enumerated by MRS-NNLP agar ...... 70

Table 2.4. BB-12 concentration in capsules as enumerated by MRS and MRS- NNLP ...... 71

Table 2.5. Selectivity test of mMRS and MRS-NNLP ...... 72

Table 2.6. BB-12 population in capsules during shelf life ...... 73

Table 2.7. BB-12 population in yogurt smoothies during shelf life ...... 74

Table 2.8. pH value of the yogurt smoothies throughout shelf life ...... 74

Table 2.9. S. thermophilus population in yogurt smoothies during shelf life ...... 75

Table 2.10. Composition of the yogurt smoothies ...... 76

Table 3.1. Calibration curves of the SCFAs standards ...... 89

Table 3.2. Demographic characteristics of participants at baseline ...... 92

Table 3.4. SCFAs concentration (µg/g) among treatments ...... 99

Table 3.5. Spearman correlations of gut transit time and fecal SCFAs ...... 102

Table 3.S1. Spearman correlation between gut transit time, short chain fatty acid, dietary and demographic data ...... 110

Table 4.1. Demographic characteristics of participants at baseline ...... 128

Table 4.2. B. animalis subsp. lactis PCR results of the participants’ stool samples ... 130

Table 4.3. Predominant fecal bacterial phyla and genera present in healthy adults before and after consuming BB-12 containing interventions in a crossover study ...... 132

Table 4.S1. Statistical analyses of the weighted UniFrac distance (beta-diversity) when grouped by the metabolic and immune parameters ...... 152

xiii

List of Abbreviations

3D Three-dimensional

AAD Antibiotic-associated diarrhea

ANOVA Analysis of variance

ATP Adenosine triphosphate

BAL Bifidobacterium animalis subsp. lactis

BB-12 Bifidobacterium animalis subsp. lactis BB-12

Bif-TRFLP Bifidobacteria specific terminal restriction fragment length polymorphism

BMI Body mass index bp Base pair

CAP Capsule

CD Crohn's disease

CDI Clostridium difficile infection

CFR Code of federal regulations

CFU Colony forming unit

CTT Colonic transit time

DAI Disease activity index

DCs Dendritic cells

DNA Deoxyribonucleic acid dNTPs Deoxynucleotide triphosphates

DSS Dextran sulphate sodium

DZ Dizygotic

ELISA Enzyme linked immunosorbent assays

xiv

EMP Earth Microbiome Project

EPX Eosinophilic protein X

FAO Food and Agriculture Organization of the United Nations

FID Flame ionization detector

FISH Fluorescence in situ hybridization

FMT Fecal microbiota transplantation

FOS Fructooligosaccharides g Acceleration of gravity (~9.81 m/s2) g Gram

G+C Guanine+cytosine

GC Gas chromatography

GET Gastric emptying time

GF Germ-free

GI Gastrointestinal

GIT Gastrointestinal tract

GMP Good manufacturing practice

GRAS Generally recognized as safe

GTT Gut transit time h Hour

HDL High-density lipoprotein

HMP Human Microbiome Project

HPLC High performance liquid chromatography

HRQL Health-related quality of life hs-CRP High-sensitivity C-reactive protein

xv

HTST High temperature/short time

IBD Inflammatory bowel disease

IBS Irritable bowel syndrome

IFN-γ Interferon-gamma

Ig Immunoglobulin

IL Interleukin

IPAQ International physical activity questionnaire

IS Internal standard

KW Kruskal-Wallis

LAB Lactic acid bacteria

LB Lactobacillus delbrueckii subsp. bulgaricus

LDA Linear discriminant analysis

LDL Low-density lipoprotein

LEfSe Linear discriminant analysis effect size

LGG Lactobacillus rhamnosus GG

LPS Lipopolysaccharide

M Molarity

MANOVA Multivariate analysis of variance

MET Metabolic equivalent of task

MetaHIT Metagenomics of the Human Intestinal Tract min Minute ml Milliliter

MRS de Man, Rogosa and Sharpe

MZ Monozygotic

xvi

n-3 PUFA n-3 polyunsaturated fatty acids

NGS Next-generation sequencing

NK Natural killer

NNLP Nalidixic acid, neomycin sulfate, lithium chloride, and paromomycin sulfate

OTU Operational taxonomic unit

PBMCs Peripheral blood mononuclear cells

PCoA Principal coordinates analysis

PCR Polymerase chain reaction

PFGE Pulsed-field gel electrophoresis

POST Post-fermentation

PRE Pre-fermentation

QIIME Quantitative insights into microbial ecology qPCR Quantitative polymerase chain reaction

® Registered

RCT Randomized clinical trial rDNA Ribosomal DNA

RNA Ribonucleic acid rpm Revolutions per minute rRNA Ribosomal ribonucleic acid

SBTT Small bowel transit time

SCFA Short chain fatty acid

SCORAD Scoring atopic dermatitis

SDS Sodium dodecyl sulfate

xvii

SNPs Single nucleotide polymorphisms

SOP Standard operating procedure

ST Streptococcus thermophilus

T1D Type-1 diabetes

TGs Triglycerides

TGF-β Transforming growth factor beta

TGGE Temperature gradient gel electrophoresis

TLR Toll-like receptor

TM Trademark

TNF-α Tumor necrosis factor alpha

T-RFLP Terminal restriction fragment length polymorphism

UC Ulcerative colitis

URTI Upper respiratory tract infection

WGTT Whole gut transit time

WHO World Health Organization

YS Yogurt smoothie

xviii

Acknowledgements

I would first like to thank Dr. Robert Roberts, my dissertation advisor, for his support, encouragement, and guidance in the past several years. I consider it was a great privilege to work in the Roberts’ lab, because I was able to learn both scientific and practical skills under Dr. Roberts’ supervision. I want to sincerely thank Dr. Roberts for sharing his knowledge and expertise in Food Microbiology and Dairy Processing with me, teaching me to become a good scientist and even better person. Dr. Roberts exhibited characteristics I hope to one day emulate.

I would also like to thank the members of my committee, Dr. Rodolphe

Barrangou, Dr. Ryan Elias, Dr. Penny Kris-Etherton, and Dr. James Rosenberger for their time and help. They were always available, interested, and provided positive encouragement and advice throughout this research.

I must thank Emily Furumoto, who taught me basic lab operations when I first came to the lab, ensured plenty of lab/office supplies, and kept us well fed. Additionally,

I would like to thank my lab mates, who made the Roberts’ lab a fun, collaborating, and drama-free working environment.

I have to thank Dr. Penny Kris-Etherton, Dr. Connie Rogers, Jennifer Fleming,

Dr. Huicui Meng, and Yujin Lee from the Department of Nutritional Sciences for their excellent collaboration and tremendous support for my dissertation project. This dissertation would not have been possible without the immeasurable amount of work they put in. Meanwhile, I must express my appreciation for Drs. David Mills and Zachary

Lewis’ assistance with the 16S rDNA sequencing and Bif-TRFLP at UC Davis.

xix

I also have to thank the stuff members of the Penn State Berkey Creamery, especially Bonnie Ford, Aaron Kenepp, and Art Conklin. It was my great pleasure working with them in the processing plant to make the study interventions.

Furthermore, I would like to express gratitude to the entire Food Science

Department faculty, staff, and students, who have been extremely helpful throughout my time here; to the nurses at the Penn State Clinical Research Center, who helped with the clinical aspect of this dissertation work; and to the volunteers, who have participated in this 7-month long study.

Finally, I would like to thank my lovely wife Di, my son Tony, and my family members back in China for their love and support.

xx

Chapter 1 - Literature Review

1.1. Probiotics

Probiotics are live microorganisms that, when administered in adequate amounts, confer a health benefit on the host (FAO/WHO 2002, Hill et al. 2014). The term

“probiotic” means “for life”, and was coined by Lilly and Stillwell to contrast with

“antibiotics”, and referred to growth-promoting factors produced by microbes that enhance the growth of other microorganisms (Lilly and Stillwell 1965). Development of an understanding of the positive role played by microbes can be dated back over 100 years to Elie Metchnikoff, a Russian Nobel Prize winner, who suggested that daily doses of lactic acid bacteria (LAB) in “soured milk” in the Bulgarian diet was one of the reasons for the unusually large number of centenarians (Mackowiak 2013). Probiotic bacteria can either produce metabolites themselves, such as organic acid and bacteriocins

(Martinez et al. 2013), or trigger host immune responses to benefit the host (Damaskos and Kolios 2008). The functionality of probiotics is believed to be genus, species, and even strain-specific. Although many genera of microorganisms are considered to be probiotic, currently the majority are members of the genera Bifidobacterium and

Lactobacillus. Commonly used strains include Bifidobacterium animalis subsp. lactis

BB-12, B. animalis subsp. lactis DN-173 010, B. longum BB536, B. breve Yakult, B. infantis 35624, Lactobacillus casei DN-114001, Lb. paracasei CRL-431, Lb. casei F19,

Lb. acidophilus LA-5, Lb. acidophilus NCFM, and some others (Guarner et al. 2008).

Members of the genera Streptococcus, Enterococcus, Bacillus, Propionibacterium, even

1 gram-negative bacteria (e.g. Escherichia coli Nissle 1917) and yeast (e.g.

Saccharomyces) can be found in the probiotic market (Gerritsen et al. 2011).

Studies have been conducted addressing the benefits of consuming probiotics or probiotic-containing products (Fuller 1989). Putative benefits studied included alleviation of diarrhea, modulation of host immune system, reduction of allergic responses, decreased bowel transit time, prevention of infections, and numerous other effects (de

Vrese and Schrezenmeir 2008, Sherman et al. 2009). Due to these health benefits, the worldwide market potential for probiotics is tremendous. The global probiotics market,

70% of which are cultured dairy products, is projected to reach 46.55 billion US Dollars by 2020 (marketsandmarkets.com 2015). In addition to traditional fermented dairy-based probiotic-containing products, many new types of probiotic-containing products such as juice-based, chocolate-based, and capsules have been developed. While these provide more options for people to access probiotics, the efficacy of these products as vehicles to deliver probiotics has not been extensively studied.

1.2. The Genus Bifidobacterium

In an excellent review article (Lee and O'Sullivan 2010), bifidobacteria are described as “non-motile, non-spore-forming, non-gas-producing, Gram-positive, anaerobic, catalase-negative bacteria with a high G+C content (55%-67%) with a morphology generally referred to as bifid or irregular V- or Y- shaped rods resembling branches”. Bifidobacteria were first identified in 1899 by Tissier in the feces of a healthy infant and named Bacillus bifidus (Poupard et al. 1973). Tissier speculated on the likely

2 association between high bifidobacteira concentration in the feces of breast-fed infants and their lower incidence of infantile gastroenteritis. This association was substantiated by a number of later studies (Bullen and Willis 1971, Collins and Gibson 1999, Gill and

Prasad 2008). Bacillus bifidus was later re-named Lactobacillus bifidus because of its similarity to Lactobacillus acidophilus (Poupard et al. 1973). After 13 recognized name changes, Bacillus bifidus is now known as Bifidobacterium bifidum (Scardovi 1986). The genus Bifidobacterium currently contains 38 species and subspecies; the most recent addition is B. stercoris (Kim et al. 2010).

Russell et al. (Russell et al. 2011) reported species of bifidobacteria inhabit seven different ecological niches: the human intestine, the human vagina, the oral cavity, food, the animal gastrointestinal tract (GIT), sewage, and the intestine of honeybees. The role of bifidobacteria in human intestinal microflora is not yet clear, but studies have suggested a likely beneficial role for the host (Lee and O'Sullivan 2010). A number of studies have reported some bifidobacteria are able to consume complex oligosaccharides that normally escape digestion in the upper GI-tract and produce short chain fatty acids

(SCFAs), which are water soluble and readily absorbed by the host (Barrangou et al.

2003, Russell et al. 2011). Other roles of bifidobacteria in the large intestine are the production of water-soluble vitamins, including many of the B group vitamins (Deguchi et al. 1985), control of undesirable bacteria (Chen et al. 1999, Alvaro et al. 2007), and stimulation of the host innate immune response (Arunachalam et al. 2000, Sonnenburg et al. 2006).

3

1.3. Bifdobacteriam animalis subsp. lactis

Bifdobacteriam animalis subsp. lactis belongs to a species that only has two subspecies, B. animalis subsp. animalis and B. animalis subsp. lactis (Cai et al. 2000). B. animlis subsp. lactis was first identified and isolated as strain UR1 by Meile et al. (Meile et al. 1997) from a French yogurt. This strain was found to grow at elevated oxygen concentration above 5% oxygen in liquid media containing 0.3mM methylviologen.

Meile et al. considered this strain as a new Bifidobacterium speices based on the fact that it had only 27% DNA homology with its most related species B. animalis DSM 20104 and named it B. lactis.

A few years later, Cai et al. (Cai et al. 2000) reclassified B. lactis as a junior subjective synonym of B. animalis because of the close relationship observed between B. lactis and B. animalis in the bifidobacterial phylogenetic tree and the high similarity

(98.8%) in 16S rRNA sequence between these two species. Within a year of reclassification, Ventura et al. (Ventura et al. 2001) developed a set of primers,

Bflact2/Bflact5, that specifically target the 16S rDNA region of the subspecies B. animalis subsp. lactis.

Masco et al. (Masco et al. 2004) further evaluated 16 representatives of B. animalis subsp. animalis and B. animalis subsp. lactis and demonstrated the genetic and phenotypic relatedness between these two subspecies. They described the optimum growth condition for B. animalis subsp. lactis as: “ The optimum growth temperature is

39 – 42°C. No growth occurs on agar plates exposed to air, but 10% oxygen in the headspace atmosphere above liquid media is tolerated. Growth occurs in milk or milk-

4 based media. The molar ratio of acetate to lactate from glucose metabolism is about 10 to

1 under anaerobic conditions.” Commonly used commercial strains include B. animalis subsp. lactis BB-12, B. animalis subsp. lactis DN-173 010, B. animalis subsp. lactis Bl-

04, B. animalis subsp. lactis Bi-07, and B. animalis subsp. lactis HN019 (Barrangou et al.

2009, Garrigues et al. 2010).

1.4. Bifidobacterium animalis ssp. lactis BB-12

Bifidobacterium animalis ssp. lactis strain BB-12 is a probiotic strain from Chr.

Hansen’s culture collection. It was considered to belong to the species Bifidobacterium bifidum when it was first isolated. Modern molecular biology based classification methods reclassified BB-12 as Bifidobacterium animalis subspecies lactis (BAL)

(Jungersen et al. 2014). This strain has been used as a food ingredient and food supplement worldwide since 1985. BB-12 has a genome size of 1,942,198 bp, which is one of smallest genomes in the genus Bifidobacterium (Lee and O'Sullivan 2010). Strains of this subspecies are believed from B. animalis species in a dairy fermentation environment, which may be a result of bacterial genome reduction (Nilsson et al. 2005).

Strains of Bifidobacterium animalis ssp. lactis, including BB-12 are oxygen and bile- tolerant, providing technological advantage over other probiotics (Garrigues et al. 2005).

This strain is also GRAS (Generally Recognized As Safe) approved by FDA (GRAS

Notice No. GRN 000049) and B. animalis has been granted Qualified Presumption of

Safety (QPS) status since 2007 by the European Food Safety Authority (EFSA).

Moreover, this organism has been extensively studied in a variety of populations, is

5 readily available, has been assessed for safety in populations ranging from infants to the elderly and has been completely sequenced (Chouraqui et al. 2004, Ouwehand et al.

2004, Garrigues et al. 2010). In addition, it has been shown to decrease fecal transit time in healthy young adults (18 – 40 years) in a dose-dependent manner (Larsen et al. 2006,

Ringel-Kulka et al. 2008) and modulate immune function in infants (Isolauri et al. 2000).

In a recent large scale (1,248 healthy subjects with low defecation frequency (2-4 d/week), mean age ~37 years), randomized, double-blind, placebo-controlled, parallel- group clinical trial (Eskesen et al. 2015), a treatment effect on average defecation frequency was found, as the frequency in BB-12 group being significantly higher than placebo group (P = 0.0065). However, the placebo also significantly increased the frequency as compared to baseline.

Based on previous research with strains of BAL, a dose level of log 10 ± 0.5 colony-forming unit (CFU) per day is considered effective for BB-12 (Bouvier et al.

2001, Marteau et al. 2002, Guyonnet et al. 2009, Guyonnet et al. 2009). A study by

Larsen et al. (Larsen et al. 2006) found that even at doses as low as 1 x 108 CFU per day in a capsule for 3 weeks resulted in a decreased transit time.

1.5. Probiotic Effects of BB-12

BB-12 has been used in over 300 human, animal, and in vitro studies, which were mainly focused on gastrointestinal (GI) and immune health. The earliest documented research examining the probiotic effect of BB-12 was conducted by Black (Black 1996).

He performed two placebo-controlled, double-blind trials on a total of 195 travelers who

6 had a 15-day tour in Egypt using a probiotic capsule containing BAL BB-12, Lb. acidophilus LA-5, Lb. bulgaricus LBY-27 and Streptococcus thermophilus STY-31. He found that only 46.8% of the probiotic group, compared to 73.3% of the placebo group, suffered from traveller’s diarrhea (P < 0.05). Since a mixture of different bacteria were used in this study, it is rather difficult to attribute the effect to BB-12 alone. Schiffrin et al. (Schiffrin et al. 1995) conducted the first study using BB-12 alone as an intervention to study the immunomodulatory effect of lactic acid bacteria in humans. A total of 28 (12 females and 16 males) were randomly assigned to two groups that received two different types of fermented milk (supplement with BB-12 or Lb. acidophilus La1) in addition to their normal diet for 3 weeks. Blood samples were measured for lymphocyte subsets or leukocyte phagocytic activity. No modification of lymphocyte subpopulation was detected. However, both groups showed enhanced phagocytosis of Escherichia coli sp. in vitro, suggesting non-specific, anti-infective mechanisms of defense can be enhanced by the ingestion of specific LAB strains.

A number of studies have evaluated the probiotic effect of BB-12 on GI and immune disease conditions. These studies are summarized below.

1.5.1. BB-12 and GI Health

Bifidobacterium animalis subsp. lactis BB-12 has shown promising effects on GI health in populations ranging from infants to the elderly in randomized clinical trials

(RCTs). Saavedra et al. (Saavedra et al. 1994) performed a double-blind, placebo- controlled trial on infants aged 5-24 months who were admitted to the Mount Washington

Pediatric Hospital, a chronic medical care facility. They were randomly assigned to two

7 groups: 1) 26 infants received standard infant formula, 2) 29 infants received the same formula supplemented with probiotic B. bifidum (currently known as BB-12) and S. thermophilus. They were monitored daily for the occurrence of diarrhea, and fecal samples, collected weekly, were analyzed for rotavirus during a total of 4,447 patient- days (within 17 months). While 8 (31%) of the 26 patients who received standard formula developed diarrhea, only 2 (7%) of the 29 infants in the probiotic group showed diarrhea symptom during the course of the study (P = 0.035). Only 3 (10%) infants in the probiotic group shed rotavirus at some time during the study, compared to 10 (39%) of the patients in the control group (P = 0.025). The authors concluded supplementation of infant formula with BB-12 and S. thermophilus could reduce the incidence of acute diarrhea and rotavirus shedding in infants admitted to hospital. The limitation of this study is that BB-12 was not the only strain used, although there is no evidence showing any probiotic effect of S. thermophilus.

Chouraqui et al. (Chouraqui et al. 2004) performed a multicenter, double-blind, controlled study to evaluate the efficacy of an acidified milk formula supplemented with viable BB-12 (BbF) in the prevention of acute diarrhea in infants younger than 8 months living in residential nurseries or foster care centers. Ninety healthy infants were given either the BbF or a conventional formula (CF) daily for a mean of 137 days and 148 days, respectively. Participants were evaluated daily for stool number and consistency.

Although there was no difference in number of infants with diarrhea (13/46 in BbF group vs. 17/44 in CF group), relative risk of diarrhea was significantly lower in the probiotic group than the control group (0.54 in BbF vs. 1 in CF, P < 0.001). The study suggested

8 viable BB-12 added to an acidified infant formula has some protective effect against acute diarrhea in healthy infants. Again, the treatment formula was acidified by S. thermophilus and Lb. helveticus, which may confound the findings.

Weizman et al. (Weizman et al. 2005) reported the preventive effect of BB-12 against diarrhea in term infants aged 4-10 months in a double-blind, placebo-controlled, randomized trial, where 201 participants were randomly assigned to control formula (n =

60), BB-12 supplemented formula (n = 73), and Lb. reuteri supplemented formula (n =

68) for 12 weeks including follow-up. Infants in both probiotic groups had fewer and shorter episodes of diarrhea than control, but no effect on respiratory infections was observed. Additionally, Lb. reuteri showed more prominent effects than BB-12.

In a double-blind, placebo-controlled, randomized study, Mohan et al. (Mohan et al. 2006) studied the effect of BB-12 on the gut microflora of preterm infants. Sixty-nine preterm infants with a gestational age of < 37 weeks were randomly assigned to the placebo (conventional infant formula, 32 infants) or probiotic (conventional infant formula supplemented with 2 × 109 CFU of BAL BB-12 per gram of powder, 37 infants) group. The interventions were given on the first day after birth and continued for 21 days.

Fecal samples were collected as fresh as possible during the study period for bacterial analyses. Both culture-dependent (selective media) and culture-independent

(Fluorescence in situ hybridization, FISH) methods found significantly higher number of

Bifidobacteria in the probiotic group than in the placebo group. However, the probiotic group had lower viable counts of Enterobacteriaceae and Clostridium ssp. than the infants in the placebo group. The study suggested supplementation of preterm infants with BB-

9

12 might have beneficial effect on their gut microbiota composition as the number of bifidobacteria was increased while the numbers of enterobacteria and clostridia, which include many potential pathogens, were reduced after BB-12 administration.

A recent randomized, double-blind, placebo-controlled study conducted by

Hojsak et al. (Hojsak et al. 2015), investigated the effect of BB-12 on the prevention of common (GI and respiratory) infections in healthy children who attended day care centers.

210 children with a median age of 4.6 years were randomized into placebo group (106) or probiotic group (104). All children were given 1 gram of powder with (109 CFU) or without BB-12 for 3 months. Participants were monitored primarily for the occurrence of common infections and the duration of symptoms during the study period. There were only 11 GIT infections in placebo group and 14 in probiotic group with an incident rate of 0.035 (95% CI 0.0173-0.0619) and 0.045 (95% CI 0.0245-0.0753), respectively (P =

0.517). No difference in duration of GIT infections was observed. Eighty-eight respiratory tract infections in placebo group and 83 in probiotic group were recorded in the same study. There was no difference in incidence rate or duration, suggesting BB-12 has no effect on the prevention of respiratory tract infection in children attending day care centers. The authors stated that the results of this study could not be easily extrapolated to other age groups, e.g. infants or adults, although they were aware that an earlier study

(Taipale et al. 2011) reported preventive effect of BB-12 on respiratory infections in infants. A more recent study performed by Ringel-Kulka et al. (Ringel-Kulka et al. 2015) further demonstrated daily supplementation of children’s diet with yogurt containing BB-

12 and inulin significantly reduced days of fever, improved social and school functioning,

10 and increased frequency of bowel movements. It is noteworthy that prebiotic inulin used in this study may confound the probiotic effect of BB-12.

Numerous studies have also reported beneficial effects of BB-12 on the GI health of healthy adults, including improvement of defecation and increase beneficial microbes of the fecal microflora (Nishida et al. 2004, Nishida et al. 2004, Larsen et al. 2006,

Pitkala et al. 2007), shortened duration of antibiotic-associated diarrhea (AAD)

(Chatterjee et al. 2013), improvement of intestinal environment of the elderly

(Matsumoto et al. 2001).

While the exact mechanisms of action of BB-12 on gut health remain unclear, there are a few putative explanations. It has been proposed that BB-12 supports gut health through inhibition of pathogens, barrier function enhancement and immune interaction

(Jungersen et al. 2014). More specific mechanisms include production of inhibitory substances (organic acids, H2O2, bacteriocins), nutrient competition, toxin removal or degradation, competition for sites of adherence (mucus, cell receptors), co-aggregation and virulence modulation, production of short chain fatty acids, and induction of host immune responses. However, further studies are needed to substantiate these mechanisms.

1.5.2. BB-12 and Immune Health

As more clinical studies reported immunomodulatory effect of probiotics, researches have began to acknowledge immune interaction as an important mode of action for probiotic bacteria. The earliest study on the effect of BB-12 on human immune status was by Link-Amster et al. (Link-Amster et al. 1994) in 1994. The intervention of

11 this study was a mixture of BB-12 and Lb. acidophilus La1 in fermented milk. Although it is difficult to attribute the observed effect to BB-12 alone due to confounding variables, this study reported a significantly higher increase in specific serum immunoglobulin A

(IgA) titre in response to attenuated Salmonella typhi Ty21a in the treatment group than in the control group suggesting that LAB may act as adjuvant to the humoral immune response.

Since the late 1990s, numerous studies have been undertaken to investigate the immuno-modulatory effect of BB-12 in populations ranging from infant to elderly.

Isolauri’s group studied the effect of BB-12 on allergic inflammation control in infants

(Isolauri et al. 2000). In this randomized, double-blind study, 27 infants average age 4.6 months were recruited while still fully breastfed. The patients were randomly weaned to a formula supplemented with BB-12; a formula supplemented with Lb. rhamnosus GG, and a formula without probiotic, respectively. The patients were measured for the extent and severity of atopic eczema, their growth and nutrition, and the concentration of circulating cytokines/chemokine and soluble cell surface adhesion molecules in serum and methyl- histamine and eosinophilic protein X (EPX) in urine. Based on SCORAD score, established by the European Task Force on Atopic Dermatitis to assess the severity of atopic eczema, both BB-12 and LGG significantly improved the skin condition after 2 months as compared to the control group (χ 2 = 12.27, P = 0.002) along with a reduction in the concentration of soluble CD4 in serum and EPX in urine. This study indicated certain probiotic strains (BB-12, LGG) could counteract the inflammatory responses in infants during the weaning period.

12

Fukushima et al. (Fukushima et al. 1998) evaluated the effect of BB-12 containing formula on intestinal IgA production in healthy Japanese children aged 15 to

31 months. A significant increase in fecal levels of IgA and anti-poliovirus IgA was observed during formula intake compared to the values before intake. Since there were only 7 children included in the study and due to a lack of appropriate controls, it is difficult to draw any conclusion regarding the probiotic effect of BB-12 in this case.

In a dose-response study in healthy young adults, no immunomodulating effect was observed, even at high doses (1011 CFU/day) of BB-12 and Lb. paracasei CRL-431

(Christensen et al. 2006). In this double blind placebo-controlled study, 71 out of 75 recruited healthy volunteers have completed the study. After a 2-week run-in period, subjects were randomly assigned to one of the following treatments: two capsules containing a total of 108, 109, 1010, 1011 CFU per day of BB-12 and CRL-431 in approximately equal amount or two placebo capsules for 21 days. The end points measured were phagocytic activity in blood leukocytes, fecal IgA concentration, and production of interferon-γ (IFN-γ) and interleukin-10 (IL-10) in blood cells. No statistically significant dose-dependent effect was observed, suggesting no probiotic effect on the immune function of young healthy adults supplemented with even high doses of BB-12 and CRL-431. The authors acknowledged the limitations of using healthy subjects in similar studies, because of the difficulty of identifying relevant immune parameters. In another RCT conducted by Hol et al. (Hol et al. 2008), 119 infants with

Cow’s milk allergy (CMA) were given either BB-12 and CRL-431 supplemented extensively hydrolyzed formula or placebo formula for 12 months. No treatment effect

13 was observed in terms of accelerating cow milk tolerance, although minor effects of probiotic intervention on lymphocytic subpopulations in peripheral blood were observed.

In an effort of evaluate the immune effect of BB-12 potentially in healthy subjects, Rizzardini et al. (Rizzardini et al. 2012) developed an influenza vaccination model. In this randomized, double-blind, placebo-controlled, parallel-group study, 211 healthy subjects (aged 33.2 ± 13.1 years) were assigned to one of the four treatments: a capsule containing 109 CFU per day of BB-12, a placebo capsule, an acidified dairy drink containing 109 CFU per day of Lb. paracasei CRL-431, or a placebo dairy drink for two weeks, respectively. Followed by seasonal influenza vaccination, each group continued with correspond intervention for another 4 weeks. Immune markers in blood and saliva were measured at baseline (2 weeks prior to vaccination) and after a 6-week intervention period (4 weeks after vaccination). Changes from baseline in vaccine-specific IgG, subclasses IgG1 and IgG3, and salivary IgA were significantly greater in each probiotic group compared with the corresponding placebo group. Additionally, significantly greater changes from baseline and mean fold increase (MFI) in total plasma IgG, IgG1,

IgG3, and total salivary IgG were observed in both probiotic groups compared with the relevant placebo group. Taken together, this study indicates that consumption of either the probiotic strains BB-12 or CRL-431 significantly increases antigen-specific immune responses in healthy individual receiving an influenza vaccination.

Smith et al. (Smith et al. 2013) reported on the effect of BB-12 containing interventions on the prevention of upper respiratory infections (URI) in healthy college students. In this randomized, placebo-controlled study, 231 healthy college students

14 living on campus were given placebo (n = 117) or BB-12 and LGG containing powder (n

= 114) for 12 weeks. 198 subjects (97 in placebo and 101 in probiotics) were analyzed for health-related quality of life (HRQL) based on The Wisconsin Upper Respiratory

Symptom Survey-21 (Barrett et al. 2009). Significantly shorter median duration of URI and lower median severity score were observed in the probiotics group in comparison with the placebo group (P < 0.001). Students in the probiotics group also missed significantly fewer school days (0.2 d) than the placebo group (P = 0.002), suggesting use of BB-12 and LGG may be beneficial among college students with URI. Although no immune marker was directly measured in this study, URI symptoms are believed to be result of the inflammatory response of the host towards the virus (Kirchberger et al.

2007). While the exact mechanism remains unclear, the findings of this study may be partially explained by modulation of the inflammatory response.

In another double-blind, placebo-controlled study, Ouwehand et al. (Ouwehand et al. 2008) studied the effect of B. longum and BB-12 on the immune status of institutionalized elderly subjects. A total of 209 elderly (84.3 ± 0.98 years) were randomly assigned to one of the following three interventions: oat-based drink supplemented with 109 CFU per day of B. longum 2C and 46 (n = 56), placebo product without probiotics (n = 67), and a commercial oat drink containing 109 CFU per day of

BB-12 (n = 86). Bifidobacterium species in the feces were measured by qualitative polymerase chain reaction (PCR) and quantitative PCR. The levels of IL-10, TNF-α, and

TGF-β1 in the serum were measured using enzyme linked immunosorbent assays

(ELISA). The results showed that consumption of BB-12 did not change the levels of

15 other bifidobacteria in elderly while B. longum administration resulted in increased levels of B. adolescentis and B. catenulatum. Serum cytokine levels were not influenced by any treatment, but significant correlations were observed between certain Bifidobacterium species and specific cytokine levels (negative correlations between B. longum and IL-10,

B. animalis and IL-10, and positive correlation between B. breve and transforming growth factor (TGF)–β1). This study suggested that consumption of probiotic intervention might modify the intestinal Bifidobacterium microbiota of the elderly, which may further influence their inflammatory responses.

It is true that some studies showed beneficial effects of BB-12 on host immune markers while others saw none. While this is confusing to both consumers and researchers, there are a few reasons that led to such discrepancy in the findings. First, the variation in population of interest may result in the variable findings in RCTs evaluating the immunomodulatory effect of BB-12 due to host-dependent factors. Moreover, different end points are measured from study to study. Also, the fact that combinations of

BB-12 with other probiotic strains were often employed and that dose levels varied, and the use of prebiotics in different studies confounds results. Finally, variations in sample size and diverse intervention periods make it problematic to observe general changes in immune markers as result of BB-12 intervention, especially when the mechanism of actions remains uncertain.

In an in vitro study (López et al. 2010) demonstrated that BB-12 was able to induce maturation of dendritic cells (DCs), antigen-presenting cells of the mammalian immune system, to a similar or higher degree than lipopolysaccharide (LPS) as measured

16 by surface expression markers while cell-free supernatant had a weak or no effect on maturation of DCs. Moreover, BB-12 induced high levels of IL-10, IFN-γ and TNF-α in peripheral blood mononuclear cells (PBMCs). In a similar study (Latvala et al. 2008) compared the ability of 9 different probiotic strains of the genus of Bifidobacterium and

Lactobacillus at various concentrations to induce maturation and cytokine/chemokine expression in human DCs. BB-12 was able to induce all cytokines tested (TNF-α, IL-6,

IL-10, IL-12, and IFN-γ) and chemokine CCL20 in a positive dose-dependent manner. A recent study (Meng et al. 2015) suggested that BB-12 interacts with peripheral myeloid cells via toll-like receptor-2 (TLR-2), which further substantiated the probiotic effect of

BB-12 on human immune function. Further studies are warranted to validate how these cytokine/chemokine induction properties of BB-12 clinically influence individuals in different populations.

1.5.3. BB-12 and Other Health Effects

Apart from GI and immune heath BB-12 has been associated with a number of other health-related topics including metabolic status of infants (Aaltonen et al. 2011), oral health of young children (Singh et al. 2011), prenatal and postnatal growth (Luoto et al. 2010), glucose regulation during and after pregnancy (Laitinen et al. 2009), chronic fatigue syndrome in adults (Sullivan et al. 2009), and blood pressure of infants (Aaltonen et al. 2008). Since 78% of the immune cells are associated with the gut mucosa

(Jungersen et al. 2014), manipulation of gut environment (microbiota) through probiotics like BB-12 to promote the health of host makes sense. However, further studies are

17 needed to better understand the exact mechanism, dosage, duration, and application of each health condition.

1.5.4. BB-12 in Comparison with Other Bifidobacteria

BB-12 has shown beneficial effects in numerous studies. Fewer studies have been conducted on other bifidobacteria strains relative to BB-12 and the results are also inconclusive. BAL DN-173 010, another widely used commercial strain, is closely related to, and shares the same pulse-field gel electrophoresis (PFGE) pattern with BB-12 (Grand et al. 2003). It has been found that DN-173 010 is able to increase stool frequency in constipated children (Tabbers et al. 2009), improve bowel function in women (Marteau et al. 2002, Guyonnet et al. 2009), in the general adult population (Guyonnet et al. 2009), and in patients with irritable bowel syndrome (IBS) (Guyonnet et al. 2007, Agrawal et al.

2009). However, Tabbers et al. (Tabbers et al. 2011) found the increased stool frequency in constipated children induced by DN-173 010 was comparable to that of the control product.

B. longum BB536 has been extensively studied in Japan in populations ranging from infants to elderly (Namba et al. 2010, Akatsu et al. 2013, Kondo et al. 2013,

Enomoto et al. 2014). It was shown that BB536 reduced allergic disease with prenatal and postnatal administration (Enomoto et al. 2014) and may improve the immunity of elderly, although the results were inconclusive and derived from a small number of studies. Gianotti et al. (Gianotti et al. 2010) found BB536 failed to adhere to the colonic mucosa of patients with colorectal cancer. Only limited studies on B. breve were available. One-year treatment of B. breve containing Yakult together with galacto-

18 oligosaccharide was found to be effective in improving the clinical condition of patients with ulcerative colitis (UC) (Ishikawa et al. 2011). B. infantis 35624 was also found to be useful to patients with UC as this microbe significantly reduced their systemic pro- inflammatory biomarkers (IL-6, TNF-α) (Groeger et al. 2013). Additionally, B. infantis

35624 was shown to alleviate some symptoms of IBS (O'Mahony et al. 2005, Whorwell et al. 2006). However, further studies with larger sample size are warranted to validate these findings and to understand the mechanisms behind them. It seems obvious that different probiotic species/strain possesses unique and specific functional properties.

1.6. Influence of Probiotic Delivery Vehicles

A number of factors may influence the survival and physiologic state of probiotic bacteria and thus the health-promoting ability of these organisms in humans (Ranadheera et al. 2010, Sanders and Marco 2010). The nature of the delivery vehicle (food matrix, tablets or capsules) – including such parameters as water activity; pH; level and types of fat, protein, carbohydrates; presence of organic acids; oxygen level; and the presence of other functional ingredients – may be important in determining how a specific probiotic organism will behave in a product and when ingested by the consumer. In a review paper,

Sanders and Marco (Sanders and Marco 2010) pointed out that, “Little is known about the food matrix and product formulation impacts on probiotic functionality even though such information is essential to scientific understanding and regulatory substantiation of health benefits.” Although dairy foods, particularly yogurts, are the most common vehicles for delivering probiotic bacteria to consumers (more than 80% of yogurts are

19 supplemented with “probiotic” bacteria) (Granato et al. 2010), increasingly non-dairy products such as juice, cereal, chocolate and other foods are supplemented with probiotics due to problems such as lactose intolerance in many people, the undesired cholesterol content of fermented dairy products, and increasing demand for vegetarian probiotic-containing products (Heenana et al. 2004). Additionally, there are numerous other options for consumers to consume probiotics including capsules, tablets and sachets, which may or may not be supplemented with other functional ingredients.

Milk-based probiotics delivery systems are highlighted in Sanders and Marco’s review (Sanders and Marco 2010) for the potential bioactive components present in milk.

They point out that there could be interactions between particular probiotic bacteria and various milk components, for instance hydrolysis of milk protein to produce bioactive peptides. Azcarate-Peril et al. (Azcarate-Peril et al. 2009) demonstrated differential expression of genes required for adherence to mucin and intestinal epithelial cells when

Lb. acidophilus was grown in milk. Vinderola et al. (Vinderola et al. 2000) reported that hige milk fat content had inhibitory effects against probiotic cultures, particularly B. bifidum BBI in yogurts. In another study (Saarela et al. 2006), BAL BB-12 showed better stability in low fat milk during 2 weeks of refrigerated storage compared to orange, grape, or passion fruit juice. Moreover, their acid and bile tolerance were also much better in milk than in fruit juice, possibly due to the additional protective effect of the buffering capacity of milk.

In a recent study (Lee et al. 2015), Lee et al. fed dextran sulfate sodium (DSS) – induced ulcerative colitis mice with a wild-type and two mutant (DltD- and RecA-)

20 probiotic strains of Lb. casei (2 × 107 CFU/feeding) in milk or a nutrient-free buffer prior to and during administration of DSS for 15 consecutive days. Live Lb. casei cells in stool samples were recovered using selective MRS media. A disease activity index (DAI) was calculated based on percent total weight loss (before-/after DSS-treatment), histology score, the presence of blood in stools, and stool consistency. Cytokines and chemokines in ileal and colonic tissues were measured using Bio-Plex ProTM mouse cytokine 23-plex panel. The results of this study showed that mice fed with Lb. casei in milk had lower

DAI than those fed with Lb. casei in nutrient-free buffer, milk only, and mutant in either milk or buffer, suggesting that milk might be the preferred delivery matrix for certain probiotic strains.

Although some evidence suggests that delivery matrix (dairy vs. non-dairy) may play a role in the performance of some probiotic strains in vivo, it is not clear whether this delivery matrix effect applies to BB-12. It is important to explore which outcomes may be influenced to what extent in clinical setups. This will help to further understand the mechanism of probiotic effect of BB-12 and better exert their beneficial effect on hosts.

1.7. Production of BB-12 containing Yogurt

Although BB-12 is commercially available in various forms including bulk powder blend, bulk finished capsules, chewable tablets, stick packs, and powder for infant formula; it has traditionally been added to fermented dairy foods (primarily yogurt products) due to perceived health benefits. However, BB-12 containing yogurt intervention used in many clinical trials often include other probiotic strains or prebiotics

21

(de Vrese et al. 2011, Savard et al. 2011, Palaria et al. 2012, Fox et al. 2015, Ringel-

Kulka et al. 2015), which make it difficult to attribute observed effects to BB-12.

Additionally, there is no industrial standard regarding how to prepare probiotic yogurt in terms of the timing of probiotic addition. It is not clear how the timing of BB-12 addition

(before or after yogurt fermentation) may influence their survival in the product and the performance of probiotic in vivo. Merenstein et al. (Merenstein et al. 2010) conducted a clinical trial using a yogurt smoothie only contained BB-12 as probiotic intervention and an identical yogurt smoothie without BB-12 as placebo to study the potential benefits of

BB-12 to healthy children. Days of missed school were recorded, although no treatment effect was observed in this study. The same intervention was used to in a following study

(Merenstein et al. 2011), which was to investigate if consumption of yogurt containing high dose of BB-12 (1010 CFU/100 ml) decreases absences in children 2-4 years attending daycare/school centers. Again, no significant treatment effect was observed.

Since many beneficial effects associated with probiotic BB-12 were observed in diseased populations, it was proposed to use BB-12 containing yogurt as a natural means to prevent and treat various diseases, such as antibiotic-associated diarrhea and type-2 diabetes (Mohamadshahi et al. 2014). In order to study the effect of BB-12 containing yogurt in populations with disease states in the United States, phase 1 safety studies are required by the Food and Drug Administration (FDA). In a recent phase 1 study

(Merenstein et al. 2015), BB-12 containing yogurt smoothie was found to be safe and well tolerated by healthy adults concurrently taking antibiotics for a respiratory infection.

22

In the same study, BB-12 was also linked to changes in whole blood expression of genes associated regulation and activation of immune cells.

1.8. Enumeration of BB-12

To ensure the concentration of viable BB-12 in a food product or intervention remains at effective level (>109 CFU/serving), accurate enumeration of BB-12 is important. In the context of yogurt it can be challenging to identify one selective medium that fits all situations, i.e. in the presence of different cultures at various levels, or limited time window. A variety of selective media were developed based on de Man-Rogosa-

Sharpe broth (MRS) (de Man et al. 1960), including MRS-NNLP (nalidixic acid, neomycin sulfate, lithium chloride, paromomycin sulfate) (Laroia and Martin 1991) and

MRS-ABC (dicloxacillin, lithium chloride, L-cysteine hydrochloride) (Chr-Hansen

2005). However, researchers have found variations in the accuracy of BB-12 enumeration in different types of fermented milk products (Lapierre et al. 1992, Darukaradhya et al.

2006), not to mention the labor and space required by these media. By taking advantage of both PetrifilmTM Aerobic Count (AC) plates and conventional selective media,

Miranda et al. (Miranda et al. 2011) came up with PetrifilmTM AC plates associated

NNLP and ABC media. These media were found to be comparable with conventional media in terms of selectively recovering BB-12 in S. thermophilus fermented milk, especially PetrifilmTM AC plate associated with ABC medium. However, further studies are warranted to evaluate the effectiveness of these media to quantify BB-12 in products that are fermented with starter cultures, such as Lactobacillus spp. and Lactococcus spp.

23

Following selective enumeration, molecular-based verification provides additional assurance of a more reliable result of BB-12 enumeration. Ventura et al.

(Ventura et al. 2001) designed a PCR method using the primers, Bflact2 and Bflact5, that only target a region of the 16S rDNA (680 bp) of BAL. This approach was able to differentiate BAL from 29 Bifidobacterium and 9 Lactobacillus species, but it could not differentiate strains within the BAL species. Since it is generally believed that the beneficial effects of a given probiotic is specific to that strain, and cannot be regarded as general for other strains of the same species (Jungersen et al. 2014); it is of utmost importance to identify the genetic fingerprint of the probiotics claimed in a product.

Lomonaco et al. (Lomonaco et al. 2015) developed a single nucleotide polymorphism

(SNP)-based assay that can accurately characterize commercial BAL strains used in fermented dairy products.

Since it is proposed that BB-12 may benefit the host through multiple mechanisms including facilitating gut transit, modifying the gut microbiota, and production of short chain fatty acids, these parameters are important when evaluating the efficacy of this probiotic bacterium (Jungersen et al. 2014).

1.9. Fecal Transit Time

Fecal transit time, also known as bowel transit time or intestinal (gut) transit time, refers to how long it takes for the food to move from the mouth to the anus. Maintaining a regular bowel transit time is essential for health and general well being (Raahave 2015).

Delayed bowel transit time may have physical and psychological consequences that can

24 considerably diminish quality of life. Specifically, slow bowel transit time increases the risk of toxicity, anal fissures, hemorrhoids, fatigue, constipation, bloating, gas, diverticulitis, and weight gain (Lewis and Heaton 1999). On the other hand, too short bowel transit time may be associated with diarrhea, colitis, and some types of IBS. A healthy bowel transit time (approximately 12-14 hours as measured by the food dye method (Medline Plus Medical Encyclopedia 2014)) reduces toxin absorption, bloating, gas, indigestion, parasite infection, and the chances of diseases development in the colon.

It is estimated that 2-27% of the population in North America are suffering from chronic constipation, affecting 4-56 million adults in the U.S. (Sanchez and Bercik 2011).

The population suffering from delayed fecal transit time is much more extensive.

Intestinal transit time varies markedly between individuals, between male and female, even within individuals. Many factors can affect intestinal transit time, including age, diet, climate, education, income, and etc. Gut transit time can be measured in a number of ways: radiopaque markers (Hinton et al. 1969), scintigraphy (Krevsky et al. 1986), and through use of a wireless motility capsule (Cassilly et al. 2008). Each method has its pro and cons in terms of cost, invasiveness, accuracy, and ease of use. The lack of standardization in methods makes it challenging to assess the effect of probiotics on gut transit.

Numerous confusing and contradictory results have been reported in clinical trials regarding the effect of probiotic interventions on gut transit/motility due to an unclear definition of obstipation, inappropriate end-point markers, insufficiently detailed symptoms questionnaires, and unsatisfactory recording of health and well-being of the

25 subjects before the study (de Vrese and Schrezenmeir 2008). Some studies have shown probiotics can improve GI symptoms and reduce bowel transit time (Shanahan 2010,

Williams et al. 2010, Waller et al. 2011). Kashyap et al. (Kashyap et al. 2013) demonstrated that gut transit time could be influenced by altering microbiota in a humanized mouse model or by changing the diet. Based on this study, the authors proposed a model (Figure 1.1) for the interactions between diet, gut microbiota, and gut transit time in the host. According to this model, diet can independently influence both GI transit time and gut microbiota of the host. However, GI transit time can partially mediate the changes in gut microbial composition altered by diet. Diet-induced changes in functionality of gut microbiota may have led to the shift of GI transit time. It appears the effect of diet, gut transit time, and gut microbiota represents a highly interdependent and interactive environment. This study provides a useful model for the investigation of these complex interrelationships that can potentially offer new strategies for manipulation of human GIT, although such relationships may be much more complicated in humans.

26

Figure 1.1. A complex interplay between diet, the gut microbiota, and GI transit. (Kashyap et al. 2013)

A number of RCTs have shown various probiotic bacteria to be effective in improving gut transit time in a range of populations. In a randomized, double-blind, placebo controlled, parallel study, Bouvier et al. (Bouvier et al. 2001) demonstrated that ingestion of a fermented milk (3 x 125 g/day) containing the strain BAL DN-173 010 for

11 days significantly reduced total colonic transit time in healthy young adults (36 males,

36 females, mean age of 30 ± 8 years) by 21% and sigmoid transit time by 39%, compared to an identical fermented milk in which bacteria were killed by heat treatment.

Since these beneficial effects were not observed in the heat-treated product, it appears both probiotic survival and/or metabolic activity are important for improving colonic transit time. The authors did point out that the reduction of colonic transit time was always within physiological limits. It is worth noting the effect of BAL DN-173 010 on gut transit time was not observed on all volunteers and was more pronounced in women than in men.

27

In another triple-blind, randomized, placebo-controlled, dose-ranging study

(Waller et al. 2011), 100 healthy adults (mean age: 44 years, 64% female) with delayed whole gut transit time (females 55 ± 34 h, males 41 ± 24 h) were randomly assigned to one of the following capsule groups: 17.2 billion CFU of BAL HN019 (n = 33), 1.8 billion CFU of BAL HN019 (n = 33), or placebo (n = 34) for 14 days. The primary end point, whole gut transit time (WGTT), was measured on days 0 and 14 using radiopaque markers. Both treatment groups had a statistically significant shorter mean WGTT over the 14-day study period (33% and 25% decrease in high and low dose group, respectively) compared to baseline with no change observed in placebo group. In addition, most GI symptoms as assessed by questionnaires were significantly improved in both treatment groups while in the control group only 2/9 symptoms showed improvement. The findings of this study suggested consumption of probiotic strain BAL

HN019 is well tolerated, decreases WGTT in a dose-dependent manner, and reduces the frequency of functional GI symptoms in adults such as abdominal pain, nausea, constipation, irregular bowel movements, and flatulence.

The influence of BB-12 on gut transit time has also been studied. In a randomized, placebo-controlled, double-blinded, parallel dose-response study (Larsen et al. 2006), 71 (46 women, 25 men, mean age 25.6) out of 75 enrolled volunteers completed the study. The volunteers were randomly assigned into 5 groups receiving either placebo capsule or a capsule containing a mixture of BB-12 and Lb. paracasei

CRL-431 in concentrations of 108, 109, 1010, or 1011 CFU per day for 2 weeks. Diaries of bowel habits and well being, blood lipids, and fecal microflora were analyzed before and

28 after interventions. The results have indicated a significant linear increase in fecal consistency (looser stool) (P = 0.018) and a tendency to an increase in the frequency of defecation (shorter gut transit time) with increasing dosage of probiotics consumption.

Since a mixture of probiotics was used in this study, even the moderate effect observed could not be attributed solely to BB-12. A similar issue exists in other studies, i.e. Ringel-

Kulka et al. (Ringel-Kulka et al. 2008) have observed a decrease in colonic transit time in volunteers who received the probiotic BB-12 and prebiotic inulin. In another study

(Ringel-Kulka et al. 2015), a greater frequency of loose stool was detected in children (n

= 76) consumed a BB-12 and inulin containing yogurt for 16 weeks compared to the children (n = 73) in placebo group.

While precise mechanisms of probiotic effects on intestinal motility remain unclear, a few potential explanations have been proposed (Bouvier et al. 2001, Marteau et al. 2002, Vandenplas and Benninga 2009, Waller et al. 2011). These mechanisms, the majority of which involve interactions between gut mucosa and gut microbiota, include an increase in fecal bacterial mass, especially lactic acid-producing bacteria that may lower colonic pH and produce SCFAs, which may stimulate peristalsis; stimulation of cholecystokinin; deconjugation of bile salts; and production of phloroglucinol like compounds by bacteria. However, Marteau et al. (Marteau et al. 2002) demonstrated that shorter WGTT in women aged 18-45 years induced by BAL DN-173 010 was not a result of deconjugation or dehydroxylation of bile salts, but speculated a compound of bacterial origin that may decrease sigmoid tonus and stimulate colonic motility. More controlled clinical studies with clearly defined end-point markers and sufficient numbers of

29 volunteers are necessary for better understanding the effect of BB-12 itself on gut transit time and moving from helpful findings to conclusive results.

1.10. Gut Microbiota

The human gastrointestinal tract contains a diverse and dynamic community of microorganisms collectively termed the “microbiota”. The catalog of these microbes and their genes is defined as the gut microbiome (Human Microbiome Project Consortium

2012, Ursell et al. 2012). Interchangeably using these two terms has caused confusion to many researchers. The gut microbiota is composed of both autochthonous (native to a particular place) and allochthonous (derived from outside a system) residents. Their relative proportions are determined by many intrinsic and extrinsic factors including change to the diet (Eckburg et al. 2005, Turroni et al. 2008). The bacterial cells contained in the GIT exceeds the population of human cells in the body by at least a factor of 10

(Palmer et al. 2007, Turroni et al. 2008), and it has become clear over the last two decades this microbial community can play a critical role in health and disease of the host. It is suggested that changes in composition and diversity of the intestinal microbiota are related to disease, including inflammatory bowel disease (IBD), IBS and obesity

(Gerritsen et al. 2011). The functions of microbiota have been intensively studied and will continue to be a topic of interest in the scientific world. Putative functions include nutrient processing, protection against infection by pathogens, protection against epithelial cell injury, and regulation of host fat storage (Eckburg et al. 2005, Ley et al.

30

2005, Turroni et al. 2008). Thus, understanding the effect of exogenous factors, including diet and probiotics, on the ecology of the GIT is of critical importance.

1.10.1. Analysis of Microbiota

Historically, knowledge of the gut microbiota was limited to culture-based techniques, an approach that has been used since the early 20th century (Fraher et al.

2012). Recent advances in technology used to identify and to analyze components of the microbiome, the collection of genes that are harbored by the microbiota, have substantially improved our knowledge of the microbial communities associated with various habitats, including the human GIT (Kuczynski et al. 2012). Over the last decade a number of culture-independent molecular techniques have been developed to examine the microbiota of the GIT including qPCR, FISH, temperature gradient gel electrophoresis

(TGGE) and denaturing gradient gel electrophoresis (DGGE), RT-PCR, cloned 16S rRNA gene sequencing, terminal restriction fragment length polymorphism (T-RFLP),

DNA microarrays, NGS (454-pyrosequencing®, Illumina®, and SOLiD®) (Turroni et al.

2008, Fraher et al. 2012). These techniques overcame the limitation of culture-based approaches based on viable counting. However, each of the above mentioned culture- independent method has limitations. For example, qPCR, DGGE/TGGE, and Cloned 16S rRNA gene sequencing have PCR biases; FISH is not able to identify unknown species;

NGS is expensive and limited to the resolution of genus level and its data analysis is computationally intensive (Fraher et al. 2012). Because of developments in bioinformatics, decreased sequencing costs, growth in the culture database, NGS of the

16S rDNA has become an increasingly popular tool for microbiota studies (Eckburg et al.

31

2005, Ley et al. 2005, Fraher et al. 2012). Among NGS techniques, Illumina® sequencing currently provides greatest yield at the lowest cost and has been found to produce the best assemblies in a complex microbial community (Mende et al. 2012).

Since Illumnia® sequencing of the 16S rDNA technique can only identify bacterial members to the genus level due to the limitation of the read length, T-RFLP is a rapid and cost efficient supplemental method for the analysis of target species/subspecies.

T-RFLP is a method based on fragmentation of the 16S rDNA amplicons by restriction endonucleases followed by band visualization (Fraher et al. 2012), which can pseudo- quantitatively estimate the abundance of the taxa present. This method may be limited if a sample is heavily dominated by one species and there are subjective elements of the data analysis. However, these limitations do not prevent T-RFLP from providing useful inexpensive data on the most prevalent species present. It is especially useful as a supplemental tool to NGS. Lewis et al. (Lewis et al. 2013) developed bifidobacterial- specific T-RFLP (Bif-TRFLP) that have proven useful when monitoring intestinal bifidobacterial populations in infant fecal samples (Huda et al. 2014, Lewis et al. 2015).

NGS is the current gold standard for phylogenetically characterizing the components of the microbiota and quantifying the relative proportions of microbes present, although true metagenomics, implemented by shortgun sequencing of the gut microbiota, provides the most powerful data, including prediction of microbiota function

(Fraher et al. 2012). When studying the human gut microbiota, many technical questions need to be considered such as exclusion criteria (antibiotic use), controls, study design, sample collection and storage, DNA isolation and amplification protocols, sequence data

32 generation, analysis, and interpretation (Goodrich et al. 2014). Although each of the above-mentioned steps could influence the outcomes, sample handling and process are of special importance. Fecal sample storage and transit were thought to be critical as they may impact the DNA yield and quality. Studies have shown that after 24 hours of storage at room temperature, fecal genomic DNA and total RNA started to fragment, which may alter the relative abundance of taxa in samples (Cardona et al. 2012). However, short- term storage conditions were found to have little effect on diversity and structure of the gut microbiota (Carroll et al. 2012, Sergeant et al. 2012). Notably though, the number of freeze-thaw cycles may influence the composition of the gut microbiota as detection of some members in the phylum Firmicutes appear to increase after re-freezing samples

(Sergeant et al. 2012). In clinical trails, it is standard procedure for the volunteers to immediately store stool samples in a home freezer. Frozen samples are then transported to a research facility and stored at -80˚C upon receiving at the research facility (Goodrich et al. 2014). To avoid repeat the freeze-thaw cycle, it is recommended the samples be divided into enough aliquots for different applications when it is first thawed.

Selection of DNA extraction and amplification methods depends on what protocol is used and what equipment is available in a laboratory. However, more and more studies use widely employed protocols (HMP, EMP, or MetaHIT) and/or commercial kits. There is a wide range of primer sets available depending on the specific research question, the protocol is going to be employed, and the sequencing platform to be used. For example, the primer set 27F/338R is biased against the amplification of bifidobacterial 16S rDNA while the 515F/806R primer set amplifies sequences from both bacteria and archaea

33

(Kuczynski et al. 2012). The later primers were developed for the amplification of V4 region of the 16S rRNA gene, which would yield optimal community clustering with

Illumnia® sequencing platform (Caporaso et al. 2011).

1.10.1.1. 16S ribosomal DNA

Most culture-independent techniques for analysis of the gut microbiota are based on analysis of the 16S ribosomal DNA (16S rDNA or the 16S rRNA gene). 16S rDNA is the gene that codes 16S ribosomal RNA, a component of the 30S subunit of prokaryotic ribosomes. This piece of DNA is often used as an evolutionary chronometer, and contains both slowly evolving regions that can be used to design broad-spectrum PCR primers and fast evolving regions that can be used to classify microorganisms at finer taxonomic levels (family or genus) (Kuczynski et al. 2012).

1.10.1.2. Operational Taxonomic Units (OTUs)

Within the many millions of estimated bacterial taxa, the vast majority are unculturable and known only by DNA sequences (Curtis et al. 2002). Thus it is impossible to classify these microbes based on their phenotypes. Developing a method that can organize bacterial DNA sequences into biologically and ecologically meaningful taxonomic units may solve this problem. The most widely used practice for organizing bacterial diversity is to cluster sequences solely on the basis of DNA similarity at a conserved locus. Sequence clusters delineated in this manner are termed operational taxonomic units (OTUs) (Koeppel and Wu 2013). The most frequently used OTUs for

34 approximate bacterial species are clustered based on the 16S rRNA gene. A cluster of reads with 97% similarity is normally considered as the same species while 95% similarity is regarded as the same genera. However, Koeppel and Wu (Koeppel and Wu

2013) pointed out the limitations of OTUs, including arbitrarily defined similarity cut- offs, do not take phylogenetic information into account for different lineages evolve at different rates. Since there is no rigorous species concept for bacteria yet, OTU naming conventions remain useful because they employ shared vocabulary to discuss sequence- based observation of the gut microbiota (Goodrich et al. 2014).

It is important to employ a consistent OTU-picking strategy as this can have a major impact on downstream findings and interpretation of the data. There are mainly three OTU-picking algorithms: de novo, open-reference, and closed-reference

(Kuczynski et al. 2011). The open-reference strategy is the most widely used as it retains all sequence data while clustering reads against a reference database such as Greengenes

(McDonald et al. 2012), ribosomal database project (RDP) (Cole et al. 2009), or SOLVA

(Quast et al. 2013).

1.10.1.3. Sequence Data Analysis

NGS data sets are massive (up to 1.5 Tb) containing billions of reads

(http://www.illumina.com/systems/hiseq-3000-4000.html). Open-source bioinformatics pipelines such as QIIME (Caporaso et al. 2010) and MOTHUR (Schloss et al. 2009) are the most widely used tools employed for the study of microbial communities.

35

For 16S rDNA-based high-throughput sequencing, the data analysis workflow typically has three phases (Weinstock 2012). The first phase is to process and filter raw sequence data including addressing sequencing errors, read quality, chimaeras, and read length (Schloss et al. 2011). Following read processing, the second phase is to produce tables of taxa and abundance by comparing the results to 16S rDNA sequence databases or by clustering the reads into OTUs. The third phase involves generating trees or other representations of the similarity of communities, abundance curves, biodiversity plots, and other ecological and statistical descriptors of community structure using the derivative data from the previous phase. Both QIIME and MOTHUR are capable of doing the work but require the users to have a certain level of proficiency in bioinformatics. Both pipelines have their own standard operation procedure (SOP) for different kinds of sequencing data. The selection of pipeline in part depends on the user’s familiarity with different tools, the sequencing technique used, and the expected downstream analysis.

Using QIIME as an example for downstream analyses, α-diversity (within-sample species richness) and β-diversity (between-sample community dissimilarity) are widely used. These are based on the calculation of the weighted or unweighted UniFrac

(Lozupone et al. 2006) distance between samples. Principle coordinates are then calculated from the Unifrac distance matrices to decrease the dimensionality of the taxonomic dataset (OTU tables) into 3D principal coordinate analysis (PCoA) plots, enabling visualization of sample relationships. To determine whether sample classification (treatment, time points) caused differences in phylogenetic or species

36 diversity, ANOSIM (Clarke 1993) and permutational MANOVA (Anderson 2008) are usually employed to test significant differences between sample groups based on Unifrac distances. Significant taxonomic differences between sample groups can also be tested using the Linear Discriminant Analysis (LDA) effect size (LEfSe) (Segata et al. 2011).

LEfSe is an algorithm for high-dimensional (OTU tables) biomarker discovery and identification of genomic features (genes, pathways, or taxa) that characterize the differences between two or more classes/treatments. It first uses the non-parametric factorial Kruskal-Wallis (KW) sum-rank test to detect taxa with significant differential abundances with respect to the class of interest (one-against-all strategy). Then LEfSe uses LDA to estimate the effect size of each differentially abundant feature.

These command line based pipelines (QIIME and MOTHUR) are extremely flexible as many commands have multiple algorithms and most commands have adjustable parameters. For example, three different methods can be used for OTU- picking, more than a dozen measurements can describe the α-diversity of a given sample, and supervised or unsupervised classification methods can be used to calculate between- sample distance, which is the basis for β-diversity calculation. The specific parameter to use at each step is determined by the nature of the sample, the quality of the data and available metadata, the research question of interest, as well as other factors. Most importantly, it is recommended to practice consistent experimental and analytic methods throughout a well-designed study with minimum confounding variables and to keep good records of all possible metadata for downstream statistical analyses (Goodrich et al.

2014).

37

1.10.2. Is There a “Core Microbiota”?

Defining the characteristics of a “healthy gut microbiota” has proven difficult due to the great variations observed within and between individuals. A widely accepted concept is that the human gut microbiota influences normal physiology and susceptibility to disease through its collective metabolic activities and host interactions (Lozupone et al. 2012). In order to promote health by targeting the gut microbiota in clinical treatments, scientists have tried to characterize the baseline healthy microbiota as the initial step. Although a few large-scale projects such as the European Metagenomics of the Human Intestinal Tract (MetaHIT) (Qin et al. 2010) and the US Human Microbiome

Project (HMP) (Human Microbiome Project Consortium 2012) have made substantial progress towards this goal, the variation observed between and within individuals complicates the definition of what the ideal gut microbiota may be within a population or an individual. It is noteworthy that stool samples are usually used as the surrogate of gastrointestinal samples in gut microbiota studies, because they are ease to obtain and non-invasive.

A number of factors have been found to contribute to variations observed between and within individuals, including age, genetics, environment, diet, human health, and medicine. For example, age and geography/cultural tradition have been identified as primary drivers for variations observed in the gut microbiota of 531 healthy individuals from Malawi (n = 115), Venezuela (n = 100), and the USA (n = 316) (Yatsunenko et al.

2012). In the same study, greater variations in microbial communities and functional gene repertoires were observed among infants than in adults. However, there were shared characteristics of microbiota across individuals and populations, both compositional

38

(many bifidobacteria, and lower species richness than in adults) and functional (more genes that encode enzymes involved in folate biosynthesis in babies). Goodrich et al.

(Goodrich et al. 2014) have conducted a study on 416 twin pairs (171 monozygotic, MZ; and 245 dizygotic, DZ) to investigate the impact of genotype and early shared environment on their gut microbiota. They found that abundances of taxa in gut microbiota were more highly correlated within MZ than DZ twin pairs. The most highly heritable taxon was the family Christensenellaceae while Bacteroidetes community was primarily influenced by environmental factors. The authors suggested that human genetic state influences our microbial phenotype and this host genetic effect varies across taxa and includes members of different phyla.

In spite of the great variation in taxa present, the microbiota of the human gut is usually dominated by the phyla Firmicutes and Bacteroidetes. It is suggested that most individuals can be categorized into one of three variants or “enterotypes” based on the dominant genera (Bacteroides, Prevotella, or Ruminococcus) (Arumugam et al. 2011). A later study characterized human gut microbiota into Bacteroides and Prevotella enterotypes, which were driven by animal fat and carbohydrates diet, respectively (Wu et al. 2011). While it remains inconclusive whether our gut harbors a “core” set of specific bacterial taxa, studies have suggested that individual adults may actually share a functional “core microbiome” rather than a “core microbiota” (Lozupone et al. 2012,

Ursell et al. 2012).

In a study that addressed the question of how host genotype, environmental exposures, and host adiposity influenced the gut microbiome, Turnbaugh et al.

39

(Turnbaugh et al. 2009) characterized the gut microbial communities in 31 MZ twin pairs, 23 DM twin pairs, and available mothers (n =46). Individuals from the same family were found to possess more similar microbiota profile than unrelated individuals.

Individual’s gut microbiota varied in relative abundance of the major gut bacterial phyla at different time points, but these temporal changes in community structure are considered minor compared to interpersonal differences. More importantly, a shared core gut microbiome appears to exist at the level of metabolic functions but no single bacterial phylotype was detectable at an abundant frequency in the guts of all 154 sampled humans.

1.10.3. Gut Microbiota and Human Health

The observation of human gut microbiota in states of health and disease started as early as the 17th century by Antonie van Leewenhoek, who observed striking difference in microbes between samples (oral and fecal) from these two states (Dobell 1920).

Knowledge of the human gut microbiota has advanced exponentially over the past several decades based primarily on developments in molecular biology, especially advancements in high-throughput next-generation sequencing (NGS) techniques and related bioinformatics. Improvement of methodologies were mainly driven by the HMP (Human

Microbiome Project Consortium 2012) and the Earth Microbiome Project (EMP) (Gilbert et al. 2014) that were launched in 2008 and 2010, respectively. Since then, study of the human gut microbiota became more common in large population due to the well- established protocols and databases, as well as dramatic reductions in the cost of sequencing. A simple literature research on PubMed using the keyword “human gut

40 microbiota” resulted in 271 articles for the time window 01/01/1900 – 12/31/2007 and

3,703 articles for the period 01/01/2008 – 08/06/2015. Mounting evidence derived from these studies have shown changes in the composition of gut microbiota from “normal” state associate with numerous disease state. This change is known as dysbiosis. For example, AAD is caused by an imbalanced gut microbiota, resulting in both decreased fermentation of indigestible carbohydrates and rapid overgrowth of opportunistic pathogens (Alonso and Guarner 2013).

The gut microbiota begins to play a role on human health in infancy. The GIT of a normal fetus is believed to be sterile (Mackie et al. 1999), although a recent study has found that the placenta harbors a unique microbiota associated with a remote history of antenatal infection (Aagaard et al. 2014). Newborn infants acquire their microbiota

(homogenously distributed across the body) as soon as twenty minutes after birth and the profile of the microbiota is strongly influenced by the mode of delivery (Dominguez-

Bello et al. 2011), which may further impact the health of the infants. For example, the microbiota of vaginally delivered infants resembles the microbiota of their mother’s vagina, which is dominated by Lactobacillus and Prevotella spp. In contrast, cesarean section (C-section) delivered infants possess a microbiota that is similar to the skin, comprising Staphylococcus, Corynebacterium, and Propionibacterium spp. (Dominguez-

Bello et al. 2010). In this case, the microbiota of the baby is no more similar to their mother’s skin than to the skin of other women. Compared to vaginally delivered infants,

C-section babies suffer from higher incidence of skin infection with methicillin-resistant

Staphylococcus aureus, atopic diseases (Penders et al. 2007), allergies, and asthma,

41 which may possibly due to the lack of the natural first inoculum (Dominguez-Bello et al.

2011). A baby continues to acquire microbiota over the first few years of life. Their gut microbiota become adult-like as early as 1 year old and may be shaped by diet (breast- milk or formula), which may have further impact on their health (Palmer et al. 2007,

Bäckhed et al. 2015). In a study with 1,032 one-month old infants recruited from the

KOALA Birth Cohort Study in the Netherlands, it was found that exclusively formula- fed infants were more often colonized with E. coli, C. difficile, Bacteroides fragilis group, and lactobacilli than were their exclusively breastfed counterparts (Penders et al.

2006). Breastfeeding is associated with lower rates of infection in infancy; and reduction in blood pressure and total blood cholesterol, and lower risks of obesity and diabetes in adult life in high-income population (Robinson and Fall 2012, Denne 2015).

Evidence is mounting for a link between gut microbiota and metabolic syndrome, especially obesity. This is well substantiated in both animal models and human studies

(Bäckhed et al. 2004, Ley et al. 2005, Ley et al. 2006, Turnbaugh et al. 2006, Vijay-

Kumar et al. 2010, Million et al. 2012). In a gnotobiotic mice model, germ-free (GF) mice were inoculated by spreading the cecal contents of conventionally raised 8-week- old mouse (CONV-R) on their fur. The resulting conventionalized mice (CONV-D) were housed in gnotobiotic isolator and fed the same way as their GF counterparts. CONV-D mice were found to have rapid increase in body fat content despite reduced chow consumption as a result of elevated metabolic rate and increased energy deposition into host adipocytes, which is linked to the suppression of Fiaf, a circulating lipoprotein lipase inhibitor (Bäckhed et al. 2004).

42

Ley et al. (Ley et al. 2005) identified Firmicutes and Bacteroidetes as key players in obesity as genetically obese mice (ob/ob) had a statistically significant 50% reduction in Bacteroidetes and significantly greater proportion of Firmicutes (P < 0.05) compared to their lean counterparts. More interestingly, the obesity phenotype was found to be transferrable to GF mouse by colonization with “obese microbiota” in two separate studies. Turnbaugh et al. (Turnbaugh et al. 2006) found that ob/ob mice extract more energy from their diet and leave less behind in the feces while Vijay-Kumar et al. (Vijay-

Kumar et al. 2010) found GF mice tend to be hungrier and consume more food after being inoculated with the cecal contents of TLR-5 knockout mice. Promisingly, a correlation was found between gut microbiota and weight loss in humans. In a clinical trial with 12 obese individuals, Ley et al. (Ley et al. 2006) showed the relative proportion of Bacteroidetes is decreased while Firmicutes is increased in obese people by comparison with lean people. Increased abundance of Bacteroidetes correlated with percentage loss of body weight in subjects that were on two types of low-calorie diet (fat- restricted, FAT-R; carbohydrate-restricted, CARB-R) for 1 year. This correlation only held true in individuals who had lost >6% of their body weight on the FAT-R diet and

>2% on the CARB-R diet. David et al. (David et al. 2014) showed that animal-based diets can alter human gut microbiota as soon as 1 day after the diet reached the distal gut microbiota and that the gut microbiota reverted to their original structure 2 days after the animal-based diet ended. They also found that animal-based diets have a greater impact on the gut microbiota than a plant-based diets.

43

In another study, depleted levels of B. animalis and Methanobrevibacter smithii and enriched Lactobacillus reuteri were associated with obesity in humans (Million et al.

2012). The impact of gut microbiota on metabolic syndrome may be explained by the following mechanisms: 1) increased energy extraction by degrading non-digestible carbohydrates to monosaccharides and SCFAs; 2) modulation of host metabolism by regulation of the composition of bile-acid pools; 3) production of proatherogenic compounds from dietary choline such as trimethylamine (Tremaroli and Bäckhed 2012,

D'Aversa et al. 2013). These observations may serve as the basis for the manipulation of gut microbiota in weight management and metabolic disease control.

Alteration of gut microbiota has been linked to inflammatory bowel disease in both animal models and human studies (Sartor 2008). It has been shown in mouse models that the microbiota can induce colitis in genetically susceptible hosts, but it can also protect them against colitis as germ-reduced and GF mice had more severe disease in a

DSS-induced model when compared with their normal counterparts (Maslowski et al.

2009, Blumberg and Powrie 2012). The proposed mechanism of increased susceptibility in GF mice is because of decreased production of SCFAs in the absence of the commensal microbes, where SCFAs bind the G-protein-coupled receptor 43 (GPR43) and stimulate host inflammatory responses (Maslowski et al. 2009).

In a study on identical twins with Crohn’s Disease (CD), healthy twins were found to share more similar fecal microbial communities than twins with CD. Overall, patients with CD had less diverse fecal microbiota compared to healthy subjects and members of Bacteroides were the main contributors to this discrepancy (Dicksved et al.

44

2008). General IBDs have been associated with a shift in gut microbiota rather than a single causal bacterial taxa. For example, the gut microbiota of IBD patients was reported to have increased relative abundance of Proteobacteria and Actinobacetria, but decreased

Bacteroidetes, Lachnospiraceae, Bifidobacterium, C. leptum, C. coccoides,

Faecalibacterium prasnitzii, and Firmicutes/Bacteroidetes ratio (Clemente et al. 2012). It is suggested that pathobionts for IBD emerge from a dysbiotic commensal microbiota that are driven by a particular genetically determined, environmental induced maladapted host response (Blumberg and Powrie 2012). This complexity explains why IBDs remain medically incurable. Thorough understanding of the gut microbiome raises the possibility of personalized approaches to treat IBDs.

It has been established that the commensal gut microbiota is an important environmental factor for autoimmune diseases such as allergies and type 1 diabetes

(T1D). Studies have shown that low diversity of the gut microbiota in early infancy is an important risk factor for subsequent development of allergy (Bisgaard et al. 2011,

Abrahamsson et al. 2012, Marsland and Salami 2015). Decreased abundance of

Lactobacillus spp., Bifidobacterium adolescentis, Clostridium difficile, and Helicobacter pylori has been detected in patients with allergies relative to healthy counterparts

(Clemente et al. 2012). The incidence of allergic diseases is much higher in developed countries than developing countries and is linked to environmental factors such as the size of household, exposure to livestock, and increased antibiotic use. All these together have led to the hygiene hypothesis: “Increased sterility in Western life-styles leads to reductions in perinatal microbial exposure, which impairs microbiota-driven protective

45 mechanisms such as immune tolerance” (Blumberg and Powrie 2012). A good example would be the protective effect of H. pylori against allergic asthma in mouse models

(Arnold et al. 2011), in which the protective effect is particularly evident in neonatally infected mice through H. pylori specific immunological tolerance. This bacterium harbors in the stomach of almost all adults in developing countries but appears in less than 10% of Western-born children (Blumberg and Powrie 2012), which may partially explain the high incidence of allergies in developed countries. T1D are very common autoimmune- mediated metabolic disorders in children and young adults (Gülden et al. 2015). In a non- obese diabetic (NOD) mice model, gut microbiota has been shown to play a critical role in the development of T1D in MyD88 (an adaptor for multiple innate immune receptors that recognize microbial stimuli)-negative NOD mouse (Wen et al. 2008). One of the possible mechanisms seems to be the interaction between Toll-like receptors and commensal microbiota, where TLR4 serves as antidiabetogenic signal and TLR2 provides prodiabetic signaling (Burrows et al. 2015). Findings like this have drawn considerable interest to the hypothesis that the gut microbiota may play a critical role in

T1D development in humans. Although the studies are still at the very early stages and mechanistic data are barely available in humans, Giongo et al. (Giongo et al. 2011) have observed increased abundance in Bacteroides ovatus, decreased Firmicutes strain CO19, and less diverse and stable gut microbiota in T1D children compared to their healthy counterparts.

Gut microbiota has also been associated with other diseases such as autism

(Burokas et al. 2015), schizophrenia (Castro-Nallar et al. 2015), colorectal cancer

46

(Gianotti et al. 2010, Blumberg and Powrie 2012), celiac’s disease, anorexia, and type 2 diabetes (Clemente et al. 2012). As has been mentioned since many disease states are associated with imbalanced gut microbiota, it seems logical that restoring a healthy microbial community to prevent and treat diseases is an appealing approach. The most successful example is the treatment of recurrent C. difficile infection (CDI) via fecal microbiota transplantation (FMT) from a healthy donor (Ursell et al. 2012, Fuentes et al.

2014), where gut microbiota of patients shifted from low-diversity disease state, dominated by Proteobacteria and Bacilli, to a more diverse ecosystem resembling that of healthy donors, dominated by Bacteroidetes and Clostridium groups, including butyrate- producing bacteria. FMT has also shown promising potential of treating other dysbiosis- related diseases such as UC (Kunde et al. 2013), methicillin-resistant Staphylococcus aureus enterocolitis (Wei et al. 2015), IBDs (Wei et al. 2015), and other diseases (Xu et al. 2015). While safety, effectiveness, and acceptability of FMT are actively under investigations, manipulation of gut microbiota by a more acceptable manner - probiotics is an attractive approach to improve and maintain health.

1.10.3.1. Gut Microbiota and Probiotics

Within the gut microbiota, bifidobacteria and lactobacilli are generally considered beneficial organisms whereas proliferation of clostridia and members in the

Enterobacteriaceae family are linked to disease states (Jungersen et al. 2014). Since it is known that gut microbiota plays an important role in human health and disease, many

47 probiotic strains have been studied for their gut microbiota restoring potentials with the expectation of promoting beneficial bacteria and inhibiting undesired ones.

Can probiotics alter the gut microbiota? Yes. Even when using traditional culture techniques, feeding elderly with BAL HN019 at levels as low as 6.5 × 107 CFU/day resulted in increased counts of bifidobacteria, enterococci and lactobacilli, and reduced levels of enterobacteria in their stool samples (Ahmed et al. 2007). A more recent study by Cox et al. (Cox et al. 2010) demonstrated, using a 16S rDNA high-density microarray approach, that microbial communities with high abundance of LGG was associated with more even and stable taxa clusters, which are ecologically more resistant to perturbation and outgrowth of pathogens. The authors have also found a negative correlation between

LGG abundance and taxa that have been previously associated with atopy. In an animal study (Veiga et al. 2010), 4-week old T-bet-/-Rag2-/- UC mice were given either BAL DN-

173 010-containing fermented milk product (BFMP), non-fermented milk product (MP), or water (sham) for 4 weeks. A decreased in cecal pH and alteration in SCFA profile was observed in BFMP mice in comparison with other two groups, which explains the significant reduction of Enterobacteriaceae, a colitogenic taxa, in BFMP compared to MP nice. Additionally, increased levels of lactate-consuming Desulfovibrio spp., lactate- consuming, butyrate-producing Anaerostipes caccae subgroup, and Eubacterium hallii were detected in response to BFMP as compared with the MP. These microbial changes result in metabolic shifts, which further create a non-permissive environment for colitogenic T-bet-/-Rag2-/- Enterobacteriaceae. Although the presence of other organisms

(Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus) and their

48 metabolites in BFMP may confound the results, this study provides a framework for evaluating and optimizing probiotic-based functional foods and mechanistic insights of probiotic effects on IBDs.

While conventional methods (culture and molecular) have already substantiated that probiotics can influence the prevalence of gut microorganisms as summarized by

Gerritsen et al. (Gerritsen et al. 2011), high-throughput sequencing approaches provide new insights in the potential effect of probiotic interventions on the gut microbiota. One of the aims of early probiotic studies was to increase the diversity of host gut microbiota, because low gut microbial diversity has long been linked to disease states such as allergies and IBDs (Dicksved et al. 2008, Bisgaard et al. 2011, Abrahamsson et al. 2012,

Marsland and Salami 2015). In contrast, a high-throughput pyrosequencing approach has revealed that autistic children exhibited a higher richness and diversity of microbes in their feces when compared with their healthy counterparts (Finegold et al. 2010, De

Angelis et al. 2013). Since it is commonly acknowledged the interaction between gut microbiota and host is much more complicated than just a high or low microbial diversity, it is more meaningful to survey the global picture of microbial and metabolic changes after probiotic interventions via metagenomic and metabolomic approaches.

In a recent randomized, double blind, placebo-controlled, crossover study

(Ferrario et al. 2014), 30 healthy volunteers were assigned to either probiotic group (n =

14, receiving a capsule contains 2.4 billions of Lactobacillus paracasei DG (LDG)) or placebo group (n = 16, receiving an identical capsule without probiotic) for 4 weeks.

Then the two groups switch over to each other’s capsule for another 4 weeks after a 4-

49 week washout period. Fecal samples were collected before and after each treatment period and analyzed for fecal microbiota and SCFAs using high-throughput Ion Torrent

PGM sequencing technology and high performance liquid chromatography (HPLC), respectively. The results demonstrated intake of probiotic LDG beneficially influenced the gut microbiota at different taxa levels. Specifically, LDG induced an increase in

Proteobacteria (P = 0.006) and Coprococcus (P = 0.009) but a decrease in Blautia (P =

0.036). High abundant Coprococcus was hypothesized to prevent allergic diseases while high incidence of Blautia might be associated with IBS. LDG intake also influenced the participants’ fecal butyrate concentration. The authors noticed the impact of LDG on the gut microbiota and SCFAs seems to strictly depend on the initial state of the host’s intestinal ecosystem and identified fecal butyrate concentration as an important biomarker for the identification of subjects who may benefit more from probiotic interventions.

It is clear that probiotics can beneficially alter the host gut microbiota. It is generally understood the health benefits of probiotics are strain-specific and cannot be extrapolated to other probiotic stains of the same species, or other species, or mixtures of strains (Kekkonen et al. 2008, Gerritsen et al. 2011, Jungersen et al. 2014). With the technological advancements in omics (metagenomics, metabolomics, transcriptomics, and proteomics) approaches, knowledge of the molecular and physiological mechanisms behind specific diseases that are associated with microbial dysbiosis is accumulating exponentially and it is not unreasonable to expect tailor-made probiotics designed for specific populations at some future time.

50

1.10.3.2. Gut Microbiota and BB-12

Both animal studies and human trials have shown that BB-12, alone or in combination with other probiotics or prebiotics, is associated with desirable changes in microbiota such as an increase in beneficial bacteria and a reduction in potentially pathogenic bacteria. In an animal study, Lesniewska et al. (Lesniewska et al. 2006) demonstrated that BAL BB-12 supplementation (with LGG and inulin) significantly increased the fecal content of lactobacilli, and decreased enterobacteria compared to the control groups. The neuropeptide (Y and YY) response of the two age groups to intervention also differed significantly, suggesting that the physiological status of the GIT has a major role in the regulation of gut microbiota through the administration of synbiotics. Beneficial effects of BB-12 on human gut microbiota have been observed in both disease-compromised and healthy individuals.

In a randomized study, 21 infants with early onset atopic eczema of whom 8 were intolerant (highly sensitized group, HSG) and 13 tolerant (Sensitized group, SG) to extensively hydrolyzed whey formula (EHF) were included (Kirjavainen et al. 2002).

Infants in the SG group were tested for the probiotic effect of BB-12 by weaning 6 of them to placebo EHF and 7 to EHF supplemented with BB-12. Fecal samples were analyzed using fluorescence in situ hybridization (FISH). The fecal Bacteroides numbers increased in all infants and E. coli numbers increased in 4/6 infants in the placebo group while Bacteroides decreased in 4/7 infants and E. coli decreased in 4/6 infants in the BB-

12 group. Although this is a small-scale trial, the results suggest that BB-12 supplementation modulates the composition of the gut microbiota during weaning in a

51 manner that could contribute to the alleviation of atopic eczema. In a previously mentioned study (Mohan et al. 2006), BB-12 has also shown promising effect in preterm infants by reducing the cell counts of Enterobacteriaceae and Clostridium spp. in their feces.

A randomized, placebo-controlled, double blind, parallel dose-response study

(Savard et al. 2011) was conducted on 58 healthy volunteers (20 men, 38 women) to investigate the impact of 4-week commercial yogurt consumption supplemented with

BB-12 and LA-5 on their fecal microbiota using conventional culture method. Viable counts of fecal lactobacilli were significantly higher (P = 0.05) after probiotic intervention while enterococci counts were significantly lower (P = 0.04) when compared to placebo. Furthermore, these counts were significantly different from baseline in the probiotic groups but not the placebo group, suggesting a probiotic effect on the gut microbiota composition of healthy adults. In another study on healthy female adults

(Nishida et al. 2004), BB-12 beneficially altered the gut microbiota by reducing

Bacteroidaceae proportion but increasing Bifidobacterium and Lactobacillus counts in the host.

In a recent clinical trial (Dotterud et al. 2015), where 415 women with 36-week pregnancy were randomized to either a fermented milk supplemented with probiotics

(BB-12, LGG, and LA-5) or a placebo milk product. Stool samples of the mothers at 30 to 36 weeks of gestation and 3 months after giving birth, and of the babies at age 10 days,

3 months, 1 year, and 2 years were analyzed for colonization of the probiotics using quantitative PCR (qPCR); and stool samples of the babies at age 3 months and 2 years

52 were analyzed for microbiota profile via Illumnia sequencing of the 16S rRNA gene. The results indicated that all probiotics colonized in mothers while only LGG transiently harbored in babies at age 10 days and 3 months. No difference in gut microbiota of the infants was observed between the probiotics group and the placebo group. Thus, no evidence was found that the BB-12 containing probiotic intervention altered the gut microbial composition of infants.

In summary, the influence of BB-12 consumption on gut microbiota composition has been mixed. Some studies have observed a significant positive influence and a few other studies have found no BB-12 effect on the composition of host gut microbiota

(Maccaferri et al. 2012, Dotterud et al. 2015). This discrepancy can be attributed to a few factors: the method of BB-12 administration (alone or in combination with other strains or ingredients, in dairy or non-dairy matrix), daily dosage of BB-12, duration of intervention, target population, end points measured (different target bacterial groups), and methodology of measurement (culture-dependent or independent, high-throughput sequencing or others). The putative mechanisms by which BB-12 may influence gut microbiota include reduction of luminal pH, competition for mucosal adhesion, and immune interaction with the host (Jungersen et al. 2014).

1.11. Fecal Short Chain Fatty Acids (SCFAs)

The metabolic composition of fecal samples provides a window for phenotypically elucidating the complex relationship between the gut microbiota and the host influenced by probiotic interventions. SCFAs are organic fatty acids with 1 to 6

53 carbon atoms, which are the main anions resulting from bacterial fermentation of carbohydrates and protein in the colon (Wong et al. 2006). They are very important compounds for maintaining a healthy GI environment as they promote the growth and differentiation of epithelial cells, and provide energy to other tissue/organs of the host such as muscle, kidney, heart and brain (Macfarlane and Macfarlane 2012). SCFAs are primarily produced in the proximal colon, where bacterial numbers, proliferation, and substrate availability are the greatest; however, the concentrations of the principal SCFAs

(acetate, propionate, and butyrate in a molar ratio of 60:20:20) in the proximal colon and distal colon were found to be similar as measured in the gut contents taken from victims of sudden death (Wong et al. 2006, Wong et al. 2012). Gas chromatography (GC) is commonly used for the measurements of SCFAs (Garcia-Villalba et al. 2012), however, high performance liquid chromatography (HPLC) has also been used in SCFA studies

(Riezzo et al. 2012, De Baerea et al. 2013). A number of factors are found to influence the production of SCFAs, including gut microbiota, gut transit time, diet, and host physiological status (Roy et al. 2006, Rahat-Rozenbloom et al. 2014). Since the vast majority of the SCFAs (>95%) are absorbed from the gut, it is not surprising that fecal

SCFAs measurements are not sensitive to treatments designed to simulate SCFAs production. Average total SCFA concentrations over 12 weeks were found to range from

36.9 to 144.4 mmol/kg as measured in a total of 86 fecal specimens of 8 healthy volunteers (McOrist et al. 2008). The same as in studies on the effect of dietary fiber on

SCFA production (Canfora et al. 2015), inconsistent results have been reported regarding the probiotic effects on fecal SCFA levels.

54

Probiotics have shown promising effects on SCFA production in mice models.

Martin et al. (Martin et al. 2008) studied probiotic modulation of the gut microbiota – host interactions in a humanized microbiota mouse model. They reported probiotic consumption modified the microbiota resulting in different SCFA profiles. Specifically, both Lactobacillus paracasei NCC2461 and Lactobacillus rhamnosus NCC4007 supplemented mice exhibited reduced levels of acetate and butyrate. Increased concentrations of isobutyrate and isovalerate were observed in the mice fed with

NCC2461. Concurrently, reduced populations of Bifidobacterium longum and

Staphylococcus aureus as measured by selective media were observed after introduction of both probiotics. The authors suggested there is a significant association between the probiotic modulation of the gut microbiota and the metabolism of SCFAs. In another mouse study, Hong et al. (Hong et al. 2010) demonstrated that probiotic lactic acid bacteria (a mixture of Lactobacillus brevis HY7401, Lactobacillus sp. HY7801, and

Bifidobacterium longum HY8004) treated mice had increased levels of acetate, butyrate, glutamine, and decreased levels of trimethylamine compared to the control group suggesting probiotics have protective effects against DSS-induced colitis in mice.

In an in vitro study, Vitali et al. (Vitali et al. 2012) examined a total of 84 different fecal metabolites belonging to the families of sulfur compounds, nitrogen compounds, aldehydes, ketones, esters, alcohols, phenols, organic acids, and hydrocarbons influenced by the addition of probiotic, prebiotic, and synbiotic to batch fecal fermentations of healthy individuals. Butyrate was found to be the metabolite most extensively affected by the addition of probiotics and prebiotics as compared to other

55 compounds. This study provides an in vitro model that could be used as predictive model to assess the metabolic implication of new probiotic organisms and prebiotic substrates.

However, great care must be taken when in vitro and animal studies are extrapolated to the humans.

Human studies also showed inconsistent findings, although variable organisms were used in different populations. In a clinical trial 25 volunteers (10 patients on long- term total enteral nutrition (TEN), and 15 healthy subjects (HVO)) (Schneider et al.

2005), were given a probiotic yeast, Saccharomyces boulardii, for 6 days. Stool samples were collected at baseline, on the last 2 days of treatment, and 9-10 days after treatment.

Fecal SCFAs and fecal microbiota were measured using GC and culture method, respectively. No treatment effect on fecal SCFAs was observed in the HVO group.

Butyrate concentrations and total SCFAs increased significantly in the TEN group during treatment compared to baseline and this difference disappeared 9 days after cessation of the intervention. Significant differences in gut microbiota were observed between the

HVO and the TEN at baseline as the latter had higher counts of aerobic bacteria and lower anaerobes/aerobes ratio. No change in gut microbiota was observed in the TEN patients during treatment; however, the conventional culture approach may not be sensitive enough to detect any microbes that could be affected by the intervention.

In another double-blind, crossover human study (Riezzo et al. 2012), 20 constipated patients (3 males and 17 females at an average age of 38.8 ± 14.4 years) were randomized given either 180 g per day of ordinary artichokes (control) or artichokes enriched with probiotic Lactobacillus paracasei IMPC 2.1 (2 × 1010 CFU per day) for 15

56 days. Stool consistency, GI symptoms, and fecal SCFAs were evaluated before and after each treatment period. Propionate concentrations were significantly increased during the probiotic period as compared to baseline, but there was no difference between control period and probiotic period. However, GI symptoms of the patients were significantly improved during the probiotic-enriched artichokes period relative to the ordinary artichokes period and 80% of the patients preferred the former treatment to later ones.

In a healthy cohort of 26 healthy adults (mean age 25 years) (Klein et al. 2008) half of the volunteers consumed 300 g/day of yogurt supplemented with probiotic strains

Lactobacillus acidophilus 74-2 (9.3 × 108 CFU/g) and BAL DGCC 420 (3.0 × 106

CFU/g), and the other half took a placebo product for 5 weeks. The outcomes measured were fecal probiotic contents, fecal SCFAs concentrations, serum lipids and plasma immune markers. Fecal probiotic proportions increased significantly (P < 0.05) so did some of the immune markers during probiotic intervention. No changes in fecal SCFAs or serum cholesterol levels were observed in any treatment group. However, serum triacylglycerol did decrease by 11.6% (P < 0.05) in the probiotic group.

BAL, alone or in combination with other probiotics and/or prebiotics, has shown varied effect on fecal SCFA production. In adult patients with atopic dermatitis (AD),

BAL LKM512 containing yogurt tended to increase fecal butyrate concentration of the patients compared to baseline (P = 0.089) (Matsumoto et al. 2007). Concurrently, serum

IFN-γ significantly increased 6 fold after LKM512 yogurt consumption (P < 0.005) and 3 fold after placebo yogurt treatment (P < 0.05) compared to baseline. Shifts in fecal microbial profile were also observed in both probiotic and placebo groups. The authors

57 concluded that LKM512 yogurt consumption may be effective against intractable adult- type AD, and this effect in part relies on the induction of IFN-γ by butyrate produced by the altered gut microbiota. In another study with 22 healthy physically active men (West et al. 2012), no changes in fecal SCFA concentrations in either treatment group

(synbiotic capsule or prebiotic capsule) were observed, although synbiotic supplementation significantly altered the composition of the fecal microbiota based on

DGGE results. Laake et al. (Laake et al. 1999) did not detect any changes in fecal SCFA concentrations in patients with ileal-pouch-anal-anastomosis (IPAA) who received either

BB-12 and LA-5 containing milk (regular Cultura) or the same milk but heat-treated

(heat-treated Cultura) for a week, although significant improvements were observed only in the regular Cultura group such as increased recovery of fecal lactobacilli and bifidobacteria and reduced stool frequency.

It is commonly understood that SCFAs present in the colon and the bloodstream are primarily from the fermentation of carbohydrates and protein by the gut microbiota

(Roy et al. 2006, Wong et al. 2006, Macfarlane and Macfarlane 2012, Wong et al. 2012,

Pluznick 2014). SCFA production has long been associated with changes in gut microbiota and gut transit time (Wichmann et al. 2013) under the influence of probiotics, prebiotics, or synbiotics, but the findings are not conclusive or consistent. To better understand their complicated relationships, it is vital to evaluate fecal SCFAs simultaneously with the gut transit; more importantly, with the survey of the gut microbiota via NGS approaches. In efforts to define the functional identity of a probiotic strain, it is recommended to use one single probiotic strain at a time.

58

1.12. Conclusion

Bifidobacterium animalis subsp. lactis BB-12 is a widely used probiotic bacterium commonly added to fermented dairy product, particularly yogurt, due to perceived health benefits. However, BB-12 has increasingly been provided in non-dairy deliver matrixes such as capsule, tablet, powder, and juices. Although the findings remain inconclusive, studies have shown beneficial effects of BB-12 to the host including accelerating gut transit, modulating immune function, improving gut microbiota, and increasing fecal SCFA production. Since BB-12 continues to gain popularity in a wide range of products (dairy vs. non-dairy), it is important to understand how the delivery matrix (dairy vs. non-dairy) and the fermentation process may influence the probiotic in the product and its performance in humans.

59 1.13. Hypothesis and Specific Aims

We hypothesize that the vehicle used to deliver probiotic bacterium B. animalis subsp. lactis BB-12 into the body influences the performance of the probiotic in vivo.

Specifically, consumption of a yogurt-based smoothie containing BB-12 will result in greater decrease in gut transit time (GTT), as well as having a greater effect on the composition of the fecal microbiota, and increased fecal short chain fatty acids (SCFAs) than the same probiotic bacterium delivered at the same level in the form of a dietary supplement (capsule).

To test the hypothesis, the specific aims were: 1) develop three different BB-12 delivery systems (BB-12 added before yogurt fermentation, BB-12 added after yogurt fermentation and BB-12 containing capsules) and study the viability of BB-12 in the yogurt smoothies over 4 weeks of storage at 4°C, 2) assess the influence of consumption of yogurt smoothie (with and without BB-12) and BB-12 containing capsules on the participants’ gut transit time using a wireless motility capsule, 3) study the effect of the interventions on the participants’ fecal SCFA concentrations, and 4) evaluate the impact of the interventions on the participants’ gut microbiota through Illumina® sequencing of the 16S rRNA gene.

60

Chapter 2 - Characterization of Bifidobacterium animalis subsp. lactis BB-12 Containing Yogurt Smoothie Interventions

2.1. Introduction

Yogurt is a traditional cultured milk product that can be dated back to as early as

6,000 B.C., when the Neolithic people of central Asia first developed the product

(Donovan and Shamir 2014). The word “yogurt” was derived from Turkish yoğurmak, meaning to thicken, coagulate, or curdle (Fisberg and Machado 2015). Yogurt was popularized as a foodstuff throughout Europe in the early 1900s by the work of Elie

Metchnikoff, a Russian Nobel Prize winner. Under the influence of Stamen Grigorov’s discovery of lactic acid bacteria (LAB) in yogurt, Metchnikoff suggested that daily doses of LAB in “soured milk” (yogurt) in the Bulgarian diet was one of the reasons for their unusually large number of centenarians (Metchnikoff 1908, Mackowiak 2013).

Nowadays, yogurt has become a common food item throughout the world in a wide range of formats represented by different names. For example, dahi is an Indian yogurt known for its characteristic taste and consistency (Satish Kumar et al. 2013). Different countries have varied legal definition for yogurt products. According to the U.S. Code of Federal

Regulations (21 CFR 131.200), yogurt is defined as a food produced by culturing cream, milk or skim milk with Lactobacillus delbrueckii subsp. bulgaricus (LB) and

Streptococcus thermophilus (ST), and contains not less than 3.25% butterfat (0.5% - 2% for low-fat yogurt), no less than 8.25% milk solids, and the mix has to be pasteurized prior to the addition of culture.

61

Yogurt is one of the original “healthy foods” and yogurt consumption has been associated with a number of gastrointestinal related health benefits including alleviation of lactose intolerance, reduction of constipation, mitigation of diarrheal diseases, protection against colon cancer, prevention of inflammatory bowel disease, inhibition against Helicobacter pylori, and prevention of allergies (Adolfsson et al. 2004). Yogurt intake has also been associated with bone health (Rizzoli 2014), and weight management

(Jacques and Wang 2014). Meanwhile, probiotics have become increasingly popular due to perceived health benefits such as alleviation of diarrhea, modulation of host immune system, reduction of allergic responses, decreased bowel transit time, prevention of infections, and numerous other effects (de Vrese and Schrezenmeir 2008, Sherman et al.

2009). Yogurt and yogurt-like products are often used as carriers of probiotics because consumers have long perceived yogurt as a healthy food with live organisms in it (Heller

2001).

Bifidobacterium animalis subsp. lactis BB-12 (BB-12) is a probiotic strain that has been used as a food ingredient and food supplement worldwide since 1985 with well- documented health benefits. It has been commonly added to fermented dairy products but is becoming available in a variety of non-dairy matrixes (capsule, tablet, powder, juices).

In vitro study has shown that BB-12 tolerates acid and bile better when stored in milk compared with those in phosphate buffered saline or juice (Saarela et al. 2006), but excessive growth of BB-12 in yogurt during fermentation may lead to unpleasant flavor due to acetic acid production (Oliveira et al. 2012). It is obviously important to monitor the concentration of BB-12 in a yogurt product, but it can be challenging to accurately

62 enumerate BB-12 in a complex food system, especially in the presence of lactic acid bacterial starter cultures ST and LB.

Little is known about the effect of delivery matrix (dairy vs. non-dairy) on the efficacy of BB-12 in humans. In addition, it is not clear how yogurt fermentation may impact the survival of BB-12 in the product or its performance in vivo. However, a recent animal study has suggested milk may be the preferred delivery matrix for certain probiotic strains (Lee et al. 2015). Therefore, the primary objective of this portion of the study was to develop three yogurt smoothies based on a product developed in our lab

(Merenstein et al. 2010, Merenstein et al. 2011, Merenstein et al. 2015); specifically 1) a control yogurt smoothie without BB-12 (YS), 2) a yogurt smoothie with BB-12 added before yogurt fermentation (PRE), and 3) a yogurt smoothie with BB-12 added after yogurt fermentation (POST). The second objective was to evaluate the survival of BB-12 in the yogurt smoothies throughout the shelf life.

2.2. Materials and Methods

2.2.1. Bacterial cultures

Culture Bifidobacterium animalis subsp. lactis BB-12 (BB-12) was obtained from

Chr. Hansen (Milwaukee, WI). The yogurt starter culture YF-L702 (a blend of different

Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus strains) was also obtained from Chr. Hansen and was added at a rate of 0.02% of the volume of yogurt mix. Cultures were obtained as concentrate, frozen pellets and held at -80°C prior to use in production.

63

2.2.2. Production of yogurt smoothie interventions

In previous studies (Merenstein et al. 2010, Merenstein et al. 2011, Merenstein et al. 2015), BB-12 was added to yogurt following fermentation (POST). Regarding this study, an additional product was developed where the probiotic was added prior to yogurt fermentation (PRE). The low-fat yogurt-based strawberry smoothies were manufactured in the Berkey Creamery at The Pennsylvania State University. The interventions with

BB-12 added either pre- or post-yogurt fermentation contained 1.1-1.3 × 108 colony- forming unit (CFU) per g. The serving size of 240 g delivered a dose level of log10 10 ±

0.5 CFU BB-12 per day. Capsules (CAP) containing log10 10 ± 0.5 CFU BB-12 were produced by Chr. Hansen specifically for this study.

To make the yogurt smoothie interventions, a yogurt mix was formulated to the target composition highlighted in Table 2.1 using the TechWizardTM software (Owl

Software, Columbia, MO). Processing details of the yogurt smoothies are listed in

Appendix I. Briefly, milk and dry ingredients for the yogurt base were selected and blended. The resulting yogurt mix was pasteurized at 84.4˚C for 41s and homogenized first stage at 2000 psi and second stage at 500 psi. Yogurt mix was pumped to a fermentation tank and heat-treated at 85°C for 30 minutes, cooled to 43.3°C and inoculated with 0.02% yogurt starter culture YF-L702. After inoculation a portion of the mix for manufacture of the PRE yogurt fermentation product was drawn off and inoculated with BB-12 (Figure I-1). Inoculated yogurt mixes were allowed to ferment to a pH of 4.6. At this time, they were cooled to 21°C and then a heat-treated slurry (heated to 85°C, held for 1 h, and then cooled to 38°C) contains strawberry puree (Sensient

Technologies, Milwaukee, WI), pectin AMD783 (Danisco, New Century, KS), 36DE

64 corn syrup solids (Ingredion, Mississauga, Ontario, Canada), sugar (Golden Barrel,

Honey Brook, PA) and water was added and blended into the yogurts in a steam kettle until uniform. The strawberry yogurt smoothies were then homogenized at first stage pressure of 1000 psi, second stage pressure of 500 psi and pumped back to the fermentation tank. Next, the yogurt smoothie was split into two groups and one group was inoculated with BB-12 (post-addition treatment, POST) while another was a BB-12 free product (Control, YS). Finally, the strawberry yogurt smoothies were homogenized again to produce a drinkable product. The separately fermented yogurt was combined with the slurry in the same steam kettle and further homogenized to produce the pre- addition treatment (PRE). The three products were identical except for the addition of

BB-12 and timing of addition.

Table 2.1. Target yogurt composition

Component Target Composition (w/w) Milk Fat 1.3% Milk Solid Non-Fat 9.0% Sucrose 3.0% Total Solids 13.3%

Table 2.2. Composition of strawberry yogurt smoothies

Component Composition in Yogurt Smoothie (w/w) Milk Fat 1.0% Milk Solids Non-Fat 6.8% Sucrose 6.1% Corn Syrup Solids 6.4% Pectin 0.4% Strawberry Puree 3.0% Total Solids 23.7%

65

2.2.3. Shelf life and safety of the yogurt smoothies

The final yogurt smoothies were evaluated for the presence of coliforms using high sensitivity PetrifilmTM Coliform Count Plates (3MTM, Burlington, NC) and

PetrifilmTM Aerobic Plate Counts according to the 17th edition of Standard Methods for the Examination of Dairy Products, #7.072 and #6.040, respectively. Shelf life of the yogurt smoothies was determined by monitoring the viable counts of BAL (target of log10

10 ± 0.5 CFU per 240 g serving) in the yogurt smoothies. Additionally, pH of the products was measured using a pH meter (VWR).

2.2.4. Selective enumeration of BAL

The population of B. animalis ssp. lactis BB-12 in the yogurt smoothies was determined following manufacture and throughout the shelf life of the products. After dilution of yogurt smoothie in 0.1% peptone water (3MTM, St. Paul, MN) using a stomacher (Gemini B.V., Apeldoorn, Netherlands), suitable dilutions were pour plated on selective media for bifidobacteria. The de Man, Rogosa, and Sharpe (MRS) agar (de Man et al. 1960) was prepared by adding 15 grams of Agar (Difco, BD, Sparks, MD) to 1000 ml MRS broth (Difco, BD). The media was autoclaved at 121°C for 15 min, tempered to

47°C and then 0.00005% (w/v) sterile dicloxacillin (Sigma-Aldrich, St. Louis, MO),

0.01% (w/v) lithium chloride (Sigma-Aldrich), and 0.005% (w/v) cysteine hydrochloride

(Sigma-Aldrich) dissolved in distilled water were added to make a modified MRS

(mMRS). The dicloxacillin, lithium chloride, and cysteine hydrochloride solutions were sterilized by filtration (0.2-µm pore size).

66

Lactobacillus bulgaricus (LB) and Streptococcus thermophilus (ST) were also enumerated following manufacture and throughout the shelf life of the products using acetic acid (VWR) titrated MRS (pH 5.4) (aMRS) agar and M17 (Difco, BD) agar supplemented with 1% lactose (Difco, BD) (M17L), respectively. Serial diluted samples were pour-plated for BAL and LB and incubated at 37°C under anaerobic conditions (5 mm Hg vacuum pressure, anaerobic mixed gas: 10% carbon dioxide, 5% hydrogen, 85% nitrogen; VWR, West Chester, PA) for 72 h. ST plates were incubated aerobically at

37°C for 48 h.

To evaluate the selectivity of the media mMRS, aMRS, and M17L, an assay was conducted by pour plating a stock culture of BB-12, YF-L702, and BB-12 + YF-L702 on each medium. The growth of ST on the BAL selective medium mMRS interfered the enumeration of BB-12 when its concentration was similar or higher than BB-12 in the yogurt. Because of this an alternative medium MRS-NNLP (nalidixic acid, neomycin sulfate, lithium chloride, paromomycin sulfate) (Laroia and Martin 1991) agar was identified as an alternative selective medium for BAL enumeration in yogurt. Nalidixic acid, neomycin sulfate, and paromomycin sulfate were purchased from Sigma-Aldrich.

The preparation of MRS-NNLP agar was the same as the modified MRS agar with the selective agents sterilized by filtration and added after tempering. Additionally, the viable count of BB-12 in capsules was evaluated periodically throughout the study to assure the product had the target count.

67

2.2.5. Verification of BAL

Species-specific PCR (Ventura et al. 2001) was employed to verify the colonies as BB-12 on selective media were indeed BAL. Specifically, five colonies were randomly picked from each countable plate and analyzed by PCR using primers Bflac2

(GTGGAGACACGGTTTCCC) and Bflac5 (CACACCACACAATCCAATAC). Both primers were obtained from IDT DNA (Coralville, IA). Cells were lysed according to the microwave method of Bollet et al. (Bollet et al. 1991) as modified by Kullen et al.

(Kullen et al. 1997). The PCR amplification mixture consists of 10 × Buffer with MgCl2

(Promega, Madison, WI), 100 µM dNTPs (Promega), 1.0 µM of each primer, and 2.5 U

Taq polymerase (Promega) brought to 25 µl with sterile distilled water. PCR for each sample was performed using the following conditions: initial denaturation, 95°C, 5 min; denaturation, 95°C, 30 s; anneal, 58°C, 60 s; elongation, 72°C, 2 min (repeat for 30 cycles); final elongation, 72°C, 7 min; cool to 4°C in a Mastercycler Personal

(Eppendorf, Hamburg, Germany). Amplicons were detected following agarose gel electrophoresis and staining with eithidium bromide. A 100-bp DNA ladder (Promega) molecular weight marker was electrophoresed at the same time to estimate the size of

PCR products. An Alpha-Imager (Alpha Innotech, San Leandro, CA) was used to visualize the bands (Figure 2.S1).

2.2.6. Composition of the yogurt smoothies

Percent fat and percent total solids of the products were measured in the Quality

Assurance Laboratory of the Berkey Creamery using CEM Smart Trac Fat and Moisture

Analysis equipment (CEM Corporation, Matthews, NC).

68

2.2.7. Statistical analysis

All data were first tested for normality using the Anderson-Darling test. The counts of bacterial colonies were converted to logarithm scale with base 10 and analyzed using the Student’s t-test or ANOVA and Tukey test (P < 0.05). Regression was used to compare changes in BB-12 concentration in the two yogurt smoothies (P < 0.05) throughout the shelf life. All analyses were performed using Minitab 17.0 software

(Minitab Inc., State College, PA).

2.3. Results and Discussion

The objectives of the study were to develop three yogurt smoothies that were identical in terms of composition (fat and total solids): a control (No BB-12) and two products that differed in the timing of BB-12 addition (pre- and post-fermentation) and to evaluate survival of BB-12 in the yogurt smoothies throughout shelf life when stored at

4°C. To minimize the number of confounding variables, it is critical to ensure the study interventions are comparable. In this case, the fat and total solids are designated indicators for composition comparison; pH is a quality parameter; and the concentration of BB-12 in the products is considered the key variable.

2.3.1. Development of PRE-added yogurt smoothie

Results from preliminary trials indicated a slightly growth of BB-12 during yogurt fermentation in the pre-added treatment (Table 2.3), but the PRE and POST added yogurt smoothies had the same concentration of viable BB-12 in the finished products.

69

Table 2.3. BB-12 population in different samples as enumerated by MRS-NNLP1 agar

CFU/g in yogurt2

BB-12 in PRE before yogurt fermentation (1.02 ± 0.1) × 108b

BB-12 in PRE after yogurt fermentation product (1.27 ± 0.1) × 108a

BB-12 in POST after yogurt fermentation product (1.26 ± 0.1) × 108a

1MRS agar supplemented with nalidixic acid (15 µg/ml), neomycin sulfate (100 µg/ml), lithium chloride (3 mg/ml), paromomycin sulfate (200 µg/ml), and cysteine hydrochloride (0.5 mg/ml). 2Data shown were mean ± SD from at least four replicates. Values in a column without a common letter are significantly different, P < 0.05. PRE yogurt smoothie with BB-12 added before fermentation, POST, yogurt smoothie with BB-12 added after fermentation.

2.3.2. S. thermophilus grows on modified MRS but not on MRS-NNLP

In previous work (Merenstein et al. 2010, Merenstein et al. 2011, Merenstein et al.

2015) where the BB-12 concentration was substantially higher than the population of ST and LB, mMRS had proven sufficiently selective for BAL. In the present study, employing low levels of BB-12, it was observed that mMRS was not sufficiently selective and that ST could grow on the mMRS agar when the ST concentration was (~1.3

× 109 CFU/g) higher than that of BB-12 (~1.5 × 108 CFU/g) in the product.

In order to overcome this issue, a different selective medium MRS-NNLP was evaluated. To assess the efficacy of MRS-NNLP agar on recovery of BB-12, an experiment was conducted assessing the viable counts of BB-12 on MRS-NNLP agar and its optimum medium MRS agar, respectively. Results indicated that MRS-NNLP agar resulted in slightly lower BB-12 count than MRS (log10 10.44 vs. log10 10.20) but the difference was not statistically significant (P = 0.16). The same was observed when enumerating BB-12 in the capsules using both MRS and MRS-NNLP (Table 2.4).

70

Previous studies (Dave and Shah 1996, Tharmaraj and Shah 2003) have found that MRS-

NNLP agar could inhibit 40%-70% of some bifidobacteria strains, but it could effectively select BAL against other LAB. The selectivity of MRS-NNLP was also confirmed using the yogurt culture YF-L702 (Table 2.5). While the inhibitory effect of MRS-NNLP agar against BAL resulted in lower absolute colonies in the present study, experiments showed

ST was inhibited. Additionally, the goal of the BB-12 enumeration was to assure the BB-

12 containing interventions (two yogurt smoothies and capsule) maintained comparable concentration (log10 10 ± 0.5 CFU per day) of the probiotic throughout the shelf life. This was achieved by enumerating BAL in all products with exactly the same procedure.

Table 2.4. BB-12 concentration in capsules as enumerated by MRS and MRS-NNLP

Replicate Media CFU/capsule Log10 CFU/capsule MRS1 3.50 × 1010 10.5 1 MRS-NNLP2 1.50 × 1010 10.2 MRS 2.52 × 1010 10.4 2 MRS-NNLP 1.29 × 1010 10.1 MRS 2.52 × 1010 10.4 3 MRS-NNLP 1.33 × 1010 10.1 MRS 2.34 × 1010 10.4 4 MRS-NNLP 1.45 × 1010 10.2

1de Man, Rogosa, and Sharpe agar 2MRS agar supplemented with nalidixic acid (15 µg/ml), neomycin sulfate (100 µg/ml), lithium chloride (3 mg/ml), paromomycin sulfate (200 µg/ml), and cysteine hydrochloride (0.5 mg/ml). Data shown are mean of three replicates.

71

Table 2.5. Selectivity test of mMRS and MRS-NNLP

mMRS1 MRS-NNLP2

YF-L702 + - YF-L702 + BB-12 + + BB-12 + +

1Modified MRS agar with 0.00005% (w/v) dicloxacillin, 0.01% (w/v) lithium chloride, and 0.005% (w/v) cysteine hydrochloride 2MRS agar supplemented with nalidixic acid (15 µg/ml), neomycin sulfate (100 µg/ml), lithium chloride (3 mg/ml), paromomycin sulfate (200 µg/ml), and cysteine hydrochloride (0.5 mg/ml). “+” Colonies, “-” No colonies

2.3.3. BB-12 remained stable throughout shelf life

The viable count of BB-12 in the yogurt smoothies and the pH of all the products were determined following manufacture and weekly throughout the shelf life. BB-12 capsules were evaluated periodically during their 18-month shelf life (Table 2.6). BB-12 in both yogurt and capsule was enumerated by pour plating appropriate dilutions on

MRS-NNLP agar. Although an earlier report based on data from 6 batches indicated no difference in the population of BB-12 between treatments PRE and POST was detected immediately following manufacture (week 0) (Ba et al. 2013), data from 27 batches revealed a statistical difference in the population of BB-12 between treatments PRE and

POST immediately following manufacture (week 0) (initial count of log10 10.55 ± 0.12

CFU/serving and log10 10.50 ± 0.14 CFU/serving, respectively). However, we believe this difference is not of practical significance.

As expected the population of BB-12 declined over the shelf life of the products.

The population was observed to decrease significantly faster in treatment POST than in treatment PRE. This trend continued and at the end of shelf life the BB-12 concentration

72

decreased significantly in both treatment PRE (log10 10.24 ± 0.13 CFU/serving) and treatment POST (log10 9.54 ± 0.25 CFU/serving) after 4 weeks’ storage (Table 2.7). The

BB-12 survived significantly better in PRE than that in POST (P < 0.001) indicating BB-

12 survives better when added before fermentation, possibly as a result of adaptation to the acidic environment.

Overall, all BB-12 interventions remained at specified effective dose level (log10

10 ± 0.5 CFU/serving) during shelf life. The PRE treatment had a lower pH (4.30 ± 0.03) than YS (4.32 ± 0.04) and POST (4.33 ± 0.04). Likely the cooling step for PRE was not as efficient as the other treatments, because it was fermented separately. Again, this difference is not of significance in terms of product development. The pH of all yogurt smoothies remained stable during shelf life (Table 2.8). To ensure participants were getting comparable amount of BB-12 every day during intervention phase, fresh yogurt smoothies were provided once every three weeks.

Table 2.6. BB-12 population in capsules during shelf life

Date 03/13/12 07/12/12 03/17/13 05/19/13 10/06/13 BB-12 Conc. (log 10 10.4 ± 0.0ab 10.4 ± 0.0a 10.3 ± 0.1ab 10.1 ± 0.2bc 10.0 ± 0.0c CFU/cap)

Data shown are mean ± SD from 4 replicates. Values in a row without a common letter are significantly different, P < 0.05.

73

Table 2.7. BB-12 population in yogurt smoothies during shelf life

BB-12 concentration (Log10 CFU/serving) Treatment Week 0 Week 1 Week 2 Week 3 Week 4 PRE 10.55 ± 0.12Aa 10.43 ± 0.13Ba 10.42 ± 0.13Ba 10.34 ± 0.13Ca 10.24 ± 0.13Da POST 10.50 ± 0.14Ab 10.16 ± 0.15Bb 10.03 ± 0.20Cb 9.77 ± 0.20Db 9.54 ± 0.25Eb

Data shown are mean ± SD from 27 batches. Values in a column without a common lowercase letter are significantly different, P < 0.05. Values in a row without a common UPPERCASE letter are significantly different, P < 0.05. PRE yogurt smoothie with BB-12 added before fermentation, POST, yogurt smoothie with BB-12 added after fermentation.

Table 2.8. pH value of the yogurt smoothies throughout shelf life

Treatment Week 0 Week 1 Week 2 Week 3 Week 4 YS 4.32 ± 0.04Ba 4.32 ± 0.04Bab 4.33 ± 0.03Aab 4.31 ± 0.04Aab 4.31 ± 0.03Ab PRE 4.30 ± 0.03Ca 4.31 ± 0.03Ca 4.29 ± 0.04Ba 4.29 ± 0.03Ba 4.29 ± 0.03Ba POST 4.33 ± 0.04Aa 4.34 ± 0.04Aa 4.33 ± 0.03Aa 4.32 ± 0.03Aa 4.32 ± 0.03Aa

Data shown are mean ± SD from 27 batches. Values in a row without a common lowercase letter are significantly different, P < 0.05. Values in a column without a common UPPERCASE letter are significantly different, P < 0.05. PRE yogurt smoothie with BB-12 added before fermentation, POST, yogurt smoothie with BB-12 added after fermentation.

2.3.4. S. thermophilus remained stable

As part of quality assessment, ST population in the yogurt smoothies was also determined following manufacture and weekly throughout the shelf life of the products

(Table 2.9). The three yogurt smoothies contained comparable levels of ST following production (YS, log10 9.01 ± 0.15 CFU/g; PRE, log10 8.97 ± 0.14 CFU/g; POST, log10

9.05 ± 0.13 CFU/g), although there was a statistically significant difference between PRE and the other two drinks. This may be due to the weak inhibitory effect of BB-12 against

ST reported in an early study (Vinderola et al. 2002). The ST population declined steadily during shelf life in all yogurt smoothies but remained within 0.2 log10 CFU, which was considered stable. This is in agreement with previous studies indicating ST and LB strains

74 survive well during cold storage at lowered pH (Dave and Shah 1997, Martensson et al.

2002, Saccaro et al. 2009, Rosburg et al. 2010).

Table 2.9. S. thermophilus population in yogurt smoothies during shelf life

ST concentration (Log10 CFU/g) Treatment Week 0 Week 1 Week 2 Week 3 Week 4 YS 9.01 ± 0.15Aa 9.02 ± 0.17Aa 9.02 ± 0.11Aa 8.94 ± 0.16ABb 8.88 ± 0.16Ac PRE 8.97 ± 0.14Ba 8.92 ± 0.13Ba 8.94 ± 0.13Ba 8.91 ± 0.13Bb 8.91 ± 0.15Bc POST 9.05 ± 0.13Aa 9.04 ± 0.12Ab 9.02 ± 0.14Aab 8.96 ± 0.15Ab 8.83 ± 0.13Ac

Data shown are mean ± SD from 27 batches. Values in a row without a common lowercase letter are significantly different, P < 0.05. Values in a column without a common UPPERCASE letter are significantly different, P < 0.05. PRE yogurt smoothie with BB-12 added before fermentation, POST, yogurt smoothie with BB-12 added after fermentation.

2.3.5. Yogurt smoothies had comparable composition

Total solids and fat content in the yogurt smoothies were measured following manufacture of each batch. Results are shown in Table 2.10. Overall, the yogurt smoothies had a total solids of 21.15% ± 0.51% and fat content of 0.69% ± 0.07%.

Although pre-added treatment (PRE) was statistically different from the other two yogurt smoothies in total solids and the POST drink had significantly higher fat content than YS and PRE, we believe these differences are not of practical significance. It is unlikely that these differences would influence the outcomes being measured in the following clinical trial.

75

Table 2.10. Composition of the yogurt smoothies

Component YS PRE POST Total solids 20.92 ± 0.36%b 21.58 ± 0.53%a 20.96 ± 0.32%b Fat 0.68 ± 0.07%b 0.68 ± 0.07%b 0.71 ± 0.07%a

Values shown in bold are mean ± SD. Data was collected from 30 batches, 3 samples per treatment per batch. Values in a row without a common letter are significantly different, P < 0.05. YS yogurt smoothie without BB-12, PRE yogurt smoothie with BB-12 added before fermentation, POST, yogurt smoothie with BB-12 added after fermentation.

2.4. Conclusions

The study identified MRS-NNLP agar as a suitable selective medium for BAL enumeration in yogurt smoothies that contain more ST than BAL. BB-12 population declined in all three probiotic interventions throughout shelf life and decreased significantly faster in the POST added yogurt smoothie than in the PRE added yogurt smoothie. In both products, BB-12 remained at the specified dose level (log10 10 ± 0.5

CFU per serving or capsule) at the end of shelf life. The three yogurt smoothies had comparable composition in terms of total solids and fat content.

76

680 bp

Figure 2.S1. An example of PCR verification of BAL. Lane 1 and 6: DNA Ladder, the Bflac2/5 primers yield an amplicon of 680 bp for B. animalis subsp. lactis; Lane 2-3, 7-9: yogurt smoothie samples; Lane 4: Positive control, B. animalis subsp. lactis DSMZ 10140; Lane 5: Negative control, contains amplification mixture without any sample.

77

Chapter 3 - The Effect of Bifidobacterium animalis subsp. lactis BB-12 and Different Delivery Vehicles on Gut Transit Time and Fecal Short Chain Fatty Acid Production in Healthy Adults

3.1. Introduction

Gut transit time (GTT), also known as bowel transit time or fecal transit time, refers to how long it takes for food to move from the mouth to the anus. Maintaining a regular bowel transit time is essential for health and general well being (Raahave 2015).

Delayed gut transit time increases the risk of toxicity, anal fissures, hemorrhoids, fatigue, constipation, bloating, gas, diverticulitis, and weight gain (Lewis and Heaton 1999). On the other hand, short bowel transit times may be associated with diarrhea, colitis, and some types of inflammatory bowel syndromes (IBS). A healthy gut transit time is approximately 12-14 hours as measured by the food dye method (Medline Plus Medical

Encyclopedia 2014). It is estimated that 4-56 million adult population in the U.S. are suffering from chronic constipation (Sanchez and Bercik 2011). Gut transit time varies markedly between individuals, male and female, even within individuals. It can be measured in a number of ways including radiopaque markers (Hinton et al. 1969), scintigraphy (Krevsky et al. 1986), and using wireless motility capsule (Cassilly et al.

2008). Each method has pro and cons in terms of cost, invasiveness, accuracy, and ease of use. The lack of standardization in methods makes it challenging to assess the effect of various interventions, including probiotics on gut transit time.

Several studies have examined the effect of probiotics on bowel transit time and gastrointestinal (GI) symptoms (Shanahan 2010, Williams et al. 2010, Waller et al.

78

2011). Bouvier et al. (Bouvier et al. 2001) demonstrated ingestion of a fermented milk containing the probiotic strain Bifidobacterium animalis subsp. lactis (BAL) DN-173 010 for 11 days significantly reduced total colonic transit time (CTT) by 21% and sigmoid transit time by 39%, compared to an identical fermented milk in which bacteria were killed by heat treatment. Waller et al. (Waller et al. 2011) observed consumption of probiotic strain BAL HN019 decreased participants’ whole gut transit time (WGTT) in a dose-dependent manner (high dose = 33% reduction, low dose = 25% reduction), and reduced the frequency of functional GI symptoms in adults such as abdominal pain, nausea, constipation, irregular bowel movements, and flatulence (P < 0.05).

Bifidobacterium animalis subsp. lactis BB-12 is a well-characterized probiotic strain that has been used as a food ingredient and food supplement worldwide for over 30 years. Larsen et al. (Larsen et al. 2006) studied the influence of BB-12 on gut transit time as determined by daily frequency of defecation and showed a significant linear increase in fecal consistency (P = 0.018) and a tendency to an increase in the frequency of defecation (shorter gut transit time) with increasing dosage of probiotic consumption.

Since a mixture of probiotics (Bifidobacterium animalis subsp. lactis BB-12 and

Lactobacillus paracasei subsp. paracasei CRL-431) was used in this study, the moderate effect observed could not be attributed solely to BB-12. Similar issue exists in other studies (Ringel-Kulka et al. 2008, Ringel-Kulka et al. 2015). While precise mechanisms of probiotic effect on intestinal motility remain unclear, one of the potential explanations is the increase in fecal bacterial mass, especially lactic acid-producing bacteria that may lower colonic pH and produce short chain fatty acids (SCFAs), which can stimulate

79 peristalsis (Waller et al. 2011). SCFAs are the main anions resulting from bacterial fermentation of carbohydrates and protein in the colon (Wong et al. 2006). SCFAs are important compounds for maintaining a healthy GI environment as they promote growth and differentiation of epithelial cells, and provide energy to other tissue/organs of the host such as muscle, kidney, heart and brain (Macfarlane and Macfarlane 2012). A number of factors are known to influence production of SCFAs, including gut microbiota, gut transit time, diet, and host physiological status (Roy et al. 2006, Rahat-Rozenbloom et al. 2014). Probiotics have shown promising effects on fecal SCFAs in vitro, in mouse models, and in human studies (Martin et al. 2008, Hong et al. 2010, Riezzo et al. 2012,

Vitali et al. 2012), although the effects are modest in healthy subjects (Klein et al. 2008).

Bifidobacterium animalis subsp. lactis, alone or in combination with other probiotics or prebiotics, has shown varied effect on fecal SCFA production (Laake et al.

1999, Matsumoto et al. 2007, West et al. 2012). It is commonly understood that SCFAs present in the colon and the bloodstream result primarily from fermentation of carbohydrates and protein by the gut microbiota (Roy et al. 2006, Wong et al. 2006,

Macfarlane and Macfarlane 2012, Wong et al. 2012, Pluznick 2014). Short chain fatty acid production has long been associated with changes in gut microbiota and gut transit time (Wichmann et al. 2013) as result of probiotic, prebiotic, or synbiotic products, but the findings are not conclusive or consistent. To better understand this complicated relationship, it is important to evaluate fecal SCFAs simultaneously with the gut transit and with composition of the gut microbiota.

80

One probiotic bacterial strain, B. animalis subsp. lactis BB-12, continues to gain popularity in a wide range of products (dairy and non-dairy). Thus, it is important to define the functional identity of this probiotic strain by itself and to understand how the delivery matrix may influence the outcomes being assessed. As part of a larger project, the primary goal of this study was to evaluate the effect of BB-12 (as delivered via yogurt smoothies or capsules) on the gut transit time and fecal SCFA production in healthy adults. Additionally, the impact of the timing of probiotic addition to the finished product

(pre- and post-fermentation) on their performance was assessed.

3.2. Materials and Methods

3.2.1. Participants

The sample size was calculated based on data from a previous study (Bouvier et al. 2001) and indicated that 28 subjects were required to identify a mean difference of 6 hours in fecal transit time. Considering the event of attrition from the study or lack of compliance, 36 healthy volunteers (18-40 years of age) with delayed gut transit (≥ 24 hours) were recruited for this study. Detailed exclusion criteria were described elsewhere

(Meng et al. 2015). Briefly, all subjects were nonsmoking, normotensive and not diagnosed with any chronic medical conditions. A written informed consent was obtained from each participant. The study was approved by the Institutional Review Board of the

Pennsylvania State University (University Park, PA). The ClinicalTrials.gov identifier is

NCT01399996.

81

3.2.2. Recruitment and screening

This clinical trial was a multidisciplinary project that involved both the

Departments of Food Science and Nutritional Sciences. Collaborators from the

Nutritional Sciences Department at Penn State were responsible for recruiting participants and for the clinical aspects of the project. Food Science developed the study interventions, measured participants’ fecal SCFAs, assisted with gut transit time measurements, and surveyed the gut microbiota (to be discussed in next chapter). The recruiting details are highlighted in Figure 3.S1. A total of 203 individuals expressed interest in participating in the study via phone or email after seeing the advertisements in local newspapers or university email lists. During the initial contact they were informed about the study and then if still interested, a telephone screening was conducted using a series of medical and lifestyle questions. Upon completion of the phone interview, 67%

(136) of the screened subjects qualified for a clinic visit. During the clinic visit at the

Penn State Clinical Research Center (CRC), potential participants were given an informed consent document detailing the requirements for their involvement in the study.

A staff member was available to answer any questions and then both the participant and the staff member had to sign and date the document. Participants then had their height, weight, waist circumference and blood pressure measured, followed by a fasting blood draw for a complete blood count and health profile (liver and kidney function, and glucose metabolism). The body mass index (BMI) of each individual was calculated from the measured weight and height. Participants also completed the Rome III (Drossman and

Dumitrascu 2006) diagnostic questionnaire to assess bowel function. After completing the screening procedures, 36 eligible volunteers were recruited for the study. To eliminate

82 any possible confounding effect of treatment order on outcomes, eligible participants were randomized to treatment sequences. An online tool was used to generate the randomization schemes (http://www.randomization.com). During the study two participants dropped out at the baseline visit due to schedule conflicts; four withdrew due to personal reasons; and one individual was excluded prior to the first treatment period due to follow-up failure. Therefore, 29 participants finished at least one treatment period.

Of the 29 participants, two were excluded due to pregnancy prior to the third and fourth treatment periods, respectively; one withdrew from the study prior to the fourth treatment period due to diagnosis of irritable bowel syndrome; one dropped out prior to the second treatment period due to schedule conflict; and two left the study due to other personal reasons prior to the second and the third treatment period, respectively. Thus, a total of 23 participants completed all treatments.

3.2.3. Design and intervention

This project was a randomized, partially blinded, 4-period crossover free-living study. The study design scheme is shown in Figure 3.1. The actual randomization schemes are shown in Appendix II. Measurements taken during the baseline visit were anthropometric assessment (age, gender, BMI, waist circumference), biochemical parameters (fasting serum glucose, insulin, and hs-CRP), a physical activity questionnaire was completed, and an immune end point assessment for use as baseline values. Then each participant began the intervention phase as specified by the randomization order.

The four treatments as described in chapter 2 were: (A) yogurt smoothie (YS, without

BB-12); (B) yogurt smoothie with BB-12 added after fermentation (POST); (C) yogurt

83 smoothie with BB-12 added before fermentation (PRE); and (D) one BB-12 containing capsule (CAP). Each treatment period lasted 4 weeks, with a two-week washout period scheduled between treatment periods.

During the yogurt smoothie treatment phases, participants consumed one 8-oz

(240 g) serving of yogurt per day. Each probiotic-containing smoothie delivered log10 10

± 0.5 colony-forming unit (CFU) of BB-12 per day. The yogurt smoothies were developed at the Pennsylvania State University (Merenstein et al. 2011) and modified for

BB-12 concentration and timing of addition for this project. When the treatment phase involv capsule consumption, participants ingested one capsule per day, which was

specifically produced by Chr. Hansen (Milwaukee, WI) to deliver log10 10 ± 0.5 CFU of

BB-12/capsule. Participants were instructed to avoid consuming any other food or supplements that may contain probiotic bacteria such as commercial yogurt, smoothies, and probiotic capsules or tablets during each 4-week intervention phase. They were also asked not to change their habitual diets during the course of the study and to maintain their body weight.

84

Figure 3.1. Schematic diagram for randomization design. Each period lasted for 4 weeks and there was a 2-week compliance break between treatments. The entire study lasted for 6.5 months for each volunteer. Stool and blood samples were collected before the first treatment and after each treatment. Only a stool sample was taken after the final washout.

3.2.4. Diet and physical activity assessment

To evaluate factors such as diet and physical activity that may contribute to the outcomes, participants were asked to provide a 24-hour dietary recall for 3 continuous days to assess their dietary intake and the International Physical Activity Questionnaires

(IPAQ) questionnaire was employed to evaluate their physical activity level as previously reported (Bouchard et al. 1983). Briefly, participants were asked to recall their food and beverages intake during meals and snacks over three continuous days according to detailed instructions including portion size of each food item provided by trained staff.

Daily intake of total calories, macronutrients, vitamins, minerals, caffeine, and alcohol was assessed based on the recorded food intake using the Food Processor SQL software

85

(ESHA Research, Salem, OR). Participants also recorded the activity they performed during each of the 96 periods of 15 min over a 24-hour period for 3 days, including one weekend day. An intensity scale, metabolic equivalent of task (MET), was used to categorize the recorded activities as previously described (Bouchard et al. 1983).

3.2.5. Gut transit time measurement

During the baseline visit, participant’s whole gut transit time (WGTT) was determined using two methods: blue food dye diluted in 6 oz. water as modified from the method by Lu et al. (Lu et al.

2009) and the SmartPill® (Given

Imaging, Duluth, GA) as introduced by Cassilly et al.

(Cassilly et al. 2008). According to the manufacture’s standard operating procedure (SOP), each participant was asked to fast for 12 Figure 3.2. Wireless motility system. (SmartPill®) hours prior to visiting the CRC for the SmartPill® approach. During the visit a standardized meal (SmartBar®) was consumed with 2-4 oz. water right before ingesting the SmartPill®. A wireless data receiver (Figure 3.2) was worn by the participant throughout the course of measurement and returned to the CRC once the pill was excreted during bowel movement (normally 3-5 days after ingestion). Unlike the blue dye method which only measures WGTT, the SmartPill® can measure regional transit time such as gastric emptying time (GET), small bowel transit time (SBTT) and colonic transit

86 time (CTT) along with temperature and pressure throughout the GI tract. Throughout the study the SmartPill® was used to measure gut transit time at the end of each intervention period. In addition, participants’ perception of constipation was assessed using the

ROME III questionnaire at the end of each intervention period.

3.2.6. Fecal SCFAs measurement

Stool samples were collected at baseline, after each treatment, and after the final washout as indicated by the black bars in Figure 3.1. The procedure for stool collection is highlighted in Appendix III. After arrived at the CRC samples were stored at -80°C prior to being aliquoted and extracted. SCFAs were extracted using a method modified from a previous study (Garcia-Villalba et al. 2012). The extraction procedure is shown in

Figure 3.3. Briefly, 0.1 g of thawed (on ice) stool sample was weighed and suspended in

1 ml of 0.5% phosphoric acid in a 2 ml screw cap tube. The tube was then homogenized at 6,500 rpm for 90 s using Precellys-24 homogenizer (Bertin Corp., Rockville, MD) followed by centrifugation at 17,000 × g for 10 min in a Thermo Fisher Sorvall ST 16R centrifuge (Langenselbold, Germany). Next, 800 µl of the supernatant was transferred to a new tube followed by addition of 800 µl of internal standard (IS, 1mM heptanoic acid) containing ethyl acetate. The sample was then vortexed for 2 min to ensure sufficient extraction and then centrifuged at 17,000 × g for 10 min. Finally, 600 µl of the organic phase was transferred into a gas chromatography (GC) vial and stored at -20°C until analysis. Fatty acid standards (Table 3.1) were purchased from Sigma-Aldrich (St. Louis,

MO, USA). Ethyl acetate, and phosphoric acid were purchased from Merck (Darmstadt,

Germany).

87

Following extraction, samples were analyzed by gas chromatography. The GC system consisted of an Agilent 6890 (Agilent Technologies, Palo Alto, CA), equipped with an automatic sampler (MPS) (Gerstel, Mülheim, Germany) and a flame ionization detector (FID). A high polarity, polyethylene glycol (PFG), fused silica capillary column

DB-WAXETR (30 m, 0.25 mm id, 0.25 mm film thickness) (Agilent) was used for separation. The GC conditions employed were based on a previous study (Garcia-Villalba et al. 2012) with minor modifications. Prior to sample analysis, a standard solution containing a mixture of standards (30 mM final concentration of each acid) in ethyl acetate was diluted to obtain a calibration curve ranging from 3 to 3,000 µM. The standard curves were constructed by plotting the concentration of each individual SCFA versus the ratio of SCFA peak area/IS peak area (Table 3.1). Each point of the standard curves corresponds to the mean value from three independent replicate injections. Three independent replicate extractions were performed per sample.

88

Figure 3.3. Diagram showing the extraction of SCFAs.

Table 3.1. Calibration curves of the SCFAs standards

Compound Calibration curves R2 Acetic acid ! = 6222.2! − 316.53 0.9995 Propionic acid ! = 2989.0! − 32.352 0.9994 Isobutyric acis ! = 2226.9! − 35.407 0.9996 Butyric acid ! = 1955.4! − 29.708 0.9996 Isovaleric acid ! = 1390.6! − 28.810 0.9998 Valeric acid ! = 1367.0! − 27.513 0.9998 Caproic acid ! = 1067.4! − 25.573 0.9998

89

3.2.7. Statistical analysis

Intention-to-treat (ITT) analyses were applied to treatment effect analyses (Fisher et al. 1990). All data were first tested for normality using the Anderson-Darling test. Log transformation was applied when necessary. The analysis of treatment effect and period effect was performed using analysis of variance (ANOVA) or the Kruskal-Wallis test where appropriate. Mann-Whitney U test was used for gender comparison and the

Spearman rho test was employed for correlation analyses, a P value < 0.05 was considered significant. All analyses were performed using Minitab 17.0 software

(Minitab Inc., State College, PA).

3.3. Results and Discussion

To our knowledge, the present study is the first clinical trial to evaluate the effect of delivery matrixes on the efficacy of the commercial probiotic strain Bifidobacterium animalis subsp. lactis BB-12 in healthy adults by assessing their regional gut transit time and fecal short chain fatty acid production. Results of this study indicated that there was no significant treatment effect on either gut transit time or fecal SCFA production due to the large variations observed among and within individuals. However, the present study did confirm that regional gut transit times of the study cohort significantly correlate with fecal SCFA concentrations.

3.3.1. Participant characteristics

Baseline characteristics of the participants (anthropometric indexes, blood pressure, biochemical features, and physical activity) are shown in Table 3.2. Twenty-

90 eight participants (17 females and 11 males) were included in the analyses, as they completed at least one of the four intervention periods. Overall, the participants were healthy, young adults (mean age of 28.2 ± 0.6 years). The average BMI was 24.1 ± 0.2 kg/m2; 16 (57.1%) participants were normal weight, 11 (39.3%) were overweight, and 1

(3.6%) was obese. Their blood pressure was normal and waist circumference, fasting blood glucose, insulin, and CRP levels were within the normal range (Table 3.2).

Physical activity, as assessed from self-reported IPAQ responses, indicated a median daily physical activity intensity of 3.0 METs (range 2.3 – 4.3 METs). The average daily total calorie intake of participants calculated from 3-day dietary recall records was estimated to be 2248 ± 83 kcal. The daily intake of macronutrients, vitamins, minerals and n-3 PUFA, caffeine and alcohol are reported in Table 3.2.

91

Table 3.2. Demographic characteristics of participants at baseline1*

Characteristics Values (n = 28) Age (yr) 28.2 ± 0.6 Male, n (%) 11 (39.3%) Body mass index (kg/m2) 24.1 ± 0.2 ≤24.9 16 (57.1%) 25.0 -29.9 11 (39.3%) ≥30 1 (3.6%) Waist Circumference (cm) 85.2 ± 0.6 Blood pressure (mm Hg) Systolic 107.7 ± 0.8 Diastolic 72.8 ± 0.6 Glucose (mg/dL) 86.6 ± 0.8 Insulin (mg/dL) 5.4 ± 0.4 hs-CRP (mg/L) 2.0 ± 0.5 Physical activity (METs)2 3.0 (2.3 – 4.3) Dietary intake2 of Total calories (kcal/d) 2248 ± 83 Carbohydrate (g/d) 285.3 ± 10.9 Protein (g/d) 90.7 ± 3.6 Fat (g/d) 84.0 ± 3.4 Vitamin C (mg/d) 68.0 ± 4.8 Vitamin D (IU/d) 99.2 ± 11.2 Vitamin E (mg/d) 3.4 ± 0.3 Iron (mg/d) 14.5 ± 0.8 Selenium (µg/d) 46.3 ± 4.5 Zinc (mg/d) 5.8 ± 0.4 n-3 PUFA (g/d) 0.6 ± 0.1 Caffeine (mg/d) 75.9 ± 10.3 Alcohol consumption (g/d) 2.6 ± 0.9

1 Values are presented as mean ± SEM or n (%) or median (range). 2 Physical activity and dietary intake were assessed from self-reported responses to IPAQ and 3-day dietary recall records, respectively. * Shared results with collaborators in the project.

92

3.3.2. Blue dye results correlate with SmartPill® measurements

In an effort to understand how well the measurements of the blue dye method and the SmartPill® agree to each other, the whole gut transit time of all participants’ at baseline was measured using both the “food dye” method and a wireless motility capsule

(SmartPill®). Since this was a baseline measurement, the data shown in Figure 3.4 are from all the recruited participants who successfully recorded their blue dye results (n =

28). Results of the two methods generally agreed. However, a few participants (7/28) reported shorter whole gut transit time as measured by blue dye than by SmartPill®.

Statistical analysis revealed a significant correlation between the blue dye and SmartPill® measurements (Spearman rho = 0.67, P < 0.0001). Insoluble and non-absorbable food dyes have been used as a simple measurement of gut transit time for many years (Higgs et al. 1975). Concerns were raised for using food dye in bowel transit time measurement for a number of reasons. First, it is challenging to define the end point and the marker travels with the liquid phase of the gut content in patients with acute gastroenteritis

(Higgs et al. 1975), the blue dye is not always visible (Slavin et al. 1981), and it is difficult to assess the highly variable time required for the dye to disappear (Fleming et al. 1983). Thus, it has been suggested a combination of dye and radiopaque pellets be used to measure the liquid phase and solid phase of gut content, respectively (Fleming et al. 1983). Results from the present study suggest that the blue dye method remains a useful screening tool for gut transit time in healthy individuals. However, because the

SmartPill® does not rely on judgment, the results for WGTT are likely more representative of actual transit time. Alternatively, difference in blue dye and SmartPill® results could indicate the rheological properties of the intestinal contents.

93

SmartPill Blue Dye

140

120

100

80

60

(h) gut Whole transit time 40

20

0

Participant's ID

Figure 3.4. Baseline whole gut transit time as measured by blue dye and SmartPill®.

94

3.3.3. Assessment of gut transit time

Whole gut transit time together with regional gut transit times (GET, SBTT, CTT) of the participants’ were measured using the SmartPill® (Appendix IV). Only 27 participants were included in this analysis, because a few data points could not be retrieved from the SmartPill® data receiver due to technical difficulties. Results of the analysis are presented in Table 3.3. No treatment effect was detected. In the present study, participants had a wide range of WGTT (7.13 h – 128.58 h), CTT (0.5 h – 122.25 h), SBTT (1 h – 19.02 h), and GET (0.98 h – 18.83 h). Subjects with extremely short

CTT (0.5 h) might have had diarrhea. Individuals had varying responses to different treatments (Figure 3.S2), which made it difficult to detect treatment effect if there was any. In agreement with a previous study (Degen and Phillips 1996), males in the present cohort had shorter CTT (P = 0.0098), WGTT (P = 0.0036), and GET (P < 0.0001) than females, but exhibited no difference in SBTT (P = 0.3).

In previous work, the effect of a BB-12 intervention (delivered in fermented milk, capsule, or fermented cereal) on host bowel movement was studied in both healthy subjects and subjects with functional bowel symptoms and promising improvements were observed (Uchida et al. 2005, Larsen et al. 2006, Pitkala et al. 2007, Ringel-Kulka et al.

2015). A recent large clinical trial with 1,248 subjects performed in eight centers across

Europe reported a treatment effect of BB-12 (delivered in capsule) on average defecation frequency (P = 0.0065) (Eskesen et al. 2015), despite the fact that the placebo group also had increased average defecation frequency compared to baseline. In the present study, we did not see a treatment effect on gut transit time. A few factors may explain the

95 discrepancy between this study and the studies showing a change. First, the present study had a relatively small sample size, which makes it difficult to detect a treatment effect from a highly variable data set if the signal is weak. Secondly, the present cohort is a generally healthy group of individuals, with whom there is only so much room for improvement in terms of bowel transit time. Finally, this study took direct measurements of the gut transit times using a wireless motility capsule in contrast to other studies that employed more subjective defecation frequency questionnaires.

While the effectiveness of BB-12 to accelerate bowel transit in certain individuals with delayed bowel transit time remains in question, the findings of this study is consistent with a previous Canadian study (Tulk et al. 2013). A total of 65 healthy participants (39 females, 26 males) were given either synbiotic (Bifidobacterium lactis

BB-12 (≥107 CFU/g), Lactobacillus acidophilus La5 (≥107 CFU/g), L. casei CRL431

(≥107 CFU/g) and 2 g inulin per 100 g yogurt) containing yogurt (200 g/day) or control yogurt for 15 days. Gut transit time was measured before and after interventions using the food dye method. No significant difference was observed either between synbiotic (30.6

± 18.3 h) and control group (31.1 ± 15.4 h) or between pre- (31.0 ± 16.6 h) and post-

(30.6 ± 18.3 h) the synbiotic treatment.

96

Table 3.3. Regional gut transit times (hours) of the participants1

WGTT CTT SBTT GET YS (A) 36.08 ± 24.95 27.76 ± 24.90 5.40 ± 3.40 2.91 ± 0.94 POST (B) 35.02 ± 17.12 27.51 ± 16.26 4.42 ± 1.81 3.07 ± 1.32 BL 40.78 ± 24.50 33.21 ± 24.25 4.69 ± 1.38 2.86 ± 0.89 PRE (C) 38.51 ± 26.63 30.80 ± 26.21 4.54 ± 1.65 3.04 ± 1.13 CAP (D) 35.54 ± 14.74 27.71 ± 14.07 4.40 ± 1.46 3.41 ± 1.27

Abbreviations: WGTT whole gut transit time, CTT colonic transit time, SBTT small bowel transit time, GET gastric emptying time, YS yogurt smoothie w/o BB-12, POST yogurt smoothie with BB-12 added post fermentation, BL baseline, PRE yogurt smoothie with BB-12 added before fermentation, CAP BB-12 containing capsule. 1 Mean ± SD.

3.3.4. Fecal SCFA concentrations

Short chain fatty acids are important compounds for maintaining a healthy GI environment as they promote the growth and differentiation of epithelial cells and provide energy to colonocytes (Roy et al. 2006). Fecal SCFA concentrations are used as indicators of SCFA production in the colon due to non-invasiveness. Fecal SCFA concentrations of the participants’ were measured at baseline, after each treatment, and after final washout (Table 3.4). In agreement with previous studies (Siigur et al. 1994,

McOrist et al. 2008, Garcia-Villalba et al. 2012), the most abundant SCFAs detected in human fecal samples were acetic acid (30.7%), propionic acid (21.3%), and butyric acid

(31.8%) and their concentrations were significantly correlated (Table 3.5). Overall, males had a higher butyric acid concentration than females in this study cohort (median 908

µg/g vs. 687 µg/g, P = 0.0233). This is in agreement with a previous study (McOrist et al.

2011), in which the authors observed a significantly higher level of butyrate excretion in males than in females (P = 0.04). High inter-individual variations were observed in the

97 content of fecal SCFAs (relative standard deviation 52%-118%). There was no difference in SCFAs level among treatments except for acetic acid, which also varied significantly from run to run. It is not clear why it was so difficult to obtain an accurate measurement of acetic acid in biological samples.

Since fecal SCFAs only account for less than 5% of the total SCFAs produced in the colon (Roy et al. 2006), the effectiveness of BB-12 in promoting SCFA production in the large intestine of healthy volunteers remains unclear. Strains of BAL have been reported to increase fecal SCFA production of individuals in disease states. An early study demonstrated consumption of yogurt containing log10 9.72 CFU of BAL LKM512 per day for 4 weeks tended to increase fecal butyrate concentration in patients with atopic dermatitis when compared to baseline (P = 0.0816) (Matsumoto et al. 2007). In another study on 16 patients with ileal pouch-anal anastomosis (IPAA) for ulcerative colitis, no difference in SCFAs content was observed after consumption of 500 ml of fermented milk (CulturaTM) containing > 108 CFU/ml of both L. acidophilus (La5) and BAL (BB-

12) for one week when compared to baseline nor between groups (Laake et al. 1999). To our knowledge, the effect of BB-12 alone or its delivery matrix on the fecal SCFA concentrations of healthy subjects has not been studied yet. The present study provides some information to this area of research.

98

Table 3.4. SCFAs concentration (µg/g) among treatments1

Treatment Baseline YS (A) POST (B) PRE (C) CAP (D) Final Washout Acetic acid 676.6 ± 402.3 943.3 ± 427.9 799.4 ± 408.8 894.3 ± 457.6 766.5 ± 333.8 993.9 ± 503.6 Propionic acid 517.6 ± 301.6 627.8 ± 319.4 538.9 ± 284.4 587.5 ± 326.2 548.3 ± 249.6 648.4 ± 335.6 Isobutyric acid 107.3 ± 82.9 82.8 ± 40.8 100.0 ± 55.9 90.6 ± 55.9 99.8 ± 59.6 106.7 ± 75.5 Butyric acid 751.4 ± 541.4 908.8 ± 456.2 802.5 ± 505.7 906.3 ± 626.3 823.8 ± 515.7 1083.8 ± 827.8 Isovaleric acid 186.5 ± 134.8 135.1 ± 67.2 170.2 ± 91.5 151.6 ± 99.2 168.3 ± 99.1 173.0 ± 128.2 Valeric acid 148.3 ± 164.5 128.0 ± 71.6 122.7 ± 82.0 133.4 ± 83.3 137.4 ± 101.6 148.5 ± 115.1 Caproic acid 38.7 ± 74.6 49.7 ± 103.9 38.5 ± 68.5 43.0 ± 71.5 68.7 ± 118.7 35.2 ± 55.1

Abbreviations: YS yogurt smoothie without BB-12, POST yogurt smoothie with BB-12 added post fermentation, PRE yogurt smoothie with BB-12 added before fermentation, CAP BB-12 containing capsule. 1 Mean ± SD.

99

3.3.5. Gut transit times correlate with predominant fecal SCFAs

To further explore possible relationships between gut transit times and fecal

SCFA concentrations, data were tested for correlations. Gut transit times negatively correlated with predominant SCFAs, but valeric acid positively correlated with small bowel transit time (Table 3.5). In Lewis and Heaton’s study (Lewis and Heaton 1997), raw wheat bran (28.3 g/day) and non-prescription drugs (Senna tablet to accelerate gut transit; Loperamide tablet to slow down transit) were given to the participants (13 healthy volunteers: 3 male and 10 female) in turn for 9 days each. There was a 2-4 weeks washout period in between treatments. The whole gut transit time and fecal SCFAs of the participants’ were measured before and at the end of each intervention period using radio-opaque marker method and GC, respectively. The authors reported negative correlations between WGTT and fecal total SCFA (r = -0.511, P = 0.001) and butyric acid (r = -0.577, P < 0.001).

In a more recent animal study (Kashyap et al. 2013), Kashyap et al. found that mice fed a diet of fermentable fructooligosaccharide (FOS) had shorter gut transit time as assessed by food dye and lower levels of SCFAs when compared with mice fed with standard control diet. The authors suspected the alterations of gut transit time and fecal

SCFA content were mediated by gut microbiota, because they observed an opposite effect in germ-free mice. However, it remains unclear whether it was the altered gut transit changed gut microbiota, which further resulted in different SCFA production, or it was the other way around. Thus, further studies are warranted to better understand the

100 role of specific pathways and metabolites in mediating the complex interactions between diet, gut microbiota, and gut transit.

In the present study, regional gut transit time and short chain fatty acid data were also screened for possible associations with self-reported dietary information and recorded demographic data (Table 3.S1). Although the recorded dietary data are not necessarily accurate due to some of the abnormally low (810 calories) or high (5,844 calories) calorie intake reported by the participants, there were some correlations detected between nutrients and gut transit times as highlighted in bold in Table 3.S1. Surprisingly, fiber intake had little to do with the gut transit time in this study cohort as fiber is generally considered a bulking agent that facilitates bowel movement. However, the fiber data assessed based on 24-hour dietary recall may not be good. On the other hand, a number of demographic characteristics correlated with gut transit time and fecal short chain fatty acid concentration. Notably, age was negatively correlated with the predominant fecal short chain fatty acids. Age was subsequently included in a general linear model as a covariate, but no treatment effect was detected either.

101 Table 3.5. Spearman correlations of gut transit time and fecal SCFAs1

Acetic Propionic Isobutyric Butyric Isovaleric Valeric WGTT CTT SBTT GET acid acid acid acid acid acid 0.97 CTT <0.001 0.26 0.12 SBTT 0.003 0.163 0.21 0.12 -0.06 GET 0.018 0.175 0.479 Acetic -0.19 -0.19 0.05 -0.29

acid 0.036 0.033 0.566 0.001 Propionic -0.25 -0.25 0.03 -0.27 0.68

acid 0.005 0.004 0.763 0.003 <0.001 Isobutyric 0.04 0.06 0.03 -0.16 0.11 0.29

acid 0.697 0.492 0.714 0.072 0.224 0.001 Butyric -0.21 -0.21 0.06 -0.36 0.86 0.73 0.23

acid 0.021 0.016 0.488 <0.001 <0.001 <0.001 0.008 Isovaleric 0.07 0.10 0.01 -0.1 -0.06 0.18 0.971 0.09

acid 0.407 0.285 0.925 0.250 0.533 0.038 <0.001 0.311 Valeric -0.1 -0.12 0.20 -0.12 0.49 0.54 0.64 0.51 0.55

acid 0.261 0.183 0.026 0.087 <0.001 <0.001 <0.001 <0.001 <0.001 Caproic 0.07 0.08 0.07 -0.01 0.24 -0.03 0.21 0.16 0.14 0.39 acid 0.417 0.350 0.407 0.897 0.006 0.759 0.018 0.076 0.107 <0.001 Cell Contents: Spearman rho P-Value 1 Numbers in red are correlations between regional gut transit times; numbers in blue are correlations between fecal short chain fatty acids; numbers in black are correlations between gut transit time and fecal short chain fatty acid; and numbers in bold are significant correlations.

102 3.4. Conclusions

The present study demonstrated the blue dye method for assessing the whole gut transit time of healthy adults was a reliable cost-efficient method for baseline screening as compared to a wireless motility capsule (SmartPill®). However, it also highlighted the difficulties of a self-reported end point. No significant treatment effect was observed on gut transit times or fecal short chain fatty acid production. However, this study is the first to demonstrate a possible relationship among regional gut transit times and fecal SCFAs in healthy adults; and the results from this study confirmed a number of previously reported correlations (Table 3.5). Notably, the predominant short chain fatty acids negatively correlated with whole gut transit time, colonic transit time, and gastric emptying time, but did not correlate with small bowel transit time. Studies with strictly selected cohorts (constipated) are needed to further explore the potential health benefits of the probiotics such as BB-12 on stool frequency in various populations.

103

Figure Legends

Figure 3.S1. Recruitment strategy used in the study. YS yogurt smoothie without BB-12, PRE yogurt smoothie with BB-12 added before fermentation, POST yogurt smoothie with BB-12 added after fermentation, CAP BB-12 containing capsule.

Figure 3.S2. Changes in regional gut transit time of the participants’ after each treatment from baseline. No clear pattern of response to treatment could be observed. YS yogurt smoothie without BB-12, PRE yogurt smoothie with BB-12 added BEFORE fermentation, POST yogurt smoothie with BB-12 added AFTER fermentation, CAP BB-12 containing capsule. (Figure 3.S2A, whole gut transit time; Figure 3.S2B, colonic transit time; Figure 3.S2C, small bowel transit time; Figure 3.S2D, gastric emptying time).

104

Figure 3.S1.

Enrollment

Assessed for eligibility (n=136)

Excluded (n=100) • Did not meet inclusion criteria (n= 41) • Declined to participate (n=0) Allocation • Other reasons (n=59)

Randomized (n=36)

Randomly allocated to receive interventions, including YS, PRE, POST, or CAP (n=36) • Received allocated intervention (n=29) • Did not receive allocated intervention (n=7) o Schedule conflict (n=2) o Lost to follow-up (n=1) o Other person reasons (n=4)

Crossover

Crossed over to receive other interventions, including YS, PRE, POST, or CAP (n=29) • Received allocated intervention (n= 29) • Did not receive allocated intervention (n=0)

Follow-Up

Lost to follow-up (n=0) Discontinued intervention after allocation (n=6) • Pregnancy (n=2) • Diagnosed of Irritable Bowel Syndrome (n=1) • Schedule conflict (n=1) • Other personal reasons (n=2)

Analysis

Analysed (n=28) Excluded from analysis (n=1)

105 Figure 3.S2A. WGTT

YS PRE POST CAP

100

50

0 2 3 4 6 7 8 9 12 13 14 15 16 17 19 20 21 22 24 25 26 28 29 30 32 33 34 35

Time (h) Time -50

-100

-150 Participant ID

106 Figure 3.S2B.

CTT

YS PRE POST CAP

100

50

0 2 3 4 6 7 8 9 12 13 14 15 16 17 19 20 21 22 24 25 26 28 29 30 32 33 34 35

Time (h) Time -50

-100

-150 Participant ID

107 Figure 3.S2C.

SBTT

YS PRE POST CAP

15

10

5

Time (h) Time 0 2 3 4 6 7 8 9 12 13 14 15 16 17 19 20 21 22 24 25 26 28 29 30 32 33 34 35

-5

-10 Participant ID

108 Figure 3.S2D.

GET

YS PRE POST CAP

5

4

3

2

1

Time (h) Time 0 2 3 4 6 7 8 9 12 13 14 15 16 17 19 20 21 22 24 25 26 28 29 30 32 33 34 35 -1

-2

-3 Participant ID

109 Table 3.S1. Spearman correlation between gut transit time, short chain fatty acid, dietary and demographic data1

Calories Protein Carbohydrates Fiber Fat Vitamin C

Protein 0.83

<0.0001

Carbohydrates 0.92 0.67

<0.0001 <0.0001

Fiber 0.68 0.66 0.67

<0.0001 <0.0001 <0.0001

Fat 0.88 0.76 0.67 0.57

<0.0001 <0.0001 <0.0001 <0.0001

Vitamin C 0.29 0.32 0.33 0.49 0.14

0.0011 0.0002 0.0001 <0.0001 0.1117

SLBTT -0.16 -0.22 -0.16 -0.134 -0.13 -0.176 0.0693 0.0158 0.0692 0.1369 0.1433 0.0628

WGTT -0.17 -0.23 -0.17 -0.13 -0.13 -0.16 0.0674 0.0106 0.0607 0.1632 0.1673 0.0844

CTT -0.12 -0.17 -0.12 -0.08 -0.09 -0.15 0.1748 0.0503 0.1719 0.3462 0.3086 0.0978

SBTT -0.03 0.02 -0.09 -0.01 0.06 -0.07 0.7546 0.7803 0.3353 0.9506 0.5246 0.4715

GET -0.19 -0.27 -0.19 -0.10 -0.11 0.06 0.0334 0.0024 0.0343 0.2775 0.1983 0.4928

Acetic acid 0.06 0.14 0.11 0.15 -0.06 0.13 0.5171 0.1223 0.2241 0.0967 0.5089 0.1408 Propionic acid -0.07 0.05 -0.03 -0.10 -0.13 -0.01 0.4646 0.5739 0.7307 0.2820 0.1321 0.9437 Butyric acid 0.08 0.19 0.11 0.12 -0.01 0.07 0.3612 0.0300 0.2251 0.1805 0.9270 0.4639

110

Table 3.S1. Spearman correlation between gut transit time, short chain fatty acid, dietary and demographic data (Continued)

Vitamin E Iron Selenium Zinc Omega-3 Alcohol Caffeine

Iron 0.26 0.0043

Selenium 0.47 0.38 <0.0001 <0.0001

Zinc 0.34 0.64 0.62 0.0002 <0.0001 <0.0001

Omega-3 0.54 0.39 0.63 0.45 <0.0001 <0.0001 <0.0001 <0.0001

Alcohol 0.11 0.19 0.15 0.14 0.12 0.2371 0.0342 0.1036 0.1328 0.1748

Caffeine 0.15 0.12 -0.07 0.04 0.20 0.10 0.0990 0.1856 0.4623 0.6377 0.0267 0.2884

SLBTT -0.12 -0.29 -0.15 -0.17 -0.19 -0.154 0.11 0.1905 0.0012 0.1069 0.0624 0.0364 0.1060 0.2425

WGTT -0.12 -0.29 -0.16 -0.17 -0.19 -0.14 0.12 0.1931 0.0010 0.1003 0.0571 0.0397 0.1109 0.1763

CTT -0.09 -0.22 -0.11 -0.12 -0.14 -0.14 0.10 0.3600 0.0139 0.2300 0.1991 0.1178 0.1262 0.2496

SBTT 0.05 -0.15 0.07 0.01 -0.01 -0.19 -0.05 0.5624 0.1013 0.4559 0.9559 0.9095 0.0320 0.5629

GET 0.06 -0.11 -0.08 -0.13 -0.04 -0.03 0.14 0.5258 0.2347 0.4264 0.1657 0.6949 0.7179 0.1187

111 Table 3.S1. Spearman correlation between gut transit time, short chain fatty acid, dietary and demographic data (Continued)

Vitamin D Vitamin E Iron Selenium Zinc Omega-3 Alcohol Caffeine

Vitamin E 0.20

0.0325

Iron 0.451 0.26

<0.0001 0.0043

Selenium 0.33 0.47 0.38

0.0003 <0.0001 <0.0001

Zinc 0.33 0.34 0.64 0.62

0.0002 0.0002 <0.0001 <0.0001

Omega-3 0.35 0.54 0.39 0.63 0.45

0.0001 <0.0001 <0.0001 <0.0001 <0.0001

Alcohol 0.11 0.11 0.19 0.15 0.14 0.12

0.2471 0.2371 0.0342 0.1036 0.1328 0.1748

Caffeine 0.07 0.15 0.12 -0.07 0.04 0.20 0.10

0.4250 0.0990 0.1856 0.4623 0.6377 0.0267 0.2884

Acetic acid 0.08 0.06 0.09 0.21 0.11 -0.03 0.22 -0.12 0.4202 0.4903 0.3339 0.0225 0.2420 0.7437 0.0122 0.1762

Propionic acid -0.05 -0.05 0.09 0.15 0.12 -0.16 -0.02 -0.32 0.6109 0.5722 0.2958 0.0979 0.1778 0.0808 0.7930 0.0002

Butyric acid 0.11 0.05 0.11 0.26 0.19 -0.05 0.11 -0.15 0.2264 0.5944 0.2139 0.0041 0.0405 0.5771 0.2295 0.0830

112

Table 3.S1. Spearman correlation between gut transit time, short chain fatty acid, dietary and demographic data (Continued)

Cholesterol LDL TGs Age BMI PA

LDL 0.85

<0.0001

TGs 0.50 0.35

<0.0001 <0.0001

Age 0.29 0.23 0.12

0.0010 0.0081 0.1717

BMI -0.01 0.13 -0.07 0.22

0.8895 0.1312 0.4682 0.0115

PA -0.19 -0.24 -0.04 0.01 -0.40

0.0293 0.0074 0.6605 0.9096 <0.0001

CTT 0.03 0.03 0.02 -0.06 0.05 -0.12 0.7138 0.7497 0.8088 0.5074 0.5697 0.1846

SBTT -0.09 -0.04 -0.17 -0.02 0.01 -0.02 0.3169 0.6779 0.0490 0.8626 0.9734 0.8399

GET 0.29 0.38 -0.06 0.15 0.14 -0.03 0.0008 <0.0001 0.5135 0.0997 0.1096 0.7025

WGTT 0.04 0.04 -0.025 -0.05 0.04 -0.15 0.6841 0.6592 0.8674 0.5964 0.6662 0.0897

SLBTT 0.02 0.02 -0.03 -0.06 0.04 -0.16 0.8624 0.8404 0.7674 0.5242 0.6790 0.0784

Acetic acid -0.20 -0.21 -0.04 -0.20 -0.14 -0.07 0.0209 0.0204 0.6396 0.0222 0.1196 0.4428 Propionic acid -0.25 -0.22 -0.10 -0.36 0.08 -0.08 0.0052 0.0111 0.2831 <0.0001 0.3648 0.3872 Butyric acid -0.29 -0.24 -0.153 -0.28 -0.09 -0.02 0.0009 0.0059 0.0865 0.0011 0.3162 0.7875

Cell content: Spearmen rho P-value 1 Dietary data and demographic data were obtained from collaborators in the Department of Nutritional Sciences at Penn State University. Abbreviations: GET gastric empty time, SBTT small bowel, SLBTT small and large bowel transit time, CTT colonic transit time, WGTT whole gut transit time, LDL low-density lipoprotein, TGs triglycerides, BMI body mass index, PA physical activity. 113

Chapter 4 - The Effect of Bifidobacterium animalis ssp. lactis BB-12 and Different Delivery Vehicles on the Gut Microbiota of Healthy Adults

4.1. Introduction

The human gastrointestinal tract (GIT) harbors a diverse and dynamic community of microorganisms collectively termed the gut “microbiota” that contributes to the homeostasis of the gut (Tuohy et al. 2005). It is estimated the gut microbiota contains over 10 trillion microbes and hundreds of species of facultative and obligate anaerobes

(Ursell et al. 2012). The balance of the resident microbiota can be disturbed by medical interventions such as antibiotics, resulting in, among other effects, decreased short chain fatty acid (SCFA) metabolism with accumulation of luminal carbohydrate, subsequent pH changes, and water absorption (Clausen et al. 1991). Other factors such as age, genetics, environment, diet, and human health can also influence the composition of the gut microbiota. Disturbed gut microbiota, termed “dysbiosis” (Tamboli et al. 2004), has been linked to various diseases such as obesity (Bäckhed et al. 2004, Ley et al. 2005, Ley et al. 2006, Turnbaugh et al. 2006, Vijay-Kumar et al. 2010, Million et al. 2012), inflammatory bowel disease (IBD) (Cadwell et al. 2010), and antibiotic associated diarrhea (AAD) (Alonso and Guarner 2013). Probiotics may help restore the microbiota of a disrupted GIT, i.e. studies have showed that probiotic interventions significantly reduced the incidence of AAD in infants and children (Arvola et al. 1999, Corrêa et al.

2005, Engelbrektson et al. 2006).

114

Probiotics, defined as “live microorganisms that, when administered in adequate amounts, confer a health benefit on the host” (FAO/WHO 2002, Hill et al. 2014), are often ingested as supplements in powder, capsule or liquid forms, designed specifically for medicinal benefit. Such supplements have shown potential benefits in treatment and prevention of numerous diseases, including diarrhea, asthma, necrotizing enterocolitis and allergies (Vanderhoof et al. 1999, Kalliomäki et al. 2001, Rosenfeldt et al. 2002,

Mastrandrea et al. 2004, Lin et al. 2005). As an alternative to delivery in supplements, probiotic organisms can be included as ingredients in fermented dairy products to produce functional foods; that is foods providing health benefits beyond their nutritional value (Pelletier et al. 2002, Pelletier et al. 2003, Hasler et al. 2009). Yogurt, for example, is a fermented milk product often considered a functional food. According to a survey conducted by Monroe Mendelsohn Research in 2001, two-thirds of primary care physicians who counsel patients about nutrition recommend consuming yogurt containing live and active cultures for the health benefits associated with this food. However, it is not clear whether probiotics delivered via dairy and non-dairy based matrixes are equally effective or which matrix is better in terms of benefiting human gut microbiota. It is also uncertain how the fermentation process may affect the performance of probiotic organisms in vivo.

Probiotics marketed as nutritional supplements and found in functional foods, such as yogurts, are principally members of the genera Bifidobacterium and

Lactobacillus. Bifidobacterium species, particularly B. animalis subsp. lactis (BAL) BB-

12, the primary focus of this study, can be found in the gastrointestinal tract (GIT) as both

115 autochthonous (native to a particular place) and allochthonous (derived from outside a system) residents (Klijn et al. 2005). Newborns, especially those that are vaginally delivered and breast-fed, are colonized with bifidobacteria within days after birth. Once the child is weaned, the population of these bacteria in the colon appears to be relatively stable until advanced age when it appears to decline (Langhendries et al. 1995, Klijn et al. 2005). Furthermore, B. animalis subsp. lactis, unlike the majority of anaerobic bifidobacteria, are moderately tolerant of oxygen and acid (Vernazza et al. 2006), and survive well in fermented milk products (Desfossés-Foucault et al. 2012). It is for this reason that B. animalis subsp. lactis is extensively used in commercial application (Meile et al. 1997).

As part of a project that evaluated the effectiveness of B. animalis subsp. lactis

BB-12 delivered in dairy (yogurt smoothie) or non-dairy (capsule) vehicles in humans, this study was the first to investigate the influence of delivery vehicles of BB-12 on the gut microbiota of healthy adults. Thus, the primary goal was to survey the gut microbiota of participants before and after different BB-12 interventions. In an effort to understand the relationships between gut microbiota and host immune and physiological responses, the gut microbiota results were also analyzed together with a variety of other parameters including sex, body mass index (BMI), blood pressure, glucose, high-density lipoprotein

(HDL), low-density lipoprotein (LDL), triglycerides (TGs), tumor necrosis factor alpha

(TNF-α), and interferon gamma (IFN-γ), etc.

116

4.2. Materials and Methods

4.2.1. Design and intervention

This project was a randomized, partially blinded, 4-period crossover free-living study. The study design scheme is shown in Figure 4.S1. Measurements taken during the baseline visit were anthropometric assessment (age, gender, BMI, waist circumference), biochemical parameters (fasting serum glucose, insulin, and hs-CRP), a physical activity questionnaire (self-reported IPAQ), and an immune end point assessment for use as baseline values. Then each participant began the intervention phase as specified by the randomization order. The four treatments as described in chapter 2 were: YS (yogurt smoothie without BB-12), POST (yogurt smoothie with BB-12 added after fermentation),

PRE (yogurt smoothie with BB-12 added before fermentation), and CAP (BB-12 containing capsule). Each treatment period lasted 4 weeks and a two-week compliance break was scheduled between treatment periods.

During the yogurt smoothie treatment phases, participants consumed one 8-oz

(240 g) serving of yogurt per day. Study interventions contained between 3.16 × 109 -

3.16 × 1010 colony-forming unit (CFU) of BB-12 per gram. The yogurt smoothies were developed in our lab at the Pennsylvania State University (Merenstein et al. 2011) and modified for this project. During the CAP treatment phase, participants ingested one capsule per day. This CAP was specifically designed by Chr. Hansen (Milwaukee, WI) to deliver log 10 ± 0.5 CFU of BB-12/capsule. Participants were instructed to avoid consuming any other food or supplements containing probiotic bacteria such as commercial yogurt, smoothies, and probiotic capsules or tablets during each 4-week

117 intervention phase. They were also asked not to change their habitual diets during the course of the study and to maintain their body weight.

4.2.2. Participants

The primary outcome of the study, gut transit time, was used for the sample size calculation. Specifically, the calculation was based on the data from a previous study

(Bouvier et al. 2001) and indicated that 28 subjects were required to identify a mean difference of 6 hours in fecal transit time. Considering the possibility of attrition from the study and potential lack of compliance, 36 healthy volunteers (18-40 years of age) with delayed gut transit (≥ 24 hours) were recruited. Detailed exclusion criteria are described elsewhere (Meng et al. 2015). Briefly, all subjects were nonsmoking, normotensive and not diagnosed with any chronic medical conditions. A written informed consent was obtained from each participant. The study was approved by the Institutional Review

Board of the Pennsylvania State University (University Park, PA). The ClinicalTrials.gov identifier is NCT01399996.

4.2.3. Recruitment and screening

This clinical trial was a multidisciplinary project that heavily involved the

Departments of Food Science and Nutritional Sciences. Recruiting details are highlighted in Figure 4.S2. A total of 203 individuals expressed interest in participating in the study via phone or email after seeing advertisements in local newspaper or university email lists. During the initial contact they were informed about the study and then if still interested, a telephone screening was conducted using a series of medical and lifestyle

118 questions. Upon completion of the phone interview, 67% (136) of the screened subjects qualified for a clinic visit. During the clinic visit at the Penn State Clinical Research

Center (CRC), potential participants were given an informed consent document detailing the requirements for their involvement in the study. A staff member was available to answer any questions and then both the participant and the staff member had to sign and date the document. Participants’ height, weight, waist circumference and blood pressure were then measured, followed by a fasting blood draw for a complete blood count and health profile (liver and kidney function, and glucose metabolism). The body mass index of each individual was calculated from the measured weight and height. Participants also completed the Rome III (Drossman and Dumitrascu 2006) diagnostic questionnaire to assess their bowel function. After completing the screening procedures, 36 eligible volunteers were recruited for the study. To eliminate any possible confounding effect of treatment order on outcomes, eligible participants were randomized to treatment sequences. An online tool was used to generate the randomization schemes

(http://www.randomization.com). Two participants dropped out at the baseline visit due to schedule conflicts; four withdrew due to personal reasons; and one individual was excluded prior to the first treatment period due to follow-up failure. Therefore, 29 participants finished at least one treatment period. Of the 29 participants, two were excluded due to pregnancy prior to the third and fourth treatment periods, respectively; one withdrew from the study prior to the fourth treatment period due to diagnosis of irritable bowel syndrome; one dropped out prior to the second treatment period due to schedule conflict; and two left the study due to other personal reasons prior to the second

119 and third treatment periods, respectively. Thus, a total of 23 participants completed all the treatments.

4.2.4. Diet and physical activity assessment

To evaluate factors such as diet and physical activity that may contribute to the outcomes assessed, participants were asked to complete a 24-hour dietary recall for 3 continuous days to assess their dietary intake and the International Physical Activity

Questionnaires (IPAQ) questionnaire was employed to evaluate their physical activity level as previously reported (Bouchard et al. 1983). Briefly, participants were asked to recall their food and beverages intake during meals and snacks over three continuous days according to detailed instructions including portion size of each food item provided by trained staff. Daily intake of total calories, macronutrients, vitamins, minerals, caffeine, and alcohol was analyzed based on the recorded food intake using the Food

Processor SQL software (ESHA Research, Salem, OR). Participants also recorded the activity they performed during each of the 96 periods of 15 min over a 24-hour period for

3 days, including one weekend day. An intensity scale was used to categorize the recorded activities.

4.2.5. Stool sample collection

As mentioned in Chapter 3, each eligible participant collected fecal samples using a stool collection kit (Appendix III) at baseline, after each treatment period, and after the final washout as indicated by black bars in Figure 4.S1. Fecal samples were stored in

120 home freezer until brought to the Penn State Clinical Research Center (CRC). After fecal samples were received, they were stored at -80°C until DNA extraction.

4.2.6. Isolation of stool DNA

DNA was isolated from fecal samples using the MOBIO PowerSoil DNA isolation kit (Cat. # 12888, Carlsbad, CA) according to the manufacture’s protocol with modifications. Briefly, 250 mg of thawed stool sample was placed into a MOBIO

PowerBead tube using a sterilized toothpick. Next, 60 µl of sodium dodecyl sulfate

(SDS) based solution C1 was added followed by intensive homogenization (6500 rpm for

2 min) using Precellys-24 homogenizer (Bertin Corp., Rockville, MD). PowerBead tubes were then centrifuged at 10,000 × g for 30 s at room temperature. Supernatant (~ 400-500

µl) was transferred to a clean 2 ml collection tube and 250 µl of solution C2, which helped to precipitate cell debris and proteins, was added. The collection tube was vortexed for 5 s, incubated on ice for 5 min and then centrifuged at room temperature for

1 min at 16,000 × g. Up to 600 µl of supernatant was transferred to another clean 2 ml collection tube and 200 µl of solution C3 was added to further remove non-DNA organic and inorganic materials. Following incubation on ice for 5 min samples were centrifuged at room temperature for 1 min at 16,000 × g then up to 750 µl of supernatant was transferred to a third 2 ml collection tube and 1.2 ml of a well-mixed high concentration salt solution (C4) was added to the tube. The mixture was vortexed for 5 s and then about

675 µl was loaded onto a silica spin filter. The spin filter was centrifuged at room temperature for 1 min at 10,000 × g and the flow through was discarded. This step was repeated until all the mixture was filtered. 500 µl of ethanol based solution (C5) was

121 added onto the spin filter and centrifuged at room temperature for 30 s at 10,000 × g to further clean the DNA bound to the silica filter membrane. After discarding the flow, the spin filter was centrifuged for another minute to remove residual ethanol. After removal of the ethanol, the filter was carefully placed in a clean 2 ml collection tube and 100 µl of elution buffer solution (C6) was added to the center of the filter membrane, incubated for

2 min and then centrifuged at room temperature for 30 s at 10,000 × g. Finally, the filter was discarded and the genomic DNA on the bottom of the tube was stored at -80°C prior downstream analysis. DNA concentration was measured using Qubit® fluorometer

(Invitrogen, Grand Island, NY) for quality control purpose.

4.2.7. Compliance check using PCR

In order to check participant compliance, DNA extracted from the stool samples was analyzed for the presence or absence of BAL using the sub-species-specific PCR method (Ventura et al. 2001). Specifically, the PCR assays were carried out using primers Bflac2 (GTGGAGACACGGTTTCCC) and Bflac5

(CACACCACACAATCCAATAC), both of which were obtained from IDT DNA

(Coralville, IA). These primers are not BB-12 but rather BAL specific. The PCR amplification mixture consists of 10 × Buffer with MgCl2 (Promega, Madison, WI), 100

µM dNTPs (Promega), 1.0 µM of each primer, and 2.5 U Taq polymerase (Promega) brought to 25 µl with sterile distilled water. PCR for each sample was performed using the following conditions: initial denaturation, 95°C, 5 min; denaturation, 95°C, 30 s; anneal, 58°C, 60 s; elongation, 72°C, 2 min (repeat for 30 cycles); final elongation, 72°C,

7 min; cool to 4°C in a Mastercycler Personal (Eppendorf, Hamburg, Germany).

122

Amplicons were detected following agarose gel electrophoresis and staining with eithidium bromide. A 100-bp DNA ladder (Promega) molecular weight marker was electrophoresed at the same time to estimate the size of PCR products. An Alpha-Imager

(Alpha Innotech, San Leandro, CA) was used to visualize the bands. An amplicon size of

680 bp indicates positive for BAL.

4.2.8. Illumina® sequencing

DNA samples were prepared as previously described (Caporaso et al. 2011) with the following modifications. Universal primers F515 (5′–

NNNNNNNNGTGTGCCAGCMGCCGCGG- TAA–3′) and R806 (5′–

GGACTACHVGGGTWTCTAAT–3′), with the forward primer modified to contain an 8- nt barcode (italicized poly-N section of the primer above) and 2-nt linker sequence (bold portion) at the 5’ end, were used to amplify the V4 region of the 16S rRNA gene. PCR reactions contained 5.0 µl 2 × GoTaq Green Master Mix (Promega), 0.4 µl 25 mM

MgCl2, 2.4 µl water, 0.2 µl reverse primer (10 mM final concentration), 1.0 µl forward primer (2 mM final concentration) and 1.0 µl genomic DNA. Reactions were held at

94°C for 3 min to denature the DNA, with amplification proceeding for 25 cycles at 94°C for 45 s, 50°C for 60 s, and 72°C for 90 s; a final extension of 10 min at 72°C was included to ensure complete amplification. The PCR products were purified using

QIAquick PCR Purification Kit (Cat # 28106, Valencia, CA). A composite sample for sequencing was created by combining equimolar ratios of amplicons from individual samples, followed by gel purification and ethanol precipitation to remove any remaining contaminants and PCR artifacts. The composite sample was sequenced at the DNA

123

Technologies Core Facility of the University of California, Davis (UC Davis) on an

Illumina® Genome Analyzer II sequencing platform.

4.2.9. Bif-TRFLP

The Bif-TRFLP assay was performed based on the method described by Lewis et al. (Lewis et al. 2013) who developed this technique. Briefly, stool DNA samples were amplified for bifidobacterial 16S rRNA gene. The PCR reactions were carried out in a mixture (50 µl) that contained 1 µl of genomic DNA, 25 µl of 2 × Promega GoTaq Green

Master Mix (Promega), 20 µl of nuclease-free water, 1 µl of each primer (NBIF389, 5’-

[HEX]-GCCTTCGGGTTGTAAAC-3', 10 µM; NBIF1018REV, 5′-

GACCATGCACCACCTGTG-3′, 10 µM), and 2 µl of MgCl2 (25mM). The reaction conditions were 95°C for 5 min followed by 30 cycles of 95°C for 1 min, 52°C for 1 min and 72°C for 1 min. A final extension at 72°C for 5 min was allowed following the cycles and then samples were stored at 4°C prior to analysis. PCR products were confirmed for positive and negative control by gel electrophoresis before purification with QIAquick

PCR Purification Kit.

A portion (8 µl) of the purified DNA was cut with two restriction enzymes (AluI,

Thermo Scientific, Wahtham, MA; HaeIII, New England BioLabs Inc., Ipswich, MA) in separate reactions, both of which involved 1 µl of enzyme (10 unit/µl) in a 10 µl reaction for 3 h at 37°C. Then the enzymes were heat inactivated at 80°C for 20 min and samples were stored at 4°C. Next, 1 µl of the digested mixture (diluted 1:20 in elution buffer) was submitted for fragment analysis on an ABI 3730 Capillary Electrophoresis Genetic

Analyzer at UC Davis. The molecular size markers used were the ROX 50-500 size

124 standards (Gel Company Inc., San Francisco, CA). The results were read using Peak

Scanner software v1.0 (Applied Biosystems, Grand Island, NY). Detailed data processing is described in the original article (Lewis et al. 2013).

4.2.10. Sequence data analysis

The data analysis pipeline used was modified based on a previous study (Bokulich et al. 2014). Briefly, the QIIME software package (Caporaso et al. 2010) was used to analyze the results of the Illumina® sequencing run. Raw Illumina® fastq files were first demultiplexed and quality filtered. Reads were truncated after a maximum number of 3 consecutive low quality scores (< 1e-5), and any read containing one or more ambiguous base calls was discarded. Reads with a minimum pairwise identity of 97% were clustered into operational taxonomic units (OTUs) using QIIME’s open-reference OTU-picking workflow, which was based on UCLUST (Edgar 2010) software. The Greengenes bacterial 16S rRNA database (13_8 release) was used for the open-reference OTU- picking (DeSantis et al. 2006). The most abundant sequence was chosen to represent each

OTU. Taxonomy was assigned to each OTU using QIIME-based wrapper of the

Ribosomal Database Project (RDP) classifier (Wang et al. 2007) against a representative subset of the Greengenes 16S rRNA database 13_8 release, using a 0.50 confidence threshold for taxonomic assignment. Bacterial 16S rRNA gene sequences were aligned using PyNAST (Caporaso et al. 2010) against a template alignment of the Greengenes core set filtered at 97% similarity. During the process, chimeras were identified and removed using the ChimeraSlayer (Haas et al. 2011) algorithm and a phylogenetic tree was built from the filtered alignment using FastTree (Price et al. 2010). Any OTU

125 representing less than 0.001% of the total filtered sequences was removed to avoid erroneous reads that could lead to inflated estimates of diversity (Bokulich et al. 2013).

After these quality-filtering steps, each sample was represented by less than 150 sequences and the filtered OTU tables were ready for downstream analyses, such as diversity comparisons and biomarker discovery.

Alpha-diversity (within-sample species richness) and beta-diversity (between- sample community dissimilarity) were calculated within QIIME based on weighted

UniFrac (Lozupone et al. 2006) distance between samples. Principle coordinates were calculated from the UniFrac distance matrices to decrease the dimensionality of the taxonomic dataset into 3D principal coordinate analysis (PCoA) plots, enabling visualization of sample relationships. To determine whether sample classification

(treatments) caused differences in phylogenetic or species diversity, ANOSIM (Clarke

1993) and permutational multivariate analysis of variance (MANOVA) (Anderson 2008) were used to test significant differences between sample groups based on weighted

UniFrac.

Significant taxonomic differences between sample groups were also tested using the Linear Discriminant Analysis (LDA) effect size (LEfSe) (Segata et al. 2011). LEfSe is an algorithm for high-dimensional (OTU tables) biomarker discovery and identification of genomic features (genes, pathways, or taxa) that characterizing the differences between two or more classes/treatments. It first uses the non-parametric factorial

Kruskal-Wallis (KW) sum-rank test to detect taxa with significant differential abundances with respect to the class of interest (one-against-all strategy). Then LEfSe

126 uses Linear Discriminant Analysis to estimate the effect size of each differentially abundant feature.

4.3. Results and Discussion

4.3.1. Participant characteristics

Baseline characteristics of the participants (anthropometric indexes, blood pressure, biochemical features, and physical activity) are shown in Table 4.1. Twenty- nine participants (18 females and 11 males) were included in the analyses, as they have completed at least one of the four intervention periods. Overall, participants were healthy, young adults (mean age of 28.1 ± 0.6 years). The average BMI was 24.1 ± 0.2 kg/m2; 16

(58.6%) participants were normal weight, 11 (37.9%) were overweight, and 1 (3.5%) was obese. Their blood pressure was normal and waist circumference, fasting blood glucose, insulin, and CRP levels were within the normal range (Table 4.1). Physical activity, as assessed from self-reported IPAQ responses, indicated a median daily physical activity of

3.0 METs (range 2.3 – 4.3 METs). The average daily total calorie intake of participants calculated from 3-day dietary recall records was estimated to be 2241 ± 83 kcal. The daily intake of macronutrients, vitamins, minerals and n-3 PUFA, caffeine and alcohol are also reported in Table 4.1.

127

Table 4.1. Demographic characteristics of participants at baseline1*

Characteristics Values (n = 29) Age (yr) 28.1 ± 0.6 Male, n (%) 11 (37.9%) Body mass index (kg/m2) 24.1 ± 0.2 ≤24.9 17 (58.6%) 25.0 -29.9 11 (37.9%) ≥30 1 (3.5%) Waist Circumference (cm) 85.1 ± 0.6 Blood pressure (mm Hg) Systolic 107.6 ± 0.8 Diastolic 72.6 ± 0.6 Glucose (mg/dL) 86.6 ± 0.8 Insulin (mg/dL) 5.3 ± 0.4 hs-CRP (mg/L) 2.0 ± 0.5 Physical activity (METs)2 3.0 (2.3 – 4.3) Dietary intake2 of Total calories (kcal/d) 2241 ± 83 Carbohydrate (g/d) 284.6 ± 10.9 Protein (g/d) 90.2 ± 3.7 Fat (g/d) 83.8 ± 3.4 Vitamin C (mg/d) 67.8 ± 4.8 Vitamin D (IU/d) 98.5 ± 11.2 Vitamin E (mg/d) 3.4 ± 0.3 Iron (mg/d) 14.4 ± 0.8 Selenium (µg/d) 46.0 ± 4.5 Zinc (mg/d) 5.7 ± 0.4 n-3 PUFA (g/d) 0.6 ± 0.1 Caffeine (mg/d) 75.3 ± 10.2 Alcohol consumption (g/d) 2.6 ± 0.9

1 Values are presented as mean ± SEM or n (%) or median (range). 2 Physical activity and dietary intake were assessed from self-reported responses to IPAQ and 3-day dietary recall records, respectively. * Shared results with collaborators in the project.

128

4.3.2. Compliance

Stool DNA samples of all participants at baseline, the end of each intervention period, and the end of final washout were tested for compliance using 16S rDNA based species-specific PCR. Overall, the results indicated good compliance status (Table 4.2).

This means most of the stool (73/78) samples were BAL positive after BB-12 interventions while all the fecal samples were BAL negative after receiving BB-12 free intervention (control yogurt smoothie) when the corresponding baselines were negative.

On the other hand, participants PRO 007, 017, 024, and 032 were BAL positive throughout the study regardless of treatments. It is possible that BAL were autochthonous to these individuals as they were BAL positive both at baseline and after final washout.

129

Table 4.2. B. animalis subsp. lactis PCR results of the participants’ stool samples ID Baseline CONTROL POST PRE CAP Final Washout PRO 002 + - N/A N/A N/A N/A PRO 003 - - + + + - PRO 004 + - + + + - PRO 006 - - N/A + + N/A PRO 007 + + + + + + PRO 008 - - + + + - PRO 009 - - + + + - PRO 011 - - + + + - PRO 012 - - N/A + N/A N/A PRO 013 - - + + - - PRO 014 - - + + + - PRO 015 + - + + + - PRO 016 - - + + + - PRO 017 + + + + + N/A PRO 019 - - + - + - PRO 020 - - + + + - PRO 021 - - - + + - PRO 022 - - + + + - PRO 024 + + + + + + PRO 025 - - + + - - PRO 026 + N/A - + + N/A PRO 028 - - + + + + PRO 029 + - + + + + PRO 030 - - + + + + PRO 032 + + + + + + PRO 033 - - + + + + PRO 034 - N/A + N/A + N/A PRO 035 - - + + + + PRO 036 - N/A N/A N/A + N/A

+ BAL positive - BAL negative N/A not applicable due to dropout Abbreviations: YS yogurt smoothie without BB-12, POST yogurt smoothie with BB-12 added post fermentation, PRE yogurt smoothie with BB-12 added before fermentation, CAP BB-12 containing capsule.

130

4.3.3. Compositional characteristics of the fecal microbiota of the participants

Illumina® sequencing of the 161 fecal samples generated over 2.6 million total reads. After removing samples with low quality and poor compliance, about 2.4 million sequences from 147 samples were used for data analyses. Sequence rarefaction curves, assessing per-sample sequence coverage, demonstrated adequate sequencing depth

(Figure 4.S3). Overall, 10 phyla and 109 genera were identified in the participants.

Firmicutes, Bacteroidetes, Actinobacteria, and Proteobacteria accounted for > 98% of the sequences at the phylum level. The predominant phyla and genera identified (represents >

0.1% of the sequences) among treatment groups are shown in Table 4.3. No difference at phylum or genus level was detected among treatment groups except that participants had significantly higher percentage of Streptococcus in their fecal microbiota after consuming control yogurt smoothie when compared to consuming capsule (P = 0.01). All yogurt smoothies tended to result in higher percentage of Streptococcus when compared to baseline, capsule, and final washout. This is likely due to the presence of the high level

(Log10 11.4 CFU/day) of the culture S. thermophilus in the yogurt interventions.

The Firmicutes/Bacteroidetes (F/B) ratio (median 6.79) of the present study cohort is at the high end when compared with the results of the Human Microbiome

Project (HMP) (Human Microbiome Project Consortium 2012). However, the results are comparable to other studies (Ley et al. 2006, Turnbaugh et al. 2009, Lin et al. 2013). The

F/B ratio is known to vary dramatically from population to population and from age group to age group (Mariat et al. 2009). De Filippo et al. compared the fecal microbiota of European children and that of children from a rural African village and observed F/B ratios of 2.8 ± 0.06 and 0.47 ± 0.05, respectively (De Filippoa et al. 2010).

131

Table 4.3. Predominant fecal bacterial phyla and genera present in healthy adults before and after consuming BB-12 containing interventions in a crossover study1

% of sequences Phylum and genus Baseline YS POST PRE CAP Final Washout SEM P Firmicutes 82 82 80 79 82 79 0.76 0.58 Blautia 8.4 8.3 7.9 6.8 8.2 7.3 0.35 0.77 Faecalibacterium 5.2 6.3 5.4 6.4 6.2 6.9 0.34 0.69 Ruminococcus 4.9 4.1 5.4 5.4 5.7 5.3 0.33 0.78 Coprococcus 2.7 3.3 3.3 2.8 3.1 3.1 0.13 0.74 Roseburia 1.7 1.8 1.8 1.4 2.0 2.6 0.19 0.62 Lachnospira 1.2 1.8 2.1 1.5 1.6 1.7 0.13 0.21 Dialister 1.0 1.1 1.1 1.3 1.4 0.8 0.12 0.77 Dorea 1.0 1.1 0.89 0.67 1.0 0.9 0.05 0.20 Streptococcus 0.59ab 1.04a 0.89ab 0.73ab 0.45b 0.38b 0.06 0.01 Clostridium 0.56 0.77 0.55 0.57 0.70 0.81 0.06 0.76 Oscillospira 0.83 0.51 0.61 0.69 0.57 0.56 0.05 0.46 Enterococcus 0.40 0.40 0.41 0.42 0.35 0.37 0.02 0.81 Lactobacillus 0.22 0.29 0.30 0.32 0.48 0.64 0.07 0.54 Lachnobacterium 0.29 0.28 0.08 0.47 0.19 0.33 0.07 0.72 Anaerostipes 0.23 0.18 0.22 0.23 0.31 0.20 0.02 0.65 Turicibacter 0.21 0.17 0.13 0.08 0.12 0.24 0.02 0.45 Megasphaera 0.08 0.07 0.10 0.21 0.10 0.13 0.03 0.79 Bacteroidetes 12 13 14 14 12 14 0.64 0.81 Bacteroides 8.3 8.9 9.9 9.2 7.9 10.3 0.53 0.81 Parabacteroides 0.92 1.0 1.1 1.0 1.0 1.1 0.08 0.96 Prevotella 0.44 0.46 0.38 0.81 0.28 0.28 0.14 0.91 Actinobacteria 3.7 2.8 3.6 4.4 3.7 3.7 0.36 0.72 Bifidobacterium 3.5 2.6 3.3 4.2 3.4 3.5 0.36 0.71 Collinsella 0.11 0.12 0.11 0.12 0.13 0.14 0.01 0.96 Proteobacteria 1.5 1.7 1.5 1.8 1.3 1.5 0.08 0.49 Sutterella 0.22 0.30 0.27 0.42 0.30 0.40 0.03 0.34 Citrobacter 0.22 0.20 0.21 0.37 0.22 0.24 0.03 0.32 Haemophilus 0.13 0.23 0.10 0.12 0.05 0.14 0.03 0.64 Verrucomicrobia 0.19 0.11 0.24 0.18 0.15 0.36 0.03 0.94 Akkermansia 0.19 0.11 0.24 0.18 0.15 0.36 0.00 0.41 1 Values are means with pooled SEMs, n = 147. Values in a row without a common letter are significantly different, P < 0.05. Abbreviations: YS yogurt smoothie without BB-12, POST yogurt smoothie with BB-12 added post fermentation, PRE yogurt smoothie with BB-12 added before fermentation, CAP BB-12 containing capsule.

132

133

Figure 4.1. Taxonomic cladogram of LDA effect size comparing relative abundance of taxa between male and female. Significantly discriminant taxon nodes are colored and branch areas are shaded according to the highest-ranked variety for that taxon. For each taxon detected, the corresponding node in the taxonomic cladogram is colored according to the highest-ranked group for that taxon. If the taxon is not significantly differentially represented between sample groups, the corresponding node is colored yellow (Bokulich et al. 2014).

Comparisons at various taxonomic levels between each treatment period and baseline were performed using LEfSe (Figure 4.S4). Overall, there were few difference between any treatment and baseline other than overrepresented Streptococcus in yogurt groups. Interestingly, a gender difference was observed in the same dataset (Figure 4.1).

Females had significantly more abundant of Paraprevotella, Butyricimonas,

Parabacteroides, Bacteroides, Collinsella, Bifidobacterium, Varibaculum,

Methanobrevibacter, and Oscillospira while males had significantly higher percentage of

Anaerostipes, Blautia, Dorea, Lachnobacterium, and Roseburia when compared with each other. Gender differences have also been observed in an animal study (Yurkovetskiy et al. 2013) and in other human studies (Mueller et al. 2006, Solano-Aguilar et al. 2013), although some studies have reported no difference (Arumugam et al. 2011) or a modest association between gender and the gut microbiota (Human Microbiome Project

Consortium 2012).

In the present study, data was re-analyzed for each gender group to determine if females and males respond differently to the interventions (Figure 4.S5 and Figure

4.S6). Results indicate that females and males did respond differently to interventions as

Haemophilus and Streptococcus were enriched in female while Lachnospira was

134 overrepresented in male. It is not clear how these differences impact the overall picture of the host gut microbiota or if there is any physiological relevance to this observation.

In summary, only limited compositional changes were detected in participants’ fecal microbiota after consuming BB-12 containing capsule and yogurt smoothie with or without BB-12 compared to baseline. Yogurt interventions appeared to result in elevated relative abundance of Streptococcus in host fecal microbiota, especially in female participants; but gender appeared to be a more significant factor in shaping the gut microbiota than the treatments at least in this study cohort.

4.3.4. Diversity of the participants’ fecal microbiota

There were no significant differences in alpha diversity indices (Chao1 richness,

Simpsons diversity and Shannon evenness) between treatment groups (Figure 4.2) or between genders (data not shown). a

135 b

c

Figure 4.2. Boxplot showing the comparison of alpha diversity indices of the fecal microbiota of participants before and after each treatment. (a, Chao1 richness; b, Simpson diversity; c, Shannon evenness). There were no significant differences between treatment groups in any index.

136

A definitive stratification according to treatment group was not evident on

UPGMA tree (Unweighted Pair Group Method with Arithmetic Mean), but samples from each individual tend to cluster together (Figure 4.3). Permutational MANOVA also

2 confirmed the similarity between treatments (RANOSIM = -0.041, P = 1.0; R ADONIS =

0.024, P = 1.0) and dissimilarity between individuals (RANOSIM = 0.925, P = 0.001;

2 R ADONIS = 0.433, P = 0.001). In this case, the results indicate that individual characteristics played a bigger role than treatments in shaping the host gut microbiota.

This is supported by recent research on both animals (Benson et al. 2010) and humans

(Goodrich et al. 2014). Particularly, Goodrich et al. studied over 1,000 stool samples from 416 twin pairs. They found that fecal microbiota were more similar overall within individuals (resampled) than between unrelated individuals (P < 0.001) and were also more similar within twin pairs than unrelated individuals (P < 0.009). Moreover, monozygotic twin pairs had more similar gut microbiota than dizygotic twin pairs (P =

0.032) (Goodrich et al. 2014).

137

Figure 4.3. UPGMA tree based on weighted UniFrac distance (beta diversity) demonstrating the hierarchical relationships between the fecal samples The code is participant’s study id followed by treatment period (i.e. TP1 is the first treatment period, BL is baseline, FL5 is final washout). The data indicates that samples from each individual tend to cluster together regardless of treatment.

138

a

b

Figure 4.4. Weighted UniFrac distance PCoA of bacterial communities with jackknife support grouped by treatment and gender. No patterns of clustering when samples were colored by treatments while samples tend to cluster based on gender, suggesting a gender difference.

139

A visible difference in community structure between treatment groups was not noted on PCoA plot; however, samples tended cluster when grouped by gender (Figure

4.4). Statistical analysis revealed that community membership was different between

2 males and females (RANOSIM = 0.096, P = 0.002; R ADONIS = 0.021, P = 0.001), even although only a small percentage of differences can be explained by the dataset. Other metadata (i.e. glucose, high/low-density lipoprotein, triglycerides, C-reactive protein, tumor necrosis factor alpha, interferon gamma, and etc.) were also screened for possible associations with the host gut microbiota data, a number of statistically significant differences were found (Table 4.S1). However, further studies are needed to validate these relationships, because either only small percentage of difference between high and low can be explained by the grouping or only a few data points were in one of the two arms.

Taken together, the diversity results are not completely surprising. The participants of the present study cohort were healthy individuals and over the last 5 years it has become well understood that a healthy gut microbiota is stable and resilient

(Human Microbiome Project Consortium 2012, Lozupone et al. 2012, Ursell et al. 2012,

Coyte et al. 2015). Although consuming a high-fat and low-fiber, or low-fat and high- fiber diet for 10 days can induce statistically significant changes in the gut microbiota, these changes in species and gene content are small compared with baseline variations that occur between individuals (Wu et al. 2011).

In previous work, McNulty et al. (McNulty et al. 2011) repeatedly sampled seven healthy adult female monozygotic twin pairs (aged 21 – 32 years, BMI 20 – 25 kg/m3) 4

140 weeks before, 7 weeks during, and 4 weeks after consumption of a commercially available fermented milk product (8 ounce/day) containing a consortium of

Bifidobacterium animalis subsp. lactis strain CNCM I-2494 (3.2 × 107 CFU/g),

Lactobacillus delbrueckii subsp. bulgaricus strains CNCM I-1632 and I-1519 (6.3 × 107

CFU/g), Lactococcus lactis subsp. cremoris strain CNCM I-1631, and Streptococcus thermophilus strain CNCM I-1630. They found the species and gene content of the twins’ gut microbial communities remained stable and were not appreciably perturbed by consuming the intervention. However, further study in mice revealed introducing the fermented milk product strains resulted in marked changes in metabolic pathways related to carbohydrate processing; although the proportional representation of their gut microbiota acquired from their human donors remained the same. The authors suggested analyses of the bacterial species and gene content of the gut microbiota/microbiome may not be an informative biomarker for understanding whether or how the interventions may have affected microbial community properties.

4.3.5. Fecal bifidobacterial distribution

To further explore the effect of the study interventions on the host gut bifidobacteria at species level, stool DNA samples were subjected to Bif-TRFLP (Lewis et al. 2013). Results are shown in Figure 4.5. The participants had significantly higher percentage of B. animalis in their feces after consuming the two BB-12 containing yogurt smoothies (PRE and POST) when compared to baseline, other interventions (YS or

CAP), and final washout (P < 0.0001). No significant difference was detected between

PRE and POST. It appears the yogurt smoothie resulted in higher relative abundance of

141

B. animalis in the stool samples than capsules. This may be due to the buffering capacity of milk components, especially milk protein (Ziarno and Zareba 2015). Precaution needs to be taken when extrapolating findings, because the results presented here are not absolute concentrations. Moreover, it is not clear what the physiological consequences are when the host has a higher abundance of B. animalis in their feces.

The physiological effect of delivery matrix on the performance of probiotics was observed in a recent animal study. Lee et al. (Lee et al. 2015) fed dextran sulfate sodium

(DSS) – induced ulcerative colitis mice with a wild-type and two mutant (DltD- and

RecA-) probiotic strains of L. casei (2 × 107 CFU/feeding) in milk or a nutrient-free buffer prior to and during administration of DSS for 15 consecutive days. Live L. casei cells in stool samples were recovered using selective MRS media. A disease activity index (DAI) was calculated by percent total weight loss (before-/after DSS-treatment), histology score, the presence of blood in stools, and stool consistency. Cytokine and chemokine in ileal and colonic tissues were measured using Bio-Plex ProTM mouse cytokine 23-plex panel. The results of this study showed mice fed with L. casei in milk had lower DAI than those fed with L. casei in nutrient-free buffer, milk only, and mutant in either milk or buffer, suggesting that milk might be the preferred delivery matrix for certain probiotic strains. Clearly additional studies are needed to evaluate other strains, other delivery matrix, and different disease conditions. More importantly, the precise mechanism needs to be further explored and clearly understood.

142 100%

90%

80%

70% B. adolescentis 60% B. animalis 50% B. breve B. bifidum/pseudocat 40% B. longum Relative proportion proportion Relative 30% B. pseudocatenulatum 20% Unknown

10%

0% Baseline YS (A) POST (B) PRE (C) CAP (D) Final Washout Treatment

Figure 4.5. Relative proportion of Bifidobacterium species in the stool DNA samples at baseline, after each treatment, and after final washout as determined by Bif-TRFLP (AluI).

143 4.4. Conclusions

The present study evaluated the effect of the probiotic Bifidobacterium animalis subsp. lactis BB-12 alone (capsule), or when in yogurt smoothies containing BB-12 added pre or post yogurt fermentation, on the composition of the gut microbiota and the bifidobacteria profile of healthy adults. No significant treatment effect on the gut microbiota was detected due to the large interpersonal and intrapersonal variations observed except that yogurt interventions generally resulted in a higher abundance of

Streptococcus. A significant gender effect was observed when comparing the gut microbiota of present study cohort. Interestingly, the two BB-12-containing yogurt smoothies (PRE and POST) resulted in significantly higher percentage of B. animalis when compared to baseline, to the BB-12-free yogurt smoothie (YS), the capsule (CAP), and final washout when analyzed by Bif-TRFLP. No difference was detected between

PRE and POST addition of BB-12.

144

Figure Legends

Figure 4.S1. Schematic diagram for randomization design. Each period lasted for 4 weeks and there was a 2-week compliance break between treatments. The whole study lasted for 6.5 months for each volunteer. Stool and blood samples were collected before the first treatment and after each treatment. Only a stool sample was taken after the final washout.

Figure 4.S2. Recruitment strategy used in the study. YS yogurt smoothie without BB-12, PRE yogurt smoothie with BB-12 added before fermentation, POST yogurt smoothie with BB-12 added after fermentation, CAP BB-12 containing capsule.

Figure 4.S3. Alpha rarefaction curves of the sequences. Each line represents one fecal sample. The data indicates most samples had sequence coverage over 10,000 reads and 5,000 reads are adequate to cover most of the OTUs.

Figure 4.S4. Histogram of the LDA scores computed for features differentially abundant between intervention or final washout and baseline. (a) OTU comparison between control yogurt smoothie and baseline. (b) OTU comparison between yogurt smoothie with BB-12 added post fermentation and baseline. (c) OTU comparison between yogurt smoothie with BB-12 added before fermentation and baseline. (d) OTU comparison between final washout and baseline.

Figure 4.S5. Histogram of the LDA scores computed for features differentially abundant between intervention or final washout and baseline in FEMALE participants. (a) OTU comparison between control yogurt smoothie and baseline. (b) OTU comparison between yogurt smoothie with BB-12 added post fermentation and baseline. (c) OTU comparison between yogurt smoothie with BB-12 added before fermentation and baseline. (d) OTU comparison between BB-12 containing capsule and baseline. (e) OTU comparison between final washout and baseline.

Figure 4.S6. Histogram of the LDA scores computed for features differentially abundant between intervention or final washout and baseline in MALE participants. (a) OTU comparison between control yogurt smoothie and baseline. (b) OTU comparison between yogurt smoothie with BB-12 added post fermentation and baseline. (c) OTU comparison between yogurt smoothie with BB-12 added before fermentation and baseline. (d) OTU comparison between final washout and baseline.

145

Figure 4.S1.

146

Figure 4.S2.

Enrollment

Assessed for eligibility (n=136)

Excluded (n=100) • Did not meet inclusion criteria (n= 41) • Declined to participate (n=0) Allocation • Other reasons (n=59)

Randomized (n=36)

Randomly allocated to receive interventions, including YS, PRE, POST, or CAP (n=36) • Received allocated intervention (n=29) • Did not receive allocated intervention (n=7) o Schedule conflict (n=2) o Lost to follow-up (n=1) o Other person reasons (n=4)

Crossover

Crossed over to receive other interventions, including YS, PRE, POST, or CAP (n=29) • Received allocated intervention (n= 29) • Did not receive allocated intervention (n=0)

Follow-Up

Lost to follow-up (n=0) Discontinued intervention after allocation (n=6) • Pregnancy (n=2)

• Diagnosed of Irritable Bowel Syndrome (n=1) • Schedule conflict (n=1) • Other personal reasons (n=2)

Analysis

Analysed (n=29) Excluded from analysis (n=0) 147

Figure 4.S3.

148

Figure 4.S4.

a

b

c

d

149

Figure 4.S5.

a

b

c

d

e

150

Figure 4.S6.

a

b

c

d

151

Table 4.S1. Statistical analyses of the weighted UniFrac distance (beta-diversity) when grouped by the metabolic and immune parameters1

ADONIS ANOSIM R2 P value R P value CRP 0.007 0.372 0.041 0.316 C_H 0.015 0.025 0.371 0.039 Glucose 0.013 0.025 0.089 0.200 Insulin 0.020 0.003 0.177 0.026 WC 0.011 0.076 0.118 0.051 TGs 0.010 0.116 -0.092 0.800 HDL 0.010 0.095 -0.118 0.963 LDL 0.016 0.006 0.015 0.414 DCs 0.015 0.008 0.020 0.018 IFN-γ 0.018 0.003 0.019 0.033 TNF-α 0.007 0.445 -0.005 0.740 IL-2 0.008 0.272 0.001 0.358

1Results of these parameters were transformed to categorical data (high or low) based on the medians. Abbreviations: CRP C-reactive protein, C_H cholesterol to high-density lipoprotein (HDL) ratio, WC waist circumference, TGs triglycerides, LDL low-density lipoprotein, DCs dendritic cells, IFN-γ interferon gamma, TNF-α tumor necrosis factor alpha, IL-2 interleukin-2.

152

Chapter 5 – Conclusions and Future Directions

5.1. Conclusions

The present study identified MRS-NNLP agar as a suitable selective medium for enumeration of Bifidobacterium animalis subsp. lactis BB-12 (BB-12) in yogurt smoothies where the concentration of Streptococcus thermophilus was higher than that of

BB-12. The population of BB-12 declined in both yogurt smoothies containing BB-12 throughout shelf life. This reduction was faster in POST addition samples than in the

PRE addition samples likely due to BB-12’s adaptation to the acidic environment during fermentation Even though the population declined, the viable counts of BB-12 remained at the specified dose level (log10 10 ± 0.5 CFU/serving) in both products at the end of shelf life. Together with a control yogurt smoothie (YS, without BB-12), three yogurt smoothies with identical nutritional composition (fat and total solids) were developed and produced for this clinical trial.

In an effort to understand the effect of BB-12, BB-12 delivery vehicles (yogurt smoothie vs. capsule), and timing of BB-12 addition on their efficacy in vivo, participants’ gut transit time (GTT) and fecal short chain fatty acid (SCFA) concentration were assessed via a wireless motility capsule (SmartPill®) and Gas Chromatography

(GC), respectively. The present study confirmed the precision of the blue dye method for assessing whole gut transit time of healthy adults and found it was a reliable cost-efficient method for baseline screening when compared to the SmartPill®. However, we noted that some participants had a difficult time detecting the blue dye end point. During the study we evaluated the effect of BB-12 alone (capsule: CAP), delivered in different matrixes

153

(yogurt vs. capsule), and when added before (PRE) or after (POST) fermentation, on

GTT and fecal SCFA production in young healthy adults. No significant treatment effect was observed on either outcome. However, this study was the first to identify a possible relationship between regional gut transit times and fecal SCFAs in healthy adults. Data from this study confirmed a number of previously reported correlations (Table 3.5).

Notably, the predominant short chain fatty acids negatively correlated with whole gut transit time (WGTT), colonic transit time (CTT), and gastric emptying time (GET), but did not correlate with small bowel transit time (SBTT).

The gut microbiota of the participants’ were analyzed for possible treatment effect on the composition, diversity, and the bifidobacterial profile using 16S rDNA based

Illumina® sequencing and bifidobacteria specific T-RFLP. No significant treatment effect on the gut microbiota was detected due to the large interpersonal and intrapersonal variations observed except that yogurt interventions tended to induce a higher relative abundance of Streptococcus. A significant gender effect was observed when comparing the gut microbiota of the study cohort. Interestingly, T-RFLP analysis revealed the two

BB-12-containing yogurt smoothies (PRE and POST) resulted in a significantly higher percentage of B. animalis in stool samples when compared to baseline, to the BB-12-free yogurt smoothie (YS), to the capsule (CAP), and to final washout samples. No difference was detected between PRE and POST treatments. However, since T-RFLP is a relative measurement, further studies are needed to validate the absolute changes in concentration of Bifidobacterium.

Immune outcomes of this project (Meng et al. 2015) indicated consumption of

154

BB-12 delivered in a yogurt smoothie significantly reduced TNF-α secretion from peripheral blood mononuclear cells (PBMCs) stimulated with heat-inactivated BB-12 or lipopolysaccharide (LPS) in the participants but not in capsule, suggesting an anti- inflammatory effect of BB-12 consumption when administered in yogurt in healthy adults. Additionally, BB-12 was found to interact with peripheral myeloid cells via Toll- like receptor 2 (TLR-2).

In summary, consumption of BB-12 delivered in yogurt or capsule did not change the gut transit time, fecal short chain fatty acid concentration, or the composition of the gut microbiota of the study cohort. However, daily consumption of BB-12 in yogurt smoothie for 4 weeks may have an anti-inflammatory effect, and result in higher relative abundance of B. animalis in young healthy adults.

5.2. Future Directions

Moving forward from the current study, it would be interesting to investigate the absolute concentration of B. animalis in the stool samples as the data indicated that the two yogurt smoothies resulted in significantly higher relative abundance of this species when compared to baseline, control yogurt smoothie, capsule, and final washout. In order to better understand any potential effect of the present study interventions may have on young healthy adults, it would be interesting to explore metagenomic and metabolomic analyses on the blood and stool samples collected. This could provide detailed pictures of hosts’ metabolic response to the interventions and gut microbial genes in respond to the interventions.

155

Overall, the limited treatment effect observed on gut transit time, fecal SCFA, and the gut microbiota of this study cohort taken together with a lack of clinical effect seen in the clinical part of this study (unpublished) and previous trials (Christensen et al. 2006,

West et al. 2012) indicate that the current approach of using the probiotic

Bifidobacterium animalis subsp. lactis BB-12 for improving the health of young healthy adults may be futile. The limitation of using healthy subjects in similar studies is that it is rather difficult to improve their already healthy and resilient physiological state, and even more challenging to identify parameters of relevance to monitor the effects of probiotics from a disease-prevention point of view when the precise mechanisms of probiotic action remain unclear.

Therefore, it is suggested for future studies to use strictly selected cohorts

(constipated or other diseases) to further explore the health benefit of the probiotic BB-12 as strains within this subspecies showed beneficial effect (increase) on stool frequency in gut transit delayed individuals (Guyonnet et al. 2007, Agrawal et al. 2009, Tabbers et al.

2009). A similar study on patients with disturbed gut microbiota (dysbiosis) should also be performed, because BB-12 may have beneficial effect on restoring homeostasis

(Gerritsen et al. 2011).

156

References

Aagaard, K., J. Ma, K. M. Antony, R. Ganu, J. Petrosino and J. Versalovic (2014). "The placenta harbors a unique microbiome." Science Translational Medicine 6(237): 237ra265.

Aaltonen, J., T. Ojala, K. Laitinen, T. J. Piirainen, T. A. Poussa and E. Isolauri (2008). "Evidence of infant blood pressure programming by maternal nutrition during pregnancy: a prospective randomized controlled intervention study." The Journal of Pediatrics 152(1): 79-84.

Aaltonen, J., T. Ojala, K. Laitinen, T. Poussa, S. Ozanne and E. Isolauri (2011). "Impact of maternal diet during pregnancy and breastfeeding on infant metabolic programming: a prospective randomized controlled study." European Journal of Clinical Nutrition 65(1): 10–19.

Abrahamsson, T. R., H. E. Jakobsson, A. F. Andersson, B. Björkstén, L. Engstrand and M. C. Jenmalm (2012). "Low diversity of the gut microbiota in infants with atopic eczema." The Journal of Allergy and Clinical Immunology 129(2): 434-440.

Adolfsson, O., S. N. Meydani and R. M. Russell (2004). "Yogurt and gut function." The American Journal of Clinical Nutrition 80(2): 245-256.

Agrawal, A., L. A. Houghton, J. Morris, B. Reilly, D. Guyonnet, N. G. Feuillerat, A. Schlumberger, S. Jakob and P. J. Whorwell (2009). "Clinical trial: the effects of a fermented milk product containing Bifidobacterium lactis DN-173 010 on abdominal distension and gastrointestinal transit in irritable bowel syndrome with constipation." Alimentary Pharmacology Therapeutics 29(1): 104-114.

Ahmed, M., J. Prasad, H. Gill, L. Stevenson and P. Gopal (2007). "Impact of consumption of different levels of Bifidobacterium lactis HN019 on the intestinal microflora of elderly human subjects." The Journal of Nutrition, Health & Aging 11(1): 26-31.

Akatsu, H., N. Iwabuchi, J. Z. Xiao, Z. Matsuyama, R. Kurihara, K. Okuda, T. Yamamoto and M. Maruyama (2013). "Clinical effects of probiotic Bifidobacterium longum BB536 on immune function and intestinal microbiota in elderly patients receiving enteral tube feeding." Journal of Parenteral & Enteral Nutrition 37(5): 631-640.

Alonso, V. R. and F. Guarner (2013). "Linking the gut microbiota to human health." British Journal of Nutrition 109(Suppl 2): S21-S26.

Alvaro, E., C. Andrieux, V. Rochet, L. Rigottier-Gois, P. Lepercq, M.Sutren, P. Galan, Y. Duval, C. Juste and J. Dore (2007). "Composition and metabolism of the intestinal microbiota in consumers and non-consumers of yogurt." British Journal of Nutrition 97: 126-133.

157

Anderson, M. J. (2008). "A new method for non-parametric multivariate analysis of variance." Austral Ecology 26(1): 32-46.

Arnold, I. C., N. Dehzad, S. Reuter, H. Martin, B. Becher, C. Taube and A. Müller (2011). "Helicobacter pylori infection prevents allergic asthma in mouse models through the induction of regulatory T cells." The Journal of Clinical Investigation 121(8): 3088- 3093.

Arumugam, M., J. Raes, E. Pelletier, D. L. Paslier, T. Yamada, D. R. Mende, G. R. Fernandes, J. Tap, T. Bruls, J. M. Batto, M. Bertalan, N. Borruel, F. Casellas, L. Fernandez, L. Gautier, T. Hansen, M. Hattori, T. Hayashi, M. Kleerebezem, K. Kurokawa, M. Leclerc, F. Levenez, C. Manichanh, H. B. Nielsen, T. Nielsen, N. Pons, J. Poulain, J. Qin, T. Sicheritz-Ponten, S. Tims, D. Torrents, E. Ugarte, E. G. Zoetendal, J. Wang, F. Guarner, O. Pedersen, W. M. Vos, S. Brunak, J. Doré, M. Consortium, M. Antolín, F. Artiguenave, H. M. Blottiere, M. Almeida, C. Brechot, C. Cara, C. Chervaux, A. Cultrone, C. Delorme, G. Denariaz, R. Dervyn, K. U. Foerstner, C. Friss, M. Guchte, E. Guedon, F. Haimet, W. Huber, J. Hylckama-Vlieg, A. Jamet, C. Juste, G. Kaci, J. Knol, O. Lakhdari, S. Layec, K. Le-Roux, E. Maguin, A. Mérieux, R. M. Minardi, C. M'rini, J. Muller, R. Oozeer, J. Parkhill, P. Renault, M. Rescigno, N. Sanchez, S. Sunagawa, A. Torrejon, K. Turner, G. Vandemeulebrouck, E. Varela, Y. Winogradsky, G. Zeller, J. Weissenbach, S. D. Ehrlich and P. Bork (2011). "Enterotypes of the human gut microbiome." Nature 473(7346): 174-180.

Arunachalam, K., H. S. Gill and R. K. Chandra (2000). "Enhancement of natural immune function by dietary consumption of Bifidobacterium lactis (HN019)." European Journal of Clinical Nutrition 54(3): 263-267.

Arvola, T., K. Laiho, S. Torkkeli, H. Mykkänen, S. Salminen, L. Maunula and E. Isolauri (1999). "Prophylactic Lactobacillus GG reduces antibiotic-associated diarrhea in children with respiratory infections: a randomized stud." Pediatrics 104(5): e64.

Azcarate-Peril, M. A., R. Tallon and T. R. Klaenhammer (2009). "Temporal gene expression and probiotic attributes of Lactobacillus acidophilus during growth in milk." Journal of Dairy Science 92(3): 870-886.

Ba, Z., E. J. Furumoto and R. F. Roberts (2013). "Survival of Bifidobacterium animalis ssp. lactis BB-12 in yogurt drink is influenced by timing of probiotic addition." Journal of Dairy Science 96(E-Suppl. 1): 99.

Bäckhed, F., H. Ding, T. Wang, L. V. Hooper, G. Y. Koh, A. Nagy, C. F. Semenkovich and J. I. Gordon (2004). "The gut microbiota as an environmental factor that regulates fat storage." Proceedings of the National Academy of Sciences of the United States of America 101(44): 15718-15723.

Bäckhed, F., J. Roswall, Y. Peng, Q. Feng, H. Jia, P. Kovatcheva-Datchary, Y. Li, Y. Xia, H. Xie, H. Zhong, M. T. Khan, J. Zhang, J. Li, L. Xiao, J. Al-Aama, D. Zhang, Y. S.

158

Lee, D. Kotowska, C. Colding, V. Tremaroli, Y. Yin, S. Bergman, X. Xu, L. Madsen, K. Kristiansen, J. Dahlgren and W. Jun (2015). "Dynamics and stabilization of the human gut microbiome during the first year of life." Cell Host & Microbe 17(5): 690-703.

Barrangou, R., E. Altermann, R. Hutkins, R. Cano and T. R. Klaenhammer (2003). "Functional and comparative genomic analyses of an operon involved in fructooligosaccharide utilization by Lactobacillus acidophilus." Proceedings of the National Academy of Sciences of the United States of America 100(15): 8957-8962.

Barrangou, R., E. P. Briczinski, L. L. Traeger, J. R. Loquasto, M. Richards, P. Horvath, A. C. Coûté-Monvoisin, G. Leyer, S. Rendulic, J. L. Steele, J. R. Broadbent, T. Oberg, E. G. Dudley, S. Schuster, D. A. Romero and R. F. Roberts (2009). "Comparison of the complete genome sequences of Bifidobacterium animalis subsp. lactis DSM 10140 and Bl-04." Journal of Bacteriology 191(13): 4144-4151.

Barrett, B., R. L. Brown, M. P. Mundt, G. R. Thomas, S. K. Barlow, A. D. Highstrom and M. Bahrainian (2009). "Validation of a short form Wisconsin Upper Respiratory Symptom Survey (WURSS-21)." Health and Quality of Life Outcomes 7: 76.

Benson, A. K., S. A. Kelly, R. Legge, F. Ma, S. J. Low, J. Kim, M. Zhang, P. L. Oh, D. Nehrenberg, K. Hua, S. D. Kachman, E. N. Moriyama, J. Walter, D. A. Peterson and D. Pomp (2010). "Individuality in gut microbiota composition is a complex polygenic trait shaped by multiple environmental and host genetic factors." Proceedings of the National Academy of Sciences of the United States of America 107(44): 18933-18938.

Bisgaard, H., N. Li, K. Bonnelykke, B. L. Chawes, T. Skov, G. Paludan-Müller, J. Stokholm, B. Smith and K. A. Krogfelt (2011). "Reduced diversity of the intestinal microbiota during infancy is associated with increased risk of allergic disease at school age." The Journal of Allergy and Clinical Immunology 128(3): 646-652.

Black, F. T. (1996). "Placebo-controlled double-blind trial of 4 Lactobacilli strains (HIP) used as prophylactic agent against traveller’s diarrhea (2 trials)." Report by G. Nirnberger, Bioconsult, GmbH, Austria.

Blumberg, R. and F. Powrie (2012). "Microbiota, disease, and back to health: a metastable journey." Science Translational Medicine 4(137): 137rv137.

Bokulich, N. A., S. Subramanian, J. J. Faith, D. Gevers, J. I. Gordon, R. Knight, D. A. Mills and J. G. Caporaso (2013). "Quality-filtering vastly improves diversity estimates from Illumina amplicon sequencing." Nature Methods 10(1): 57-59.

Bokulich, N. A., J. H. Thorngate, P. M. Richardson and D. A. Mills (2014). "Microbial biogeography of wine grapes is conditioned by cultivar, vintage, and climate." Proceedings of the National Academy of Sciences of the United States of America 111(1): E139-E148.

159

Bollet, C., M. J. Gevaudan, X. d. Lamballerie, C. Zandotti and P. d. Micco (1991). "A simple method for the isolation of chromosomal DNA from gram positive or acid-fast bacteria." Nucleic Acids Research 19(8): 1955.

Bouchard, C., A. Tremblay, C. Leblanc, G. Lortie, R. Savard and G. Thériault (1983). "A method to assess energy expenditure in children and adults." The American Journal of Clinical Nutrition 37(3): 461-467.

Bouvier, M., S. Meance, C. Bouley, J. L. Berta and J. C. Grimaud (2001). "Effects of consumption of a milk fermented by the probiotic strain Bifidobacterium animalis DN- 173010 on colonic transit times in healthy humans." Bioscience Microflora 20(2): 43-48.

Bullen, C. L. and A. T. Willis (1971). "Resistance of the breast-fed infant to gastroenteritis." British Medical Journal 3: 338-343.

Burokas, A., R. D. Moloney, T. G. Dinan and J. F. Cryan (2015). "Microbiota regulation of the Mammalian gut-brain axis." Advances in Applied Microbiology 91: 1-62.

Burrows, M. P., P. Volchkov, K. S. Kobayashi and A. V. Chervonsky (2015). "Microbiota regulates type 1 diabetes through Toll-like receptors." Proceedings of the National Academy of Sciences of the United States of America 112(32): 9973-9977.

Cadwell, K., K. K. Patel, N. S. Maloney, T. C. Liu, A. C. Ng, C. E. Storer, R. D. Head, R. Xavier, T. S. Stappenbeck and H. W. Virgin (2010). "Virus-plus-susceptibility gene interaction determines Crohn's disease gene Atg16L1 phenotypes in intestine." Cell 141(7): 1135-1145.

Cai, Y., M. Matsumoto and Y. Benno (2000). "Bifidobacterium lactis Meile et al. 1997 is a subjective synonym of Bifidobacterium animalis (Mitsuoka 1969) Scardovi and Trovatelli 1974." Microbiology and Immunology 44(10): 815-820.

Canfora, E. E., J. W. Jocken and E. E. Blaak (2015). "Short-chain fatty acids in control of body weight and insulin sensitivity." Nature Review: Endocrinology 11(10): 577-591.

Caporaso, J. G., K. Bittinger, F. D. Bushman, T. Z. DeSantis, G. L. Andersen and R. Knight (2010). "PyNAST: A flexible tool for aligning sequences to a template alignment." Bioinformatics 26(2): 266-267.

Caporaso, J. G., J. Kuczynski, J. Stombaugh, K. Bittinger, F. D. Bushman, E. K. Costello, N. Fierer, A. G. Pena, J. K. Goodrich, J. I. Gordon, G. A. Huttley, S. T. Kelley, D. Knights, J. E. Koenig, R. E. Ley, C. A. Lozupone, D. McDonald, B. D. Muegge, M. Pirrung, J. Reeder, J. R. Sevinsky, P. J. Turnbaugh, W. A. Walters, J. Widmann, T. Yatsunenko, J. Zaneveld and R. Knight (2010). "QIIME allows analysis of high- throughput community sequencing data." Nature Methods 7(5): 335-336.

160

Caporaso, J. G., C. L. Lauber, W. A. Walters, D. Berg-Lyons, C. A. Lozupone, P. J. Turnbaugh, N. Fierer and R. Knight (2011). "Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample." Proceedings of the National Academy of Sciences of the United States of America 108(Suppl 1): 4516-4522.

Cardona, S., A. Eck, M. Cassellas, M. Gallart, C. Alastrue, J. Dore, F. Azpiroz, J. Roca, F. Guarner and C. Manichanh (2012). "Storage conditions of intestinal microbiota matter in metagenomic analysis." BMC Microbiology 12: 158.

Carroll, I. M., T. Ringel-Kulka, J. P. Siddle, T. R. Klaenhammer and Y. Ringel (2012). "Characterization of the fecal microbiota using high-throughput sequencing reveals a stable microbial community during storage." PLoS One 7(10): e46953.

Cassilly, D., S. Kantor, L. C. Knight, A. H. Maurer, R. S. Fisher, J. Semler and H. P. Parkman (2008). "Gastric emptying of a non-digestible solid: assessment with simultaneous SmartPill pH and pressure capsule, antroduodenal manometry, gastric emptying scintigraphy." Neurogastroenterology & Motility 20(4): 311-319.

Castro-Nallar, E., M. L. Bendall, M. Pérez-Losada, S. Sabuncyan, E. G. Severance, F. B. Dickerson, J. R. Schroeder, R. H. Yolken and K. A. Crandall (2015). "Composition, taxonomy and functional diversity of the oropharynx microbiome in individuals with schizophrenia and controls." PeerJ 3: e1140.

Chatterjee, S., P. Kar, T. Das, S. Ray, S. Gangulyt, C. Rajendiran and M. Mitra (2013). "Randomised placebo-controlled double blind multicentric trial on efficacy and safety of Lactobacillus acidophilus LA-5 and Bifidobacterium BB-12 for prevention of antibiotic- associated diarrhoea. JAPI 2013;61:708-712." The Journal of the Association of Physicians of India 61(10): 708-712.

Chen, R. M., J. J. Wu, S. C. Lee, A. H. Huang and H. M. Wu (1999). "Increase of intestinal Bifidobacterium and suppression of coliform bacteria with short-term yogurt ingestion." Journal of Dairy Science 82: 2308-2314.

Chouraqui, J. P., L. D. V. Egroo and M. C. Fichot (2004). "Acidified milk formula supplemented with Bifidobacterium lactis: Impact on infant diarrhea in residential care settings." Journal of Pediatric Gastroenterology and Nutrition 38(3): 288-292.

Chr-Hansen, A. S. (2005). "L. acidophilus, L. casei and bifidobacteria in fermented milk products - Guidelines." Technical Bulletin F-6(Methods for Counting Probiotic Bacteria): 2-8.

Christensen, H. R., C. N. Larsen, P. Kaestel, L. B. Rosholm, C. Sternberg, K. F. Michaelsen and H. Frokiaer (2006). "Immunomodulating potential of supplementation with probiotics: a dose-response study in healthy young adults." FEMS Immunology and Medical Microbiology 47(3): 380-390.

161

Clarke, K. R. (1993). "Non-parametric multivariate analyses of changes in community structure." Australian Journal of Ecology 18(1): 117-143.

Clausen, M. R., H. Bonnen, M. Tvede and P. B. Mortensen (1991). "Colonic fermentation to short-chain fatty acids is decreased in antibiotic-associated diarrhea." Gastroenterology 101(6): 1497-1504.

Clemente, J. C., L. K. Ursell, L. W. Parfrey and R. Knight (2012). "The impact of the gut microbiota on human health: an integrative view." Cell 148(6): 1258-1270.

Cole, J. R., Q. Wang, E. Cardenas, J. Fish, B. Chai, R. J. Farris, A. S. Kulam-Syed- Mohideen, D. M. McGarrell, T. Marsh, G. M. Garrity and J. M. Tiedje (2009). "The Ribosomal Database Project: improved alignments and new tools for rRNA analysis." Nucleic Acids Research 37(Database issue): D141-D145.

Collins, M. D. and G. R. Gibson (1999). "Probiotics, prebiotics, and synbiotics approaches for modulating the microbial ecology of the gut." The American Journal of Clinical Nutrition 69(5): 1052S-1057S.

Corrêa, N. B., L. A. P. Filho, F. J. Penna, F. M. Lima and J. R. Nicoli (2005). "A randomized formula controlled trial of Bifidobacterium lactis and Streptococcus thermophilus for prevention of antibiotic-associated diarrhea in infants." Journal of Clinical Gastroenterology 39(5): 385-389.

Cox, M. J., Y. J. Huang, K. E. Fujimura, J. T. Liu, M. KcKean, H. A. Boushey, M. R. Segal, E. L. Brodie, M. D. Cabana and S. V. Lynch (2010). "Lactobacillus casei abundance is associated with profound shifts in the infant gut microbiome." PLoS One 5(1): e8745.

Coyte, K. Z., J. Schluter and K. R. Foster (2015). "The ecology of the microbiome: Networks, competition, and stability." Science 350(6261): 663-666.

Curtis, T. P., W. T. Sloan and J. W. Scannell (2002). "Estimating prokaryotic diversity and its limits." Proceedings of the National Academy of Sciences of the United States of America 99(16): 10494-10499.

D'Aversa, F., A. Tortora, G. Ianiro, F. R. Ponziani, B. E. Annicchiarico and A. Gasbarrini (2013). "Gut microbiota and metabolic syndrome." Internal and Emergency Medicine 8(Suppl 1): S11-S15.

Damaskos, D. and G. Kolios (2008). "Probiotics and prebiotics in inflammatory bowel disease: Microflora 'on the scope'." British Journal of Clinical Pharmacology 65(4): 453- 467.

Darukaradhya, J., M. Phillips and K. Kailasapathy (2006). "Selective enumeration of Lactobacillus acidophilus, Bifidobacterium spp., starter lactic acid bacteria and non-

162 starter lactic acid bacteria from Cheddar cheese." International Dairy Journal 16(5): 439- 445.

Dave, R. I. and N. P. Shah (1996). "Evaluation of media for selective enumeration of Streptococcus thermophilus, Lactobacillus delbrueckii ssp. bulgaricus, Lactobacillus acidophilus, and bifidobacteria." Journal of Dairy Science 79(9): 1529-1536.

Dave, R. I. and N. P. Shah (1997). "Viability of yoghurt and probiotic bacteria in yoghurts made from commercial starter cultures." International Dairy Journal 7(1): 31- 41.

David, L. A., C. F. Maurice, R. N. Carmody, D. B. Gootenberg, J. E. Button, B. E. Wolfe, A. V. Ling, A. S. Devlin, Y. Varma, M. A. Fischbach, S. B. Biddinger, R. J. Dutton and P. J. Turnbaugh (2014). "Diet rapidly and reproducibly alters the human gut microbiome." Nature 505(7484): 559-563.

De Angelis, M., M. Piccolo, L. Vannini, S. Siragusa, A. D. Giacomo, D. I. Serrazzanetti, F. Cristofori, M. E. Guerzoni, M. Gobbetti and R. Francavilla (2013). "Fecal microbiota and metabolome of children with autism and pervasive developmental disorder not otherwise specified." PLoS One 8(10): e76993.

De Baerea, S., V. Eeckhaut, M. Steppe, C. D. Maesschalck, P. D. Backer, F. V. Immerseel and S. Croubels (2013). "Development of a HPLC-UV method for the quantitative determination of four short-chain fatty acids and lactic acid produced by intestinal bacteria during in vitro fermentation." Journal of Pharmaceutical and Biomedical Analysis 80: 107-115.

De Filippoa, C., D. Cavalieria, M. D. Paolab, M. Ramazzottic, J. B. Poulletd, S. Massartd, S. Collinib, G. Pieraccinie and P. Lionetti (2010). "Impact of diet in shaping gut microbiota revealed by a comparative study in children from Europe and rural Africa." Proceedings of the National Academy of Sciences of the United States of America 107(33): 14691–14696. de Man, J. D., M. Rogosa and M. E. Sharpe (1960). "A medium for the cultivation of Lactobacilli." Journal of Applied Microbiology 23(1): 130-135. de Vrese, M., H. Kristen, P. Rautenberg, C. Laue and J. Schrezenmeir (2011). "Probiotic lactobacilli and bifidobacteria in a fermented milk product with added fruit preparation reduce antibiotic associated diarrhea and Helicobacter pylori activity." Journal of Dairy Research 78(4): 396-403. de Vrese, M. and J. Schrezenmeir (2008). "Probiotics, prebiotics, and synbiotics." Advances in Biochemical Engineering/Biotechnology 111: 1-66.

Degen, L. P. and S. F. Phillips (1996). "Variability of gastrointestinal transit in healthy women and men." Gut 39(2): 299-305.

163

Deguchi, Y., T. Morishita and M. Mutai (1985). "Comparative studies on synthesis of water-soluble vitamins among human species of bifidobacteria." Agricultural and Biological Chemistry 49(1): 13-19.

Denne, S. C. (2015). "Neonatal nutrition." Pediatric Clinics of North America 62(2): 427- 438.

DeSantis, T. Z., P. Hugenholtz, N. Larsen, M. Rojas, E. L. Brodie, K. Keller, T. Huber, D. Dalevi, P. Hu and G. L. Andersen (2006). "Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB." Applied and Environmental Microbiology 72(7): 5069-5072.

Desfossés-Foucault, E., V. Dussault-Lepage, C. L. Boucher, P. Savard, G. Lapointe and D. Roy (2012). "Assessment of probiotic viability during cheddar cheese manufacture and ripening using propidium monoazide-PCR quantification." Frontiers in Microbiology 3: 350.

Dicksved, J., J. Halfvarson, M. Rosenquist, G. Järnerot, C. Tysk, J. Apajalahti, L. Engstrand and J. K. Jansson (2008). "Molecular analysis of the gut microbiota of identical twins with Crohn's disease." The ISME Journal 2(7): 716-727.

Dobell, C. (1920). "The discovery of the intestinal protozoa of man." Proceedings of the Royal Society of Medicine 13(Sect Hist Med): 1-15.

Dominguez-Bello, M. G., M. J. Blaser, R. E. Ley and R. Knight (2011). "Development of the human gastrointestinal microbiota and insights from high-throughput sequencing." Gastroenterology 140(6): 1713-1719.

Dominguez-Bello, M. G., E. K. Costello, M. Contreras, M. Magris, G. Hidalgo, N. Fierer and R. Knight (2010). "Delivery mode shapes the acquisition and structure of the initial microbiota across multiple body habitats in newborns." Proceedings of the National Academy of Sciences of the United States of America 107(26): 11971-11975.

Donovan, S. M. and R. Shamir (2014). "Introduction to the yogurt in nutrition initiative and the first global summit on the health effects of yogurt." The American Journal of Clinical Nutrition 99(Suppl 5): 1209S-1211S.

Dotterud, C. K., E. Avershina, M. Sekelja, M. R. Simpson, K. Rudi, O. Storro, R. Johnsen and T. Oien (2015). "Does maternal perinatal probiotic supplementation alter the intestinal microbiota of mother and child?" Journal of Pediatric Gastroenterology and Nutrition 61(2): 200-207.

Drossman, D. A. and D. L. Dumitrascu (2006). "Rome III: new standard for functional gastrointestinal disorders." Journal of Gastrointestinal and Liver Diseases 15(3): 237-241.

164

Eckburg, P. B., E. M. Bik, C. N. Bernstein, E. Purdom, L. Dethlefsen, M. Sargent, S. R. Gill, K. E. Nelson and D. A. Relman (2005). "Diversity of the human intestinal microbial flora." Science 308(5728): 1635-1638.

Edgar, R. C. (2010). "Search and clustering orders of magnitude faster than BLAST." Bioinformatics 26(19): 2460-2461.

Engelbrektson, A. L., J. R. Korzenik, M. E. Sanders, B. G. Clement, G. Leyer, T. R. Klaenhammer and C. L. Kitts (2006). "Analysis of treatment effects on the microbial ecology of the human intestine." FEMS Microbiology Ecology 57(2): 239-250.

Enomoto, T., M. Sowa, K. Nishimori, S. Shimazu, A. Yoshida, K. Yamada, F. Furukawa, T. Nakagawa, N. Yanagisawa, N. Iwabuchi, T. Odamaki, F. Abe, J. Nakayama and J. Z. Xiao (2014). "Effects of bifidobacterial supplementation to pregnant women and infants in the prevention of allergy development in infants and on fecal microbiota." Allergology International 63(4): 575-585.

Eskesen, D., L. Jespersen, B. Michelsen, P. J. Whorwell, S. Müller-Lissner and C. M. Morberg (2015). "Effect of the probiotic strain Bifidobacterium animalis subsp. lactis, BB-12®, on defecation frequency in healthy subjects with low defecation frequency and abdominal discomfort: a randomised, double-blind, placebo-controlled, parallel-group trial." British Journal of Nutrition 114(10): 1638-1646.

FAO/WHO (2002). "Report of a Joint FAO/WHO Working Group on Drafting Guidelines for the Evaluation of Probiotics in Food. London, Ontario."

Ferrario, C., V. Taverniti, C. Milani, W. Fiore, M. Laureati, I. D. Noni, M. Stuknyte, B. Chouaia, P. Riso and S. Guglielmetti (2014). "Modulation of fecal Clostridiales bacteria and butyrate by probiotic intervention with Lactobacillus paracasei DG varies among healthy adults." The Journal of Nutrition 144(11): 1787-1796.

Finegold, S. M., S. E. Dowd, V. Gontcharova, C. Liu, K. E. Henley, R. D. Wolcott, E. Youn, P. H. Summanen, D. Granpeesheh, D. Dixon, M. Liu, D. R. Molitoris and J. A. Green (2010). "Pyrosequencing study of fecal microflora of autistic and control children." Anaerobe 16(4): 444-453.

Fisberg, M. and R. Machado (2015). "History of yogurt and current patterns of consumption." Nutrition Reviews 73(Suppl 1): 4-7.

Fisher, L. D., D. O. Dixon, J. Herson, R. K. Frankowski, M. S. Hearon and K. E. Pearce (1990). Intention to treat in clinical trials. Statistical issues in drug research and development. K. E. Pearce. New York, Marcel Dekker: 331-350.

Fleming, S. E., D. Marthinsen and H. Kuhnlein (1983). "Colonic function and fermentation in men consuming high fiber diets." The Journal of Nutrition 113(12): 2535- 2544.

165

Fox, M. J., K. D. Ahuja, I. K. Robertson, M. J. Ball and R. D. Eri (2015). "Can probiotic yogurt prevent diarrhoea in children on antibiotics? A double-blind, randomised, placebo- controlled study." BMJ Open 5(1): e006474.

Fraher, M. H., P. W. O'Toole and E. M. Quigley (2012). "Techniques used to characterize the gut microbiota: a guide for the clinician." Nature Reviews: Gastroenterology & Hepatology 9(6): 312-322.

Fuentes, S., E. Nood, S. Tims, I. H. Jong, C. J. t. Braak, J. J. Keller, E. G. Zoetendal and W. M. Vos (2014). "Reset of a critically disturbed microbial ecosystem: faecal transplant in recurrent Clostridium difficile infection." The ISME Journal 8(8): 1621-1633.

Fukushima, Y., Y. Kawata, H. Hara, A. Terada and T. Mitsuoka (1998). "Effect of a probiotic formula on intestinal immunoglobulin A production in healthy children." International Journal of Food Microbiology 42(1-2): 39-44.

Fuller, R. (1989). "Probiotics in man and animals." The Journal of Applied Bacteriology 66(5): 365-378.

Garcia-Villalba, R., J. A. Gimenez-Bastida, M. T. GarcIa-Conesa, F. A. Tomas-Barberan and J. C. E. M. Larrosa (2012). "Alternative method for gas chromatography-mass spectrometry analysis of short-chain fatty acids in faecal samples." Journal of Separation Science 35(15): 1906-1913.

Garrigues, C., E. Johansen and M. B. Pedersen (2010). "Complete genome sequence of Bifidobacterium animalis subsp lactis BB-12, a widely consumed probiotic strain." Journal of Bacteriology 192(9): 2467-2468.

Garrigues, C., B. Stuer-Lauridsen and E. Johansen (2005). "Characterisation of Bifidobacterium animalis subsp. lactis BB-12 and other probiotic bacteria using genomics, transcriptomics and proteomics." Australian Journal of Dairy Technology 60(2): 84-92.

Gerritsen, J., H. Smidt, G. T. Rijkers and W. M. Vos (2011). "Intestinal microbiota in human health and disease: the impact of probiotics." Genes & Nutrition 6(3): 209-240.

Gianotti, L., L. Morelli, F. Galbiati, S. Rocchetti, S. Coppola, A. Beneduce, C. Gilardini, D. Zonenschain, A. Nespoli and M. Braga (2010). "A randomized double-blind trial on perioperative administration of probiotics in colorectal cancer patients." World Journal of Gastroenterology 16(2): 167-175.

Gilbert, J. A., J. K. Jansson and R. Knight (2014). "The earth microbiome project: successes and aspirations." BMC Biology 12: 69.

Gill, H. and J. Prasad (2008). "Probiotics, immunomodulation, and health benefits." Advances in Experimental Medicine and Biology 606: 423-454.

166

Giongo, A., K. A. Gano, D. B. Crabb, N. Mukherjee, L. L. Novelo, G. Casella, J. C. Drew, J. Ilonen, M. Knip, H. Hyöty, R. Veijola, T. Simell, O. Simell, J. Neu, C. H. Wasserfall, D. Schatz, M. A. Atkinson and E. W. Triplett (2011). "Toward defining the autoimmune microbiome for type 1 diabetes." The ISME Journal 5(1): 82-91.

Goodrich, J. K., S. C. D. Rienzi, A. C. Poole, O. Koren, W. A. Walters, J. G. Caporaso, R. Knight and R. E. Ley (2014). "Conducting a microbiome study." Cell 158(2): 250- 262.

Goodrich, J. K., J. L. Waters, A. C. Poole, J. L. Sutter, O. Koren, R. Blekhman, M. Beaumont, W. V. Treuren, R. Knight, J. T. Bell, T. D. Spector, A. G. Clark and R. E. Ley (2014). "Human genetics shape the gut microbiome." Cell 159(4): 789-799.

Granato, D., A. G. Branco, F. Nazzaro, A. G. Cruz and J. A. F. Faria (2010). "Functional foods and nondairy probiotic food development: trends, concepts, and products." Comprehensive Reviews in Food Science and Food Safety 9(3): 292-302.

Grand, M., M. Küffer and A. Baumgartner (2003). "Quantitative analysis and molecular identification of bifidobacteria strains in probiotic milk products." European Food Research and Technology 217(1): 90-92.

Groeger, D., L. O'Mahony, E. F. Murphy, J. F. Bourke, T. G. Dinan, B. Kiely, F. Shanahan and E. M. Quigley (2013). "Bifidobacterium infantis 35624 modulates host inflammatory processes beyond the gut." Gut Microbes 4(4): 325-339.

Guarner, F., A. G. Khan, J. Garisch, R. Eliakim, A. Gangl, A. Thomson, J. Krabshuis, T. L. Mair, P. Kaufmann, J. A. Paula, R. Fedorak, F. Shanahan, M. E. Sanders and H. Szajewska (2008). "Probiotics and prebiotics." World Gastroenterology Organisation Practice Guideline.

Gülden, E., F. S. Wong and L. Wen (2015). "The gut microbiota and Type 1 Diabetes." Clinical Immunology: doi: 10.1016/j.clim.2015.1005.1013.

Guyonnet, D., O. Chassany, P. Ducrotte, C. Picard, M. Mouret, C. H. Mercier and C. Matuchansky (2007). "Effect of a fermented milk containing Bifidobacterium animalis DN-173 010 on the health-related quality of life and symptoms in irritable bowel syndrome in adults in primary care: a multicentre, randomized, double-blind, controlled trial." Alimentary Pharmacology & Therapeutics 26(3): 475-486.

Guyonnet, D., A. Schlumberger, L. Mhamdi, S. Jakob and O. Chassany (2009). "Fermented milk containing Bifidobacterium lactis DN-173 010 improves gastrointestinal well-being and digestive symptoms in women reporting minor digestive symptoms: a randomised, double-blind, parallel, controlled study." British Journal of Nutrition 102(11): 1654-1662.

167

Guyonnet, D., A. Woodcock, B. Stefani, C. Trevisan and C. Hall (2009). "Fermented milk containing Bifidobacterium lactis DN-173 010 improved self-reported digestive comfort amongst a general population of adults: A randomized, open-label, controlled, pilot study." Journal of Digestive Diseases 10(1): 61-70.

Haas, B. J., D. Gevers, A. M. Earl, M. Feldgarden, D. V. Ward, G. Giannoukos, D. Ciulla, D. Tabbaa, S. K. Highlander, E. Sodergren, B. Methé, T. Z. DeSantis, Human Microbiome Consortium, J. F. Petrosino, R. Knight and B. W. Birren (2011). "Chimeric 16S rRNA sequence formation and detection in Sanger and 454-pyrosequenced PCR amplicons." Genome Research 21(3): 494-504.

Hasler, C. M., A. C. Brown and A. D. Association (2009). "Position of the American Dietetic Association: Functional foods." Journal of the American Dietetic Association 109(4): 735-746.

Heenana, C. N., M. C. Adamsa, R. W. Hoskena and G. H. Fleetb (2004). "Survival and sensory acceptability of probiotic microorganisms in a non-fermented frozen vegetarian dessert." LWT-Food Science and Technology 37(4): 461-466.

Heller, K. J. (2001). "Probiotic bacteria in fermented foods: product characteristics and starter organisms." The American Journal of Clinical Nutrition 73(2 Suppl): 374S-379S.

Higgs, R. H., R. B. Ellis-Pegler and H. P. Lambert (1975). "Assessment of simple methods of measuring intestinal transit times in children with gastroenteritis." Gut 16(6): 458-461.

Hill, C., F. Guarner, G. Reid, G. R. Gibson, D. J. Merenstein, B. Pot, L. Morelli, R. B. Canani, H. J. Flint, S. Salminen, P. C. Calde and M. E. Sanders (2014). "Expert consensus document. The International Scientific Association for Probiotics and Prebiotics consensus statement on the scope and appropriate use of the term probiotic." Nature Review: Gastroenterology & Hepatology 11(8): 506-514.

Hinton, J. M., J. E. Lennard-Jones and A. C. Young (1969). "A new method for studying gut transit times using radioopaque markers." Gut 10(10): 842-847.

Hojsak, I., A. M. Pavić, T. Kos, J. Dumančić and S. Kolaček (2015). "Bifidobacterium animalis subsp. lactis in prevention of common infections in healthy children attending day care centers: randomized, double blind, placebo-controlled study." Clinical Nutrition: doi:10.1016/j.clnu.2015.1005.1004.

Hol, J., E. H. Leer, B. E. E. Schuurman, L. F. Ruiter, J. N. Samsom, W. Hop, H. J. Neijens, J. C. Jongste and E. E. Nieuwenhuis (2008). "The acquisition of tolerance toward cow's milk through probiotic supplementation: a randomized, controlled trial." Journal of Allergy and Clinical Immunology 121(6): 1448-1454.

168

Hong, Y. S., Y. T. Ahn, J. C. Park, J. H. Lee, H. Lee, C. S. Huh, D. H. Kim, H. Ryu-do and G. S. Hwang (2010). "1H NMR-based metabonomics assessment of probiotic effects in a colitis mouse model." Archives of Pharmacal Research 33(7): 1091-1101.

Huda, M. N., Z. T. Lewis, K. M. Kalanetra, M. Rashid, S. M. Ahmad, R. Raqib, F. Qadri, M. A. Underwood, D. A. Mills and C. B. Stephensen (2014). "Stool microbiota and vaccine responses of infants." Pediatrics 134(2): e362-e372.

Human Microbiome Project Consortium (2012). "A framework for human microbiome research." Nature 486(7402): 215-221.

Human Microbiome Project Consortium (2012). "Structure, function and diversity of the healthy human microbiome." Nature 486(7402): 207-214.

Ishikawa, H., S. Matsumoto, Y. Ohashi, A. Imaoka, H. Setoyama, Y. Umesaki, R. Tanaka and T. Otani (2011). "Beneficial effects of probiotic Bifidobacterium and galacto- oligosaccharide in patients with ulcerative colitis: a randomized controlled study." Digestion 84(2): 128-133.

Isolauri, E., T. Arvola, Y. Sütas, E. Moilanen and S. Salminen (2000). "Probiotics in the management of atopic eczema." Clinical and Experimental Allergy 30(11): 1604-1610.

Jacques, P. F. and H. Wang (2014). "Yogurt and weight management." The American Journal of Clinical Nutrition 99(Suppl S): 1229S-1234S.

Jungersen, M., A. Wind, E. Johansen, J. E. Christensen, B. Stuer-Lauridsen and D. Eskesen (2014). "The science behind the probiotic strain Bifidobacterium animalis subsp. lactis BB-12®." Microorganisms 2(2): 92-110.

Kalliomäki, M., S. Salminen, H. Arvilommi, P. Kero, P. Koskinen and E. Isolauri (2001). "Probiotics in primary prevention of atopic disease: a randomised placebo-controlled trial." Lancet 357(9262): 1076-1079.

Kashyap, P. C., A. Marcobal, L. K. Ursell, M. Larauche, H. Duboc, K. A. Earle, E. D. Sonnenburg, J. A. Ferreyra, S. K. Higginbottom, M. Mulugeta, Y. Tache, P. J. Pasricha, R. Knight, G. Farrugia and J. L. Sonnenburg (2013). "Complex interactions among diet, gastrointestinal transit, and gut microbiota in humanized mice." Gastroenterology 144(5): 967-977.

Kekkonen, R. A., N. Lummela, H. Karjalainen, S. Latvala, S. Tynkkynen, S. Jarvenpaa, H. Kautiainen, I. Julkunen, H. Vapaatalo and R. Korpela (2008). "Probiotic intervention has strain-specific anti-inflammatory effects in healthy adults." World Journal of Gastroenterology 14(13): 2029-2036.

169

Kim, M. S., S. W. Roh and J. W. Bae (2010). "Bifidobacterium stercoris sp. nov., isolated from human faeces." International Journal of Systematic Evolutionary Microbiology 60(Pt 12): 2823-2827.

Kirchberger, S., O. Majdic and J. Stockl (2007). "Modulation of the immune system by human rhinoviruses." International Archieves of Allergy and Immunology 142(1): 1-10.

Kirjavainen, P. V., T. Arvola, S. J. Salminen and E. Isolauri (2002). "Aberrant composition of gut microbiota of allergic infants: a target of bifidobacterial therapy at weaning?" Gut 51(1): 51-55.

Klein, A., U. Friedrich, H. Vogelsang and G. Jahreis (2008). "Lactobacillus acidophilus 74-2 and Bifidobacterium animalis subsp lactis DGCC 420 modulate unspecific cellular immune response in healthy adults." European Journal of Clinical Nutrition 62(5): 584- 593.

Klijn, A., A. Mercenier and F. Arigoni (2005). "Lessons from the genomes of bifidobacteria." FEMS Microbiology Reviews 29(3): 491-509.

Koeppel, A. F. and M. Wu (2013). "Surprisingly extensive mixed phylogenetic and ecological signals among bacterial operational taxonomic units." Nucleic Acids Research 14(10): 5175-5188.

Kondo, J., J. Z. Xiao, A. Shirahata, M. Baba, A. Abe, K. Ogawa and T. Shimoda (2013). "Modulatory effects of Bifidobacterium longum BB536 on defecation in elderly patients receiving enteral feeding." World Journal of Gastroenterology 19(14): 2162-2170.

Krevsky, B., L. S. Malmud, F. D'Ercole, A. H. Maurer and R. S. Fisher (1986). "Colonic transit scintigraphy. A physiologic approach to the quantitative measurement of colonic transit in humans." Gastroenterology 91(5): 1102-1112.

Kuczynski, J., C. L. Lauber, W. A. Walters, L. W. Parfrey, J. C. Clemente, D. Gevers and R. Knight (2012). "Experimental and analytical tools for studying the human microbiome." Nature Reviews: Genetics 13(1): 47-58.

Kuczynski, J., J. Stombaugh, W. A. Walters, A. Gonzalez, J. G. Caporaso and R. Knight (2011). "Unit 10.7: Using QIIME to analyze 16S rRNA gene sequences from microbial communities." Current Protocols in Bioinformatics.

Kullen, M. J., M. M. Amann, M. J. O'Shaughnessy, D. J. O'Sullivan, F. F. Busta and L. J. Brady (1997). "Differentiation of ingested and endogenous bifidobacteria by DNA fingerprinting demonstrates the survival of an unmodified strain in the gastrointestinal tract of humans." The Journal of Nutrition 127(1): 89-94.

Kunde, S., A. Pham, S. Bonczyk, T. Crumb, M. Duba, H. Conrad, D. Cloney and S. Kugathasan (2013). "Safety, tolerability, and clinical response after fecal transplantation

170 in children and young adults with ulcerative colitis." Journal of Pediatric Gastroenterology & Nutrition 56(6): 597-601.

Laake, K. O., A. Bjørneklett, A. Bakka, T. Midtvedt, K. E. Norin, T. J. Eide, M. B. Jacobsen, E. Lingaas, A. K. Axelsen, T. Løtveit and M. H. Vatn (1999). "Influence of fermented milk on clinical state, fecal bacterial counts and biochemical characteristics in patients with ileal- pouch- anal-anastomosis." Microbial Ecology in Health and Disease 11(4): 211-217.

Laitinen, K., T. Poussa, E. Isolauri and A. Nutrition, Mucosal Immunology and Intestinal Microbiota Group (2009). "Probiotics and dietary counselling contribute to glucose regulation during and after pregnancy: a randomised controlled trial." British Journal of Nutrition 101(11): 1679-1687.

Langhendries, J. P., J. Detry, J. V. Hees, J. M. Lamboray, J. Darimont, M. J. Mozin, M. C. Secretin and J. Senterre (1995). "Effect of a fermented infant formula containing viable bifidobacteria on the fecal flora composition and pH of healthy full-term infants." Journal of Pediatric Gastroenterology and Nutrition 21(2): 177-181.

Lapierre, L., P. Undeland and L. J. Cox (1992). "Lithium chloride-sodium propionate agar for the enumeration of bifidobacteria in fermented dairy products." Journal of Dairy Science 75(5): 1192-1196.

Laroia, S. and J. H. Martin (1991). "Methods for enumerating and propagating bifidobacteria." Cultured Dairy Products Journal 26(2): 32-34.

Larsen, C. N., S. Nielsen, P. Kaestel, E. Brockmann, M. Bennedsen, H. R. Christensen, D. C. Eskesen, B. L. Jacobsen and K. F. Michaelsen (2006). "Dose-response study of probiotic bacteria Bifidobacterium animalis subsp. lactis BB-12 and Lactobacillus paracasei subsp. paracasei CRL-341 in healthy young adults." European Journal of Clinical Nutrition 60(11): 1284-1293.

Latvala, S., T. E. Pietila, V. Veckman, R. A. Kekkonen, S. Tynkkynen, R. Korpela and I. Julkunen (2008). "Potentially probiotic bacteria induce efficient maturation but differential cytokine production in human monocyte-derived dendritic cells." World Journal of Gastroenterology 14(36): 5570-5583.

Lee, B., X. Yin, S. M. Griffey and M. L. Marco (2015). "Attenuation of colitis by Lactobacillus casei BL23 is dependent on the dairy delivery matrix." Applied and Environmental Microbiology: doi: 10.1128/AEM.01360-01315.

Lee, J. H. and D. J. O'Sullivan (2010). "Genomic insights into bifidobacteria." Microbiology and Molecular Biology Reviews 74(3): 378-416.

Lesniewska, V., I. Rowland, P. D. Cani, A. M. Neyrinck, N. M. Delzenne and P. J. Naughton (2006). "Effect on components of the intestinal microflora and plasma

171 ueuropeptide levels of feeding Lactobacillus delbrueckii, Bifidobacterium lactis, and inulin to adult and elderly rats." Applied and Environmental Microbiology 72(10): 6533- 6538.

Lewis, S. J. and K. W. Heaton (1997). "Increasing butyrate concentration in the distal colon by accelerating intestinal transit." Gut 41(2): 245-251.

Lewis, S. J. and K. W. Heaton (1999). "The metabolic consequences of slow colonic transit." The American Journal of Gastroenterology 94(8): 2010-2016.

Lewis, Z. T., N. A. Bokulich, K. M. Kalanetra, S. Ruiz-Moyano, M. A. Underwood and D. A. Mills (2013). "Use of bifidobacterial specific terminal restriction fragment length polymorphisms to complement next generation sequence profiling of infant gut communities." Anaerobe 19: 62-69.

Lewis, Z. T., S. M. Totten, J. T. Smilowitz, M. Popovic, E. Parker, D. G. Lemay, M. L. V. Tassell, M. J. Miller, Y. S. Jin, J. B. German, C. B. Lebrilla and D. A. Mills (2015). "Maternal fucosyltransferase 2 status affects the gut bifidobacterial communities of breastfed infants." Microbiome 3: 13.

Ley, R. E., F. Backhed, P. Turnbaugh, C. A. Lozupone, R. D. Knight and J. I. Gordon (2005). "Obesity alters gut microbial ecology." Proceedings of the National Academy of Sciences of the United States of America 102(31): 11070-11075.

Ley, R. E., P. J. Turnbaugh, S. Klein and J. I. Gordon (2006). "Human gut microbes associated with obesity." Nature 444: 1022-1023.

Lilly, D. M. and R. H. Stillwell (1965). "Probiotics: Growth-promoting factors produced by microorganisms." Science 147(No. 3659): 747-748.

Lin, A., E. M. Bik, E. K. Costello, L. Dethlefsen, R. Haque, D. A. Relman and U. Singh (2013). "Distinct distal gut microbiome diversity and composition in healthy children from Bangladesh and the United States." PLoS One 8(1): e53838.

Lin, H. C., B. H. Su, A. C. Chen, T. W. Lin, C. H. Tsai, T. F. Yeh and W. Oh (2005). "Oral probiotics reduce the incidence and severity of necrotizing enterocolitis in very low birth weight infants." Pediatrics 115(1): 1-4.

Link-Amster, H., F. Rochat, K. Y. Saudan, O. Mignot and J. M. Aeschlimann (1994). "Modulation of a specific humoral immune response and changes in intestinal flora mediated through fermented milk intake." FEMS Immunology and Medical Microbiology 10(1): 55-63.

Lomonaco, S., E. J. Furumoto, J. R. Loquasto, P. Morra, A. Grassi and R. F. Roberts (2015). "Development of a rapid SNP-typing assay to differentiate Bifidobacterium

172 animalis ssp. lactis strains used in probiotic-supplemented dairy products." Journal of Dairy Science 98(2): 804-812.

López, P., M. Gueimonde, A. Margolles and A. Suárez (2010). "Distinct Bifidobacterium strains drive different immune responses in vitro." International Journal of Food Microbiology 138(1-2): 157-165.

Lozupone, C., M. Hamady and R. Knight (2006). "UniFrac: an online tool for comparing microbial community diversity in a phylogenetic context." BMC Bioinformatics 7: 371.

Lozupone, C. A., J. I. Stombaugh, J. I. Gordon, J. K. Jansson and R. Knight (2012). "Diversity, stability and resilience of the human gut microbiota." Nature 489(7415): 220- 230.

Lu, W. Z., G. H. Song, K. A. Gwee and K. Y. Ho (2009). "The effects of melatonin on colonic transit time in normal controls and IBS patients." Digestive Diseases and Sciences 54(5): 1087-1093.

Luoto, R., K. Laitinen, M. Nermes and E. Isolauri (2010). "Impact of maternal probiotic- supplemented dietary counselling on pregnancy outcome and prenatal and postnatal growth: a double-blind, placebo-controlled study." British Journal of Nutrition 103(12): 1792-1799.

Maccaferri, S., M. Candela, S. Turroni, M. Centanni, M. Severgnini, C. Consolandi, P. Cavina and P. Brigidi (2012). "IBS-associated phylogenetic unbalances of the intestinal microbiota are not reverted by probiotic supplementation." Gut Microbes 3(5): 406-413.

Macfarlane, G. T. and S. Macfarlane (2012). "Bacteria, colonic fermentation, and gastrointestinal health." Journal of AOAC International 95(1): 50-60.

Mackie, R. I., A. Sghir and H. R. Gaskins (1999). "Developmental microbial ecology of the neonatal gastrointestinal tract." The American Journal of Clinical Nutrition 69(5): 1035S-1045S.

Mackowiak, P. A. (2013). "Recycling Metchnikoff: probiotics, the intestinal microbiome and the quest for long life." Frontiers in Public Health 1(52): 1-3.

Mariat, D., O. Firmesse, F. Levenez, V. D. Guimarăes, H. Sokol, J. Doré, G. Corthier and J.-P. Furet (2009). "The Firmicutes/Bacteroidetes ratio of the human microbiota changes with age." BMC Microbiology 9: 123. marketsandmarkets.com (2015). Probiotic ingredients market by function (regular, preventative, therapy), application (food & beverage, dietary supplements, & animal feed), end use (human & animal probiotics), ingredient (bacteria & yeast), and by region - global trends & forecast to 2020, Accessed on November 19th, 2015.

173

Marsland, B. J. and O. Salami (2015). "Microbiome influences on allergy in mice and humans." Current Opinion in Immunology 36: 94-100.

Marteau, P., E. Cuillerier, S. Meance, M. F. Gerhardt, A. Myara, M. Bouvier, C. Bouley, F. Tondu, G. Bommelaer and J. C. Grimaud (2002). "Bifidobacterium animalis strain DN-173 010 shortens the colonic transit time in healthy women: a double-blind, randomized, controlled study." Alimentary Pharmacology Therapeutics 16(3): 587-593.

Martensson, O., R. Oste and O. Holst (2002). "The effect of yoghurt culture on the survival of probiotic bacteria in oat-based, non-dairy products." Food Research International 35(8): 775-784.

Martin, F. P., Y. Wang, N. Sprenger, I. K. Yap, T. Lundstedt, P. Lek, S. Rezzi, Z. Ramadan, P. Bladeren, L. B. Fay, S. Kochhar, J. C. Lindon, E. Holmes and J. K. Nicholson (2008). "Probiotic modulation of symbiotic gut microbial-host metabolic interactions in a humanized microbiome mouse model." Molecular Systems Biology 4: 157.

Martinez, F. A., E. M. Balciunas, A. Converti, P. D. Cotter and R. P. S. Oliveira (2013). "Bacteriocin production by Bifidobacterium spp.: A review." Biotechnology Advances 31(4): 482-488.

Masco, L., M. Ventura, R. Zink, G. Huys and J. Swings (2004). "Polyphasic taxonomic analysis of Bifidobacterium animalis and Bifidobacterium lactis reveals relatedness at the subspecies level: reclassification of Bifidobacterium animalis as Bifidobacterium animalis subsp. animalis subsp. nov. and Bifidobacterium lactis as Bifidobacterium animalis subsp. lactis subsp. nov." International Journal of Systematic and Evolutionary Microbiology 54(Pt 4): 1137-1143.

Maslowski, K. M., A. T. Vieira, A. Ng, J. Kranich, F. Sierro, D. Yu, H. C. Schilter, M. S. Rolph, F. Mackay, D. Artis, R. J. Xavier, M. M. Teixeira and C. R. Mackay (2009). "Regulation of inflammatory responses by gut microbiota and chemoattractant receptor GPR43." Nature 461(7268): 1282-1286.

Mastrandrea, F., G. Coradduzza, G. Serio, A. Minardi, M. Manelli, S. Ardito and L. Muratore (2004). "Probiotics reduce the CD34+ hemopoietic precursor cell increased traffic in allergic subjects." European annals of allergy and clinical immunology 36(4): 118-122.

Matsumoto, M., A. Aranami, A. Ishige, K. Watanabe and Y. Benno (2007). "LKM512 yogurt consumption improves the intestinal environment and induces the T-helper type 1 cytokine in adult patients with intractable atopic dermatitis." Clinical and Experimental Allergy 37(3): 358-370.

174

Matsumoto, M., H. Ohishi and Y. Benno (2001). "Impact of LKM512 yogurt on improvement of intestinal environment of the elderly." FEMS Immunology and Medical Microbiology 31(3): 181-186.

McDonald, D., M. N. Price, J. Goodrich, E. P. Nawrocki, T. Z. DeSantis, A. Probst, G. L. Andersen, R. Knight and P. Hugenholtz (2012). "An improved Greengenes taxonomy with explicit ranks for ecological and evolutionary analyses of bacteria and archaea." The ISME Journal 6(3): 610-618.

McNulty, N. P., T. Yatsunenko, A. Hsiao, J. J. Faith, B. D. Muegge, A. L. Goodman, B. Henrissat, R. Oozeer, S. Cools-Portier, G. Gobert, C. Chervaux, D. Knights, C. A. Lozupone, R. Knight, A. E. Duncan, J. R. Bain, M. J. Muehlbauer, C. B. Newgard, A. C. Heath and J. I. Gordon (2011). "The impact of a consortium of fermented milk strains on the gut microbiome of gnotobiotic mice and monozygotic twins." Science Translational Medicine 3(106): 106ra106.

McOrist, A. L., G. C. Abell, C. Cooke and K. Nyland (2008). "Bacterial population dynamics and faecal short-chain fatty acid (SCFA) concentrations in healthy humans." British Journal of Nutrition 100(1): 138-146.

McOrist, A. L., R. B. Miller, A. R. Bird, J. B. Keogh, M. Noakes, D. L. Topping and M. A. Conlon (2011). "Fecal butyrate levels vary widely among individuals but are usually increased by a diet high in resistant starch." The Journal of Nutrition 141(5): 883-889.

Medline Plus Medical Encyclopedia (2014). "Bowel transit time U.S. National Library of Medicine." National Institutes of Health(Online, accessed on Nov. 20th, 2015).

Meile, L., W. Ludwig, U. Rueger, C. Gut, P. Kaufmann, G. Dasen, S. Wenger and M. Teuber (1997). "Bifidobacterium lactis sp. nov., a moderately oxygen tolerant species isolated from fermented milk." Systematic and Applied Microbiology 20(1): 57-64.

Mende, D. R., A. S. Waller, S. Sunagawa, A. I. Järvelin, M. M. Chan, M. Arumugam, J. Raes and P. Bork (2012). "Assessment of metagenomic assembly using simulated next generation sequencing data." PLoS One 7(2): e31386.

Meng, H., Z. Ba, Y. Lee, J. Peng, J. Lin, J. A. Fleming, E. J. Furumoto, R. F. Roberts, P. M. Kris-Etherton and C. J. Rogers (2015). "Consumption of Bifidobacterium animalis subsp. lactis BB-12 in yogurt reduced expression of TLR-2 on peripheral blood-derived monocytes and pro-inflammatory cytokine secretion in young adults." European Journal of Nutrition(Accepted).

Meng, H., Y. Lee, Z. Ba, J. A. Fleming, E. J. Furumoto, R. F. Roberts, P. M. Kris- Etherton and C. J. Rogers (2015). "In vitro production of IL-6 and IFN-γ is influenced by dietary variables and predicts upper respiratory tract infection incidence and severity respectively in young adults." Frontiers in Immunology 6: 94.

175

Merenstein, D., J. Gonzalez, A. G. Young, R. F. Roberts, M. E. Sanders and S. Petterson (2011). "Study to investigate the potential of probiotics in children attending school." European Journal of Clinical Nutrition 65(4): 447-453.

Merenstein, D. J., K. H. Smith, M. Scriven, R. F. Roberts, M. E. Sanders and S. Petterson (2010). "The study to investigate the potential benefits of probiotics in yogurt, a patient- oriented, double-blind, cluster-randomised, placebo-controlled, clinical trial." European Journal of Clinical Nutrition 64(7): 685-691.

Merenstein, D. J., T. P. Tan, A. Molokin, K. H. Smith, R. F. Roberts, N. M. Shara, M. Mete, M. E. Sanders and G. Solano-Aguilar (2015). "Safety of Bifidobacterium animalis subsp. lactis (B. lactis) strain BB-12-supplemented yogurt in healthy adults on antibiotics: a phase I safety study." Gut Microbes 6(1): 66-77.

Metchnikoff, E. (1908). The prolongation of life: optimistic studies, New York & London: G. P. Putnam's Sons, The Knickerbocker Press.

Million, M., M. Maraninchi, M. Henry, F. Armougom, H. Richet, P. Carrieri, R. Valero, D. Raccah, B. Vialettes and D. Raoult (2012). "Obesity-associated gut microbiota is enriched in Lactobacillus reuteri and depleted in Bifidobacterium animalis and Methanobrevibacter smithii." International Journal of Obesity 36(6): 817-825.

Miranda, R. O., G. G. Neto, R. Freitas, A. F. Carvalho and L. A. Nero (2011). "Enumeration of bifidobacteria using Petrifilm AC in pure cultures and in a fermented milk manufactured with a commercial culture of Streptococcus thermophilus." Food Microbiology 28(8): 1509-1513.

Mohamadshahi, M., M. Veissi, F. Haidari, A. Z. Javid, F. Mohammadi and E. Shirbeigi (2014). "Effects of probiotic yogurt consumption on lipid profile in type 2 diabetic patients: A randomized controlled clinical trial." Journal of Research in Medical Sciences 19(6): 531-536.

Mohan, R., C. Koebnick, J. Schildt, S. Schmidt, M. Mueller, M. Possner, M. Radke and M. Blaut (2006). "Effects of Bifidobacterium lactis BB-12 supplementation on intestinal microbiota of preterm infants: a double-blind placebo controlled randomized study." Jounrnal of Clinical Microbiology 44(11): 4025-4031.

Mueller, S., K. Saunier, C. Hanisch, E. Norin, L. Alm, T. Midtvedt, A. Cresci, S. Silvi, C. Orpianesi, M. C. Verdenelli, T. Clavel, C. Koebnick, H. J. Zunft, J. Doré and M. Blaut (2006). "Differences in fecal microbiota in different European study populations in relation to age, gender, and country: a cross-sectional study." Applied and Environmental Microbiology 72(2): 1027-1033.

Namba, K., M. Hatano, T. Yaeshima, M. Takase and K. Suzuki (2010). "Effects of Bifidobacterium longum BB536 administration on influenza infection, influenza vaccine

176 antibody titer, and cell-mediated immunity in the elderly." Bioscience, Biotechnology, and Biochemistry 74(5): 939-945.

Nilsson, A. I., S. Koskiniemi, S. Eriksson, E. Kugelberg, J. C. D. Hinton and D. I. Andersson (2005). "Bacterial genome size reduction by experimental evolution." Proceedings of the National Academy of Sciences of the United States of America 102(34): 12112-12116.

Nishida, S., M. Gotou, S. Akutsu, M. Ono, Y. Hitomi, T. Nakamura and H. Iino (2004). "Effect of yoghurt containing Bifidobacterium lactis BB-12 on improvement of defecation and fecal microflora of healthy female adults." Milk Science 53(2): 71-80.

Nishida, S., M. Gotou, S. Akutsu, M. Ono, Y. Hitomi, T. Nakamura and H. Iino (2004). "Improvement effect of yogurt containing Bifidobacterium lactis strain BB-12 on defecation and fecal microflora and its safety to healthy human adults." Milk Science 53(3): 133-140.

O'Mahony, L., J. McCarthy, P. Kelly, G. Hurley, F. Luo, K. Chen, G. C. O'Sullivan, B. Kiely, J. K. Collins, F. Shanahan and E. M. Quigley (2005). "Lactobacillus and Bifidobacterium in irritable bowel syndrome: symptom responses and relationship to cytokine profiles." Gastroenterology 128(3): 541-551.

Oliveira, R. P. S., P. Perego, M. N. d. Oliveira and A. Converti (2012). "Growth, organic acids profile and sugar metabolism of Bifidobacterium lactis in co-culture with Streptococcus thermophilus: The inulin effect." Food Research International 48(2012): 21-27.

Ouwehand, A. C., N. Bergsma, R. Parhiala, S. Lahtinen, M. Gueimonde, F. inne-Soveri, T. Strandberg, K. Pitkala and S. Salminen (2008). "Bifidobacterium microbiota and parameters of immune function in elderly subjects." FEMS Immunology and Medical Microbiology 53(1): 18-25.

Ouwehand, A. C., M. Saxelin and S. Salminen (2004). "Assessment of potential risk factors and related properties of clinical, faecal and dairy Bifidobacterium isolates." Bioscience Microflora 23(1): 37-42.

Palaria, A., I. Johnson-Kanda and D. J. O'Sullivan (2012). "Effect of a synbiotic yogurt on levels of fecal bifidobacteria, clostridia, and enterobacteria." Applied and Environmental Microbiology 78(4): 933-940.

Palmer, C., E. M. Bik, D. B. DiGiulio, D. A. Relman and P. O. Brown (2007). "Development of the human infant intestinal microbiota." PLoS Biology 5(7): e177.

Pelletier, S., S. Kundrat and C. M. Hasler (2002). "Effects of an educational program on intent to consume functional foods." Journal of the American Dietetic Association 102(9): 1297-1300.

177

Pelletier, S., S. Kundrat and C. M. Hasler (2003). "Effects of a functional foods nutrition education program with cardiac rehabilitation patients." Journal of Cardiopulmonary Rehabilitation 23(5): 334-340.

Penders, J., C. Thijs, P. A. Brandt, I. Kummeling, B. Snijders, F. Stelma, H. Adams, R. Ree and E. E. Stobberingh (2007). "Gut microbiota composition and development of atopic manifestations in infancy: the KOALA Birth Cohort Study." Gut 56(5): 661-667.

Penders, J., C. Thijs, C. Vink, F. F. Stelma, B. Snijders, I. Kummeling, P. A. Brandt and E. E. Stobberingh (2006). "Factors influencing the composition of the intestinal microbiota in early infancy." Pediatrics 118(2): 511-521.

Pitkala, K. H., T. E. Strandberg, U. H. Finne-Soveri, A. C. Ouwehand, T. Poussa and S. Salminen (2007). "Fermented cereal with specific bifidobacteria normalizes bowel movements in elderly nursing home residents. A randomized, controlled trial." Journal of Nutritional Health and Aging 11(4): 305-311.

Pluznick, J. (2014). "A novel SCFA receptor, the microbiota, and blood pressure regulation." Gut Microbes 5(2): 202-207.

Poupard, J. A., I. Husain and R. F. Norris (1973). "Biology of the bifidobacteria." Bacteriological Review 37(2): 136-165.

Price, M. N., P. S. Dehal and A. P. Arkin (2010). "FastTree 2—approximately maximum- likelihood trees for large alignments." PLoS One 5(3): e9490.

Qin, J., R. Li, J. Raes, M. Arumugam, K. S. Burgdorf, C. Manichanh, T. Nielsen, N.Pons, F. Levenez, T. Yamada, D. R. Mende, J. Li, J. Xu, S. Li, D. Li, J. Cao, B. Wang, H. Liang, H. Zheng, Y. Xie, J. Tap, P. Lepage, M. Bertalan, J.-M. Batto, T. Hansen, D. L. Paslier, A. Linneberg, H. B. Nielsen, E. Pelletier, P. Renault, T. Sicheritz-Ponten, K. Turner, H. Zhu, C. Yu, S. Li, M. Jian, Y. Zhou, Y. Li, X. Zhang, S. Li, N. Qin, H. Yang, J. Wang, S. Brunak, J. Doré, F. Guarner, K. Kristiansen, O. Pedersen, J. Parkhill, J. Weissenbach, M. Consortium, P. Bork, S. D. Ehrlich and J. Wang (2010). "A human gut microbial gene catalogue established by metagenomic sequencing." Nature 464: 59-65.

Quast, C., E. Pruesse, P. Yilmaz, J. Gerken, T. Schweer, P. Yarza, J. Peplies and F. O. Glöckner (2013). "The SILVA ribosomal RNA gene database project: improved data processing and web-based tools." Nucleic Acids Research 41(Database issue): D590- D596.

Raahave, D. (2015). "Faecal retention: a common cause in functional bowel disorders, appendicitis and haemorrhoids--with medical and surgical therapy." Danish Medical Journal 61(3): B5031.

178

Rahat-Rozenbloom, S., J. Fernandes, G. B. Gloor and T. M. Wolever (2014). "Evidence for greater production of colonic short-chain fatty acids in overweight than lean humans." International Journal of Obesity 38(12): 1525-1531.

Ranadheera, R. D. C. S., S. K. Baines and M. C. Adams (2010). "Importance of food in probiotic efficacy." Food Research International 43(1): 1-7.

Riezzo, G., A. Orlando, B. D'Attoma, V. Guerra, F. Valerio, P. Lavermicocca, S. D. Candia and F. Russo (2012). "Randomised clinical trial: efficacy of Lactobacillus paracasei-enriched artichokes in the treatment of patients with functional constipation--a double-blind, controlled, crossover study." Alimentary Pharmacology & Therapeutics 35(4): 441-450.

Ringel-Kulka, T., J. B. Kotch, E. T. Jensen, E. Savage and D. J. Weber (2015). "Randomized, double-blind, placebo-controlled study of synbiotic yogurt effect on the health of children." The Journal of Pediatrics 166(6): 1475-1481.

Ringel-Kulka, T., O. Palsson, D. Maier and Y. Ringel (2008). "Yogurt containing the probiotic bacteria Bifidobacterium lactis BB-12 and prebiotic inulin significantly improves colonic transit time in subjects with functional bowel symptoms." American Journal of Gastroenterology 103(S1): S479.

Rizzardini, G., D. Eskesen, P. C. Calder, A. Capetti, L. Jespersen and M. Clerici (2012). "Evaluation of the immune benefits of two probiotic strains Bifidobacterium animalis ssp. lactis, BB-12 (R) and Lactobacillus paracasei ssp. paracasei, L. casei 431 (R) in an influenza vaccination model: a randomised, double-blind, placebo-controlled study." British Journal of Nutrition 107(6): 876-884.

Rizzoli, R. (2014). "Dairy products, yogurts, and bone health." The American Journal of Clinical Nutrition 99(Suppl 5): 1256S-1262S.

Robinson, S. and C. Fall (2012). "Infant nutrition and later health: a review of current evidence." Nutrients 4(8): 859-874.

Rosburg, V., T. Boylston and P. White (2010). "Viability of bifidobacteria strains in yogurt with added oat beta-glucan and corn starch during cold storage." Journal of Food Science 75(5): C439-C444.

Rosenfeldt, V., K. F. Michaelsen, M. Jakobsen, C. N. Larsen, P. L. Møller, P. Pedersen, M. Tvede, H. Weyrehter, N. H. Valerius and A. Paerregaard (2002). "Effect of probiotic Lactobacillus strains in young children hospitalized with acute diarrhea." The Pediatric Infectious Disease Journal 21(5): 411-416.

Roy, C. C., C. L. Kien, L. Bouthillier and E. Levy (2006). "Short-chain fatty acids: ready for prime time?" Nutrition in Clinical Practice 21(4): 351-366.

179

Russell, D. A., R. P. Ross, G. F. Fitzgerald and C. Stanton (2011). "Metabolic activities and probiotic potential of bifidobacteria." International Journal of Food Microbiology 149: 88-105.

Saarela, M., I. Virkajärvia, H.-L. Alakomia, P. Sigvart-Mattilab and J. Mättöa (2006). "Stability and functionality of freeze-dried probiotic Bifidobacterium cells during storage in juice and milk." International Dairy Journal 16(12): 1477-1482.

Saavedra, J. M., N. A. Bauman, I. Oung, J. A. Perman and R. H. Yolken (1994). "Feeding of Bifidobacterium bifidum and Streptococcus thermophilus to infants in hospital for prevention of diarrhea and shedding of rotavirus." The Lancet 344(8929): 1046-1049.

Saccaro, D. M., A. Y. Tamime, A. L. O. P. S. Pilleggi and M. N. Oliveira (2009). "The viability of three probiotic organisms grown with yoghurt starter cultures during storage for 21 days at 4 degrees C." International Journal of Dairy Technology 62(3): 397-404.

Sanchez, M. and P. Bercik (2011). "Epidemiology and burden of chronic constipation." Canadian Journal of Gastroenterology 25(Suppl B): 11B-15B.

Sanders, M. E. and M. L. Marco (2010). "Food formats for effective delivery of probiotics." Annual Review of Food Science and Technology 1: 65-85.

Sartor, R. B. (2008). "Microbial influences in inflammatory bowel diseases." Gastroenterology 134(2): 577-594.

Satish Kumar, R., P. Kanmani, N. Yuvaraj, K. A. Paari, V. Pattukumar and V. Arul (2013). "Traditional Indian fermented foods: a rich source of lactic acid bacteria." International Journal of Food Sciences and Nutrition 64(4): 415-428.

Savard, P., B. Lamarche, M. E. Paradis, H. Thiboutot, E. Laurin and D. Roy (2011). "Impact of Bifidobacterium animalis subsp. lactis BB-12 and, Lactobacillus acidophilus LA-5-containing yoghurt, on fecal bacterial counts of healthy adults." International Journal of Food Microbiology 149(1): 50-57.

Scardovi, V. (1986). "Genus Bifidobacterium." In: Mair N.S., editor. Bergey's manual of systematic bacteriology. New York: Williams & Wilkins: 1418–1438.

Schiffrin, E. J., F. Rochat, H. Link-Amster, J. M. Aeschlimann and A. Donnet-Hughes (1995). "Immunomodulation of human blood cells following the ingestion of lactic acid bacteria." Journal of Dairy Science 78(3): 491-497.

Schloss, P. D., D. Gevers and S. L. Westcott (2011). "Reducing the effects of PCR amplification and sequencing artifacts on 16S rRNA-based studies." PLoS One 6(12): e27310.

180

Schloss, P. D., S. L. Westcott, T. Ryabin, J. R. Hall, M. Hartmann, E. B. Hollister, R. A. Lesniewski, B. B. Oakley, D. H. Parks, C. J. Robinson, J. W. Sahl, B. Stres, G. G. Thallinger, D. J. V. Horn and C. F. Weber (2009). "Introducing mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities." Applied and Environmental Microbiology 75(23): 7537-7541.

Schneider, S. M., F. Girard-Pipau, J. Filippi, X. Hebuterne, D. Moyse, G. C. Hinojosa, A. Pompei and P. Rampal (2005). "Effects of Saccharomyces boulardii on fecal short-chain fatty acids and microflora in patients on long-term total enteral nutrition." World Journal of Gastroenterology 11(39): 6165-6169.

Segata, N., J. Izard, L. Waldron, D. Gevers, L. Miropolsky, W. S. Garrett and C. Huttenhower (2011). "Metagenomic biomarker discovery and explanation." Genome Biology 12(6): R60.

Sergeant, M. J., C. Constantinidou, T. Cogan, C. W. Penn and M. J. Pallen (2012). "High-throughput sequencing of 16S rRNA gene amplicons: effects of extraction procedure, primer length and annealing temperature." PLoS One 7(5): e38094.

Shanahan, F. (2010). "Gut microbes: from bugs to drugs." The American Journal of Gastroenterology 105(2): 275-279.

Sherman, P. M., J. C. Ossa and K. Johnson-Henry (2009). "Unraveling mechanisms of action of probiotics." Nutrition in Clinical Practice 24(1): 10-14.

Siigur, U., K. E. Norin, G. Allgood, T. Schlagheck and T. Midtvedt (1994). "Concentrations and correlations of faecal short-chain fatty acids and faecal water content in man." Microbial Ecology in Health and Disease 7(6): 287-294.

Singh, R. P., S. G. Damle and A. Chawla (2011). "Salivary mutans streptococci and lactobacilli modulations in young children on consumption of probiotic ice-cream containing Bifidobacterium lactis Bb12 and Lactobacillus acidophilus La5." Acta Odontologica Scandinavica 69(6): 389-394.

Slavin, J. L., C. T. Sempos, P. M. Brauer and J. A. Marlett (1981). "Limits of predicting gastrointestinal transit time from other measures of bowel function." The American Journal of Clinical Nutrition 34(10): 2111-2116.

Smith, T. J., D. Rigassio-Radler, R. Denmark, T. Haley and R. Touger-Decker (2013). "Effect of Lactobacillus rhamnosus LGG® and Bifidobacterium animalis ssp. lactis BB- 12® on health-related quality of life in college students affected by upper respiratory infections." British Journal of Nutrition 109(11): 1999-2007.

Solano-Aguilar, G., K. P. Fernandez, H. Ets, A. Molokin, B. Vinyard, J. F. Urban and M. F. Gutierrez (2013). "Characterization of fecal microbiota of children with diarrhea in

181 two locations in Colombia South America." Journal of Pediatric Gastroenterology and Nutrition 56(5): 503-511.

Sonnenburg, J. L., C. T. L. Chen and J. I. Gordon (2006). "Genomic and metabolic studies of the impact of probiotics on a model gut symbiont and host. ." PLoS Biology 4(12): e413.

Sullivan, A., C. E. Nord and B. Evengard (2009). "Effect of supplement with lactic-acid producing bacteria on fatigue and physical activity in patients with chronic fatigue syndrome." Nutrition Journal 8: 4.

Tabbers, M. M., A. Chmielewska, M. G. Roseboom, C. Boudet, C. Perrin, H. Szajewska and M. A. Benninga (2009). "Effect of the consumption of a fermented dairy product containing Bifidobacterium lactis DN-173 010 on constipation in childhood: a multicentre randomised controlled trial." BMC Pediatrics 9: 22.

Tabbers, M. M., A. Chmielewska, M. G. Roseboom, N. Crastes, C. Perrin, J. B. Reitsma, O. Norbruis, H. Szajewska and M. A. Benninga (2011). "Fermented milk containing Bifidobacterium lactis DN-173 010 in childhood constipation: a randomized, double- blind, controlled trial." Pediatrics 127(6): e1392-e1399.

Taipale, T., K. Pienihakkinen, E. Isolauri, C. Larsen, E. Brockmann, P. Alanen, J. Jokela and E. Soderling (2011). "Bifidobacterium animalis subsp lactis BB-12 in reducing the risk of infections in infancy." British Journal of Nutrition 105(3): 409-416.

Tamboli, C. P., C. Neut, P. Desreumaux and J. F. Colombel (2004). "Dysbiosis in inflammatory bowel disease." Gut 53(1): 1-4.

Tharmaraj, N. and N. P. Shah (2003). "Selective enumeration of Lactobacillus delbrueckii ssp. bulgaricus, Streptococcus thermophilus, Lactobacillus acidophilus, Bifidobacteria, Lactobacillus casei, Lactobacillus rhamnosus, and Propionibacteria." Journal of Dairy Science 86(7): 2288-2296.

Tremaroli, V. and F. Bäckhed (2012). "Functional interactions between the gut microbiota and host metabolism." Nature 489(7415): 242-249.

Tulk, H. M., D. C. Blonski, L. A. Murch, A. M. Duncan and A. J. Wright (2013). "Daily consumption of a synbiotic yogurt decreases energy intake but does not improve gastrointestinal transit time: a double-blind, randomized, crossover study in healthy adults." Nutrition Journal 12: 87.

Tuohy, K. M., G. C. Rouzaud, W. M. Brück and G. R. Gibson (2005). "Modulation of the human gut microflora towards improved health using prebiotics--assessment of efficacy." Current Pharmaceutical Design 11(1): 75-90.

182

Turnbaugh, P. J., M. Hamady, T. Yatsunenko, B. L. Cantarel, A. Duncan, R. E. Ley, M. L. Sogin, W. J. Jones, B. A. Roe, J. P. Affourtit, M. Egholm, B. Henrissat, A. C. Heath, R. Knight and J. I. Gordon (2009). "A core gut microbiome in obese and lean twins." Nature 457(7228): 480-484.

Turnbaugh, P. J., R. E. Ley, M. A. Mahowald, V. Magrini, E. R. Mardis and J. I. Gordon (2006). "An obesity-associated gut microbiome with increased capacity for energy harvest." Nature 444(7122): 1027-1031.

Turroni, F., A. Ribbera, E. Foroni, D. Sinderen and M. Ventura (2008). "Human gut microbiota and bifidobacteria: from composition to functionality." Antonie Van Leeuwenhoek 94(1): 35-50.

Uchida, K., K. Akashi, A. Kusunoki, T. Ikeda, N. Katano, H. Motoshima and Y. Benno (2005). "Effect of fermented milk containing Bifidobacterium lactis BB-12 on stool frequency, defecation, fecal microbiota and safety of excessive ingestion in healthy female students." Journal of Nutritional Food 8(1): 39-51.

Ursell, L. K., J. L. Metcalf, L. W. Parfrey and R. Knight (2012). "Defining the human microbiome." Nutrition Reviews 70(Suppl 1): S38-S44.

Vandenplas, Y. and M. Benninga (2009). "Probiotics and functional gastrointestinal disorders in children." Journal of Pediatric Gastroenterology and Nutrition 48(Suppl 2): S107-S109.

Vanderhoof, J. A., D. B. Whitney, D. L. Antonson, T. L. Hanner, J. V. Lupo and R. J. Young (1999). "Lactobacillus GG in the prevention of antibiotic-associated diarrhea in children." The Journal of Pediatrics 135(5): 564-568.

Veiga, P., C. A. Gallini, C. Beal, M. Michaud, M. L. Delaney, A. DuBois, A. Khlebnikov, J. E. v. H. Vlieg, S. Punit, J. N. Glickman, A. Onderdonk, L. H. Glimcher and W. S. Garrett (2010). "Bifidobacterium animalis subsp. lactis fermented milk product reduces inflammation by altering a niche for colitogenic microbes." PNAS 107(42): 18132-18137.

Ventura, M., R. Reniero and R. Zink (2001). "Specific identification and targeted characterization of Bifidobacterium lactis from different environmental isolates by a combined multiplex-PCR approach." Applied and Environmental Microbiology 67(6): 2760-2765.

Vernazza, C. L., G. R. Gibson and R. A. Rastall (2006). "Carbohydrate preference, acid tolerance and bile tolerance in five strains of Bifidobacterium." Journal of Applied Microbiology 100(4): 846-853.

Vijay-Kumar, M., J. D. Aitken, F. A. Carvalho, T. C. Cullender, S. Mwangi, S. Srinivasan, S. V. Sitaraman, R. Knight, R. E. Ley and A. T. Gewirtz (2010). "Metabolic

183 syndrome and altered gut microbiota in mice lacking Toll-like receptor 5." Science 328(5975): 228-231.

Vinderola, C. G., N. Bailo and J. A. Reinheimer (2000). "Survival of probiotic microflora in Argentinian yoghurts during refrigerated storage." Food Research International 33(2): 97-102.

Vinderola, C. G., P. Mocchiutti and J. A. Reinheimer (2002). "Interactions among lactic acid starter and probiotic bacteria used for fermented dairy products." Journal of Dairy Science 85(4): 721-729.

Vitali, B., M. Ndagijimana, S. Maccaferri, E. Biagi, M. E. Guerzoni and P. Brigidi (2012). "An in vitro evaluation of the effect of probiotics and prebiotics on the metabolic profile of human microbiota." Anaerobe 18(4): 386-391.

Waller, P. A., P. K. Gopal, G. J. Leyer, A. C. Ouwehand, C. Reifer, M. E. Stewart and L. E. Miller (2011). "Dose-response effect of Bifidobacterium lactis HN019 on whole gut transit time and functional gastrointestinal symptoms in adults." Scandinavian Journal of Gastroenterology 46(9): 1057-1064.

Wang, Q., G. M. Garrity, J. M. Tiedje and J. R. Cole (2007). "Na ̈ıve bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy." Applied and Environmental Microbiology 73(16): 5261-5267.

Wei, Y., J. Gong, W. Zhu, D. Guo, L. Gu, N. Li and J. Li (2015). "Fecal microbiota transplantation restores dysbiosis in patients with methicillin resistant Staphylococcus aureus enterocolitis." BMC Infectious Diseases 15: 265.

Wei, Y., W. Zhu, J. Gong, D. Guo, L. Gu, N. Li and J. Li (2015). "Fecal microbiota transplantation improves the quality of life in patients with inflammatory bowel disease." Gastroenterology Research and Practice 2015(517597): 1-5.

Weinstock, G. M. (2012). "Genomic approaches to studying the human microbiota." Nature 489(7415): 250-256.

Weizman, Z., G. Asli and A. Alsheikh (2005). "Effect of a probiotic infant formula on infections in child care centers: comparison of two probiotic agents." Pediatrics 115: 5-9.

Wen, L., R. E. Ley, P. Y. Volchkov, P. B. Stranges, L. Avanesyan, A. C. Stonebraker, C. Hu, F. S. Wong, G. L. Szot, J. A. Bluestone, J. I. Gordon and A. V. Chervonsky (2008). "Innate immunity and intestinal microbiota in the development of Type 1 diabetes." Nature 455(7216): 1109-1113.

West, N. P., D. B. Pyne, A. W. Cripps, C. T. Christophersen, M. A. Conlon and P. A. Fricker (2012). "Gut balance, a synbiotic supplement, increases fecal Lactobacillus

184 paracasei but has little effect on immunity in healthy physically active individuals." Gut Microbes 3(3): 221-227.

Whorwell, P. J., L. Altringer, J. Morel, Y. Bond, D. Charbonneau, L. O'Mahony, B. Kiely, F. Shanahan and E. M. Quigley (2006). "Efficacy of an encapsulated probiotic Bifidobacterium infantis 35624 in women with irritable bowel syndrome." The American Journal of Gastroenterology 101(7): 1581-1590.

Wichmann, A., A. Allahyar, T. U. Greiner, H. Plovier, G. Ö. Lundén, T. Larsson, D. J. Drucker, N. M. Delzenne, P. D. Cani and F. Bäckhed (2013). "Microbial modulation of energy availability in the colon regulates intestinal transit." Cell Host & Microbe 14(5): 582-590.

Williams, M. D., C. Y. Ha and M. A. Ciorba (2010). "Probiotics as therapy in gastroenterology: a study of physician opinions and recommendations." Journal of Clinical Gastroenterology 44(9): 631-636.

Wong, J. M., A. Esfahani, N. Singh, C. R. Villa, A. Mirrahimi, D. J. Jenkins and C. W. Kendall (2012). "Gut microbiota, diet, and heart disease." Journal of AOAC International 95(1): 24-30.

Wong, J. M., R. d. Souza, C. W. Kendall, A. Emam and D. J. Jenkins (2006). "Colonic health: fermentation and short chain fatty acids." Journal of Clinical Gastroenterology 40(3): 235-243.

Wu, G. D., J. Chen, C. Hoffmann, K. Bittinger, Y. Y. Chen, S. A. Keilbaugh, M. Bewtra, D. Knights, W. A. Walters, R. Knight, R. Sinha, E. Gilroy, K. Gupta, R. Baldassano, L. Nessel, H. Li, F. D. Bushman and J. D. Lewis (2011). "Linking long-term dietary patterns with gut microbial enterotypes." Science 334(6052): 105-108.

Xu, M. Q., H. L. Cao, W. Q. Wang, S. Wang, X. C. Cao, F. Yan and B. M. Wang (2015). "Fecal microbiota transplantation broadening its application beyond intestinal disorders." World Journal of Gastroenterology 21(1): 102-111.

Yatsunenko, T., F. E. Rey, M. J. Manary, I. Trehan, M. G. Dominguez-Bello, M. Contreras, M. Magris, G. Hidalgo, R. N. Baldassano, A. P. Anokhin, A. C. Heath, B. Warner, J. Reeder, J. Kuczynski, J. G. Caporaso, C. A. Lozupone, C. Lauber, J. C. Clemente, D. Knights, R. K. R and J. I. Gordon (2012). "Human gut microbiome viewed across age and geography." Nature 486(7402): 222-227.

Yurkovetskiy, L., M. Burrows, A. A. Khan, L. Graham, P. Volchkov, L. Becker, D. Antonopoulos, Y. Umesaki and A. V. Chervonsky (2013). "Gender bias in autoimmunity is influenced by microbiota." Immunity 39(2): 400-412.

185

Ziarno, M. and D. Zareba (2015). "Effects of milk components and food additives on survival of three bifidobacteria strains in fermented milk under simulated gastrointestinal tract conditions." Microbial Ecology in Health and Disease 26: 27812.

186

Appendix I: The Procedure of Yogurt and Yogurt Smoothie Preparation

1. Batch product as per batch sheet calculated using Tech Wizard. Typical batch size is 150 (130) gallons (1308/1133.6 lb, 593.3/514.2 kg) a. Take a sample (take duplicate samples) of raw mix for compositional analysis 2. Pasteurize the mix 3. Transfer the mix to PT6 (About 130 gallons) 4. Double check the volume of the mix in PT6 with the measure stick, if the tank has more than 130 gallons mix, remove the extra 5. Heat-treat the mix to 180°F (82°C) for 30 minutes 6. Cool to 110-111oF 7. Take 2 samples from PT6 for Coliform analysis after cooling to incubation temperature 8. Withdraw about 1350 g white mass with sanitized 10-pound white bucket and mix it with 150 g YF-L702 to make 10% starter culture stock 9. Add 1028.4 g (0.02%)yogurt starter stock into PT6-PREWEIGHED IN BONNIE’S LAB 10. Mix for 20 minutes. 11. Turn agitators off. a. Take a sample for INITIAL pH 12. Sanitize the three-way valve 13. Transfer about 78 lb of the mix to each of the two sanitized, pre-weighed milk cans through the three-way valve 14. ADD 39.7 g of BB-12 into them and mix them well with dairy mixer (Take two samples for initial BB-12 count and pH) 15. Incubate to a pH of 4.65 (Monitor carefully) 16. Turn the agitator on low speed For the milk cans, use the steam kettle in the pilot plant and ice to cool them down to 70oF 17. Turn cooling (city water) on. VERIFY THE STEAM VALVE IS OFF! 18. Cool the vat to 70°F (21°C) 19. Increase speed of the agitator to full speed. 20. Add PECTIN SLURRY and Strawberry (Sensient #242529384) --Pre cooled to100°F (38°C)--from the Lee kettle to PT6 a. Draw 50.7 lb slurry/Strawberry for MILK CANS through the homogenizer into a can at 120oF and set it aside for later use b. Pump remaining slurry/strawberry through the homogenizer into PT6

187

21. Pump a portion of the yogurt back through the lines to the steam kettle and then back to the PT6 to “clear the lines” of excess slurry 22. Agitate for at least 10 minutes to assure uniform distribution 23. Homogenize (500/1000 psi) about 40 gallons into the Cream Pasteurizer (CONTROL “A”, YS) 24. Bottle the CONTROL “A”. Take a sample from the bottle machine at the beginning, middle and end of the run for coliforms and composition 25. Pump 55 (479.6 lb/217.54 kg) gallons of the remained mixture into the Lee kettle 26. Add 878.5 g of BB12 into the remained mixture in Lee kettle 27. Mix Well (at least 5 minutes) 28. Homogenize the TREATMENT “B” (POST) PRODUCT into the Cream Pasteurizer 29. Bottle the TREATMENT “B” Take a sample from the bottle machine at the beginning, middle and end of the run for coliforms and composition. 30. Dump the two cans of yogurt and the slurry into the Lee Kettle 31. Mix them well (at least 5 minutes) 32. Homogenize the TREATMENT “C” (PRE) PRODUCT through the funnel and collect the product in CLEAN AND SANITIZED MILK CANS 33. Dump the TREATMENT “C” PRODUCT into the CLEANED AND SANITIZED filling tank of the bottle line 34. Bottle the TREATMENT “C” Take a sample from the bottle machine at the beginning, middle and end of the run for coliforms and composition 35. Clean up! Byron is in charge of step 8, 9 as well as the fermentation in milk cans (Treatment C)

SLURRY COMPOSITION

INGREDIENT AMOUNT (% W/W) SUGAR 18 36DE SS 29.8 PECTIN 1.86 WATER 50.34 Note the product contains 21.5% slurry

188

PREPARATION (For 130 gallons batch of YOGURT white mass)

1. For 130 gallons of white mass need 322.8 lb of slurry 2. Add 162.5 lb water to the Lee Kettle 3. Turn the agitator on. 4. Add 96.2 lb CSS 5. Add 58.1 lb of sugar (use a portion to cut the pectin) 6. Sprinkle in 6.0 lb pectin-dry blended with sugar 7. Heat the SLURRY to 180-190°F and mix until the pectin is dissolved. 8. Mix for a MINIMUM of 1 hour (Until the slurry appears uniform) 9. COOL the slurry to 100°F 10. Add the STRAWBERRY 45.1 lb and mix 11. Homogenize slurry prior to addition to the yogurt 12. Add as in step 20 above

Figure I-1. Processing flow diagram

189

Data Sheets

Table I-1. Processing data PT6 MILK CAN Item Target Actual Target Actual Amount of yogurt batched 150 Gallons 18 Amount of yogurt in Tank 110 Gallons Gallons Amount of 10% starter culture 1028.4 Grams stock XXXX 79.3 Amount of BB12 Added XXXXX X Grams Time of starter addition Time when product reaches pH

4.5 Time when product reaches

75°F Amount of slurry added 272.1 lb 50.7 lb Amount of BB 12 added (For 878.5 Grams Treatment B)

190

Appendix II: Treatment Sequences

Subject ID Treatment Sequence PRO002 ADCB PRO003 CABD PRO004 BCDA PRO006 ADCB PRO007 CABD PRO008 DBAC PRO009 ADCB PRO011 DBAC PRO012 CABD PRO013 ADCB PRO014 DBAC PRO015 CABD PRO016 BCDA PRO017 BCDA PRO019 DBAC PRO020 DACB PRO021 CADB PRO022 ADCB PRO023 BCDA PRO024 DBAC PRO025 ADCB PRO026 CDBA PRO028 BDCA PRO029 DACB PRO030 DABC PRO032 DCBA PRO033 DCBA PRO034 DBAC PRO035 DACB

A – yogurt smoothie without BB-12, B – yogurt smoothie with BB-12 added post-fermentation, C – yogurt smoothie with BB-12 added pre-fermentation, D – BB-12 containing capsule.

191

Appendix III: Stool Collection Instructions for Participant

Stool is to be collected at baseline, after each treatment, and after final washout. At the time of enrollment you will be provided with stool collection materials. This kit will include: disposable gloves, two stool containers (vials), a plastic zipper bag and a stool collection device (a “stool hat”). At the time of stool collection: 1. Wash your hands. 2. Put on disposable gloves. 3. Write your STUDY ID number and the date and time the stool was collected. Do this 3 times: once on the label of each stool container (vials) and once on the label on the plastic zipper bag. Please do not write your name on the labels. 4. Lift the toilet seat. Align the “stool hat” with the rear of the toilet rim (see Photo A). Use the toilet as you normally would, letting the stool collect in the “hat”. 5. Unscrew the cap from the stool container. Use the spoon attached at the end of the cap of the stool container to fill the container to the fill line indicated on the vial (see Photo B). Screw the cap back onto the container. Make sure the cap is tightly closed. Repeat for both vials. 6. Discard remaining stool into the toilet and flush. Discard stool hat. 7. Place the stool containers into the plastic zipper bag. Seal the bag. 8. Place the sealed plastic zipper bag into the FREEZER immediately for storage. 9. Wash your hands. 10. Bring the stool samples to CRC when you return the wireless receiver.

Photo A Photo B

192

Appendix IV: Gut transit times of all participants

Subject ID Treatment Sequence Gender WGTT (h) CTT (h) SBTT (h) GET (h) PRO 002 Baseline 0 F 47.83 39.22 5.92 2.67 PRO 002 YS 1 F 22.50 17.25 1.40 3.85 PRO 003 Baseline 0 F 27.15 21.38 3.92 1.83 PRO 003 YS 2 F 49.92 40.45 6.22 3.23 PRO 003 POST 3 F 56.82 49.92 4.53 2.35 PRO 003 PRE 1 F 30.70 23.72 4.37 2.58 PRO 003 CAP 4 F 46.68 40.53 3.68 2.45 PRO 004 YS 4 F 123.58 117.27 2.97 3.33 PRO 004 POST 1 F 28.13 21.87 3.33 2.93 PRO 004 Baseline 0 F 57.05 51.37 1.82 3.85 PRO 004 PRE 2 F 24.92 16.00 4.85 4.05 PRO 004 CAP 3 F 47.77 38.70 6.38 2.67 PRO 006 Baseline 0 F 44.85 32.72 8.15 3.97 PRO 006 YS 1 F 24.42 14.45 5.90 4.05 PRO 006 PRE 3 F 21.52 9.70 7.27 4.53 PRO 006 CAP 2 F 45.63 17.43 9.35 18.83 PRO 007 Baseline 0 F 46.70 39.92 5.02 1.75 PRO 007 YS 2 F 97.22 87.07 7.40 2.73 PRO 007 POST 3 F 79.65 70.50 6.92 2.22 PRO 007 PRE 1 F 101.93 92.45 6.87 2.58 PRO 007 CAP 4 F 26.30 20.40 4.00 1.87 PRO 008 YS 3 F 24.85 17.67 3.35 3.82 PRO 008 POST 2 F 31.98 26.03 2.57 3.37 PRO 008 Baseline 0 F 27.37 22.08 2.80 2.47 PRO 008 PRE 4 F 34.17 28.17 2.15 3.83 PRO 008 CAP 1 F 33.45 29.27 2.53 1.65 PRO 009 YS 1 M 25.72 19.15 5.30 1.25 PRO 009 POST 4 M 30.37 21.57 6.28 2.50 PRO 009 Baseline 0 M 26.45 17.22 6.45 2.77 PRO 009 PRE 3 M 23.13 14.90 2.53 5.70 PRO 009 CAP 2 M 7.13 0.50 5.42 1.22 PRO 011 YS 3 F 22.73 17.13 3.45 2.13 PRO 011 PRE 4 F 23.33 16.38 4.72 2.22 PRO 011 CAP 1 F 24.90 18.37 3.98 2.52

193

PRO 011 POST 2 F 48.50 40.93 4.63 2.92 PRO 012 Baseline 0 M 21.17 14.83 4.60 1.72 PRO 012 PRE 1 M 26.03 18.18 1.00 3.83 PRO 013 YS 1 M 25.65 17.00 5.17 3.47 PRO 013 POST 4 M 25.32 17.48 5.02 2.80 PRO 013 Baseline 0 M 24.62 15.92 4.32 4.35 PRO 013 PRE 3 M 30.65 25.73 2.10 2.80 PRO 013 CAP 2 M 27.68 22.52 3.22 1.93 PRO 014 YS 3 F 26.38 20.65 4.73 0.98 PRO 014 Baseline 0 F 14.75 7.87 3.85 3.02 PRO 014 PRE 4 F 27.18 20.37 5.22 1.58 PRO 014 CAP 1 F 26.27 20.83 4.10 1.32 PRO 015 YS 2 F 25.32 12.25 10.20 2.85 PRO 015 POST 3 F 23.48 15.35 5.38 2.73 PRO 015 Baseline 0 F 79.55 70.07 4.92 4.55 PRO 015 PRE 1 F 47.22 38.77 5.00 3.43 PRO 015 CAP 4 F 54.00 45.23 4.50 4.25 PRO 016 YS 4 M 24.90 19.18 3.52 2.20 PRO 016 POST 1 M 49.02 42.42 4.07 2.52 PRO 016 Baseline 0 M 12.25 6.42 3.18 2.65 PRO 016 PRE 2 M 14.65 7.25 4.80 2.58 PRO 016 CAP 3 M 49.98 43.52 4.57 1.88 PRO 017 YS 4 M 48.48 40.02 6.35 2.08 PRO 017 POST 1 M 47.50 41.47 4.13 1.88 PRO 017 Baseline 0 M 33.85 26.15 5.23 2.45 PRO 017 PRE 2 M 27.42 20.88 4.20 2.30 PRO 017 CAP 3 M 26.45 18.97 5.75 1.72 PRO 019 YS 3 M 26.15 20.20 3.45 2.47 PRO 019 POST 2 M 7.43 2.17 3.50 1.75 PRO 019 Baseline 0 M 27.03 20.93 4.23 1.87 PRO 019 PRE 4 M 23.25 17.03 4.33 1.87 PRO 019 CAP 1 M 8.95 3.40 3.00 2.53 PRO 020 YS 2 F 34.82 26.63 3.13 5.03 PRO 020 POST 4 F 26.88 18.83 2.72 5.32 PRO 020 Baseline 0 F 49.03 40.12 5.18 3.72 PRO 020 PRE 3 F 120.57 111.22 4.63 4.72 PRO 020 CAP 1 F 60.73 52.32 4.30 4.12

194

PRO 021 YS 2 M 23.85 17.20 4.28 2.35 PRO 021 POST 3 M 17.60 11.75 3.10 2.72 PRO 021 Baseline 0 M 25.88 20.83 2.68 2.37 PRO 021 PRE 1 M 11.93 7.15 2.60 2.17 PRO 021 CAP 4 M 14.23 8.37 3.30 2.55 PRO 022 YS 1 M 49.02 39.32 7.27 2.43 PRO 022 POST 4 M 23.62 19.17 1.00 3.43 PRO 022 Baseline 0 M 50.28 43.32 4.47 2.48 PRO 022 PRE 3 M 23.60 15.58 4.95 3.05 PRO 022 CAP 2 M 47.92 40.80 5.08 2.02 PRO 024 YS 3 F 36.05 27.38 5.12 3.53 PRO 024 POST 2 F 33.40 21.47 4.47 7.45 PRO 024 Baseline 0 F 57.93 49.55 5.78 2.58 PRO 024 PRE 4 F 23.72 14.85 7.33 1.52 PRO 024 CAP 1 F 34.27 26.33 4.30 3.60 PRO 025 YS 1 F 49.50 41.43 4.18 3.88 PRO 025 POST 4 F 35.13 26.45 6.22 2.45 PRO 025 Baseline 0 F 61.82 52.38 4.17 5.23 PRO 025 PRE 3 F 54.75 43.93 6.90 3.90 PRO 025 CAP 2 F 48.67 37.52 6.38 4.77 PRO 026 POST 3 F 72.30 62.15 4.63 5.52 PRO 026 Baseline 0 F 51.03 42.40 5.93 2.67 PRO 026 PRE 1 F 78.88 72.85 3.25 2.75 PRO 026 CAP 2 F 33.43 27.25 2.65 3.52 PRO 028 YS 4 M 8.25 2.38 4.25 1.58 PRO 028 POST 2 M 16.75 10.82 4.02 1.90 PRO 028 Baseline 0 M 128.58 122.25 3.78 2.55 PRO 028 PRE 3 M 54.23 44.98 6.90 2.35 PRO 028 CAP 1 M 31.18 25.15 4.58 1.42 PRO 029 YS 2 F 27.15 21.62 2.37 3.13 PRO 029 POST 4 F 25.97 18.83 3.58 3.53 PRO 029 Baseline 0 F 24.28 18.48 3.30 2.48 PRO 029 PRE 3 F 30.35 21.33 3.87 5.13 PRO 029 CAP 1 F 22.32 13.38 3.20 5.70 PRO 030 YS 2 M 29.27 19.70 6.38 3.17 PRO 030 POST 3 M 30.15 22.43 4.63 3.07 PRO 030 Baseline 0 M 10.07 2.37 4.88 2.80

195

PRO 030 PRE 4 M 52.67 45.67 4.30 2.67 PRO 030 CAP 1 M 62.63 55.10 4.20 3.30 PRO 032 YS 4 M 25.50 19.45 3.73 2.32 PRO 032 POST 3 M 30.88 24.40 4.65 1.80 PRO 032 Baseline 0 M 52.92 45.02 4.98 2.88 PRO 032 PRE 2 M 53.45 46.28 5.12 2.05 PRO 032 CAP 1 M 25.18 19.22 3.38 2.57 PRO 033 YS 4 F 35.05 12.43 19.02 3.60 PRO 033 POST 3 F 27.62 21.38 4.12 2.12 PRO 033 Baseline 0 F 29.15 22.48 4.67 1.98 PRO 033 PRE 2 F 29.77 23.93 4.13 1.67 PRO 033 CAP 1 F 33.93 27.87 4.45 1.60 PRO 034 POST 2 F 47.57 34.12 10.23 3.20 PRO 034 Baseline 0 F 46.68 38.27 5.13 3.27 PRO 034 CAP 1 F 45.07 36.68 5.22 3.15 PRO 035 YS 2 F 15.65 6.62 5.77 3.27 PRO 035 POST 4 F 24.50 18.83 2.40 3.25 PRO 035 Baseline 0 F 22.78 13.23 7.27 2.27 PRO 035 PRE 3 F 11.17 3.55 4.53 3.07 PRO 035 CAP 1 F 39.30 30.80 2.98 5.50

Abbreviations: GET gastric empty time, SBTT small bowel, CTT colonic transit time, WGTT whole gut transit time, YS yogurt smoothie without BB-12, PRE yogurt smoothie with BB-12 added before fermentation, POST yogurt smoothie with BB-12 added after fermentation, CAP BB-12 containing capsule, F female, M male.

196

VITA – Zhaoyong Ba

Education

Ph.D., Food Science The Pennsylvania State University May 2016 * Graduate minor in Statistics M.E., Food Biotechnology, China Agricultural University, Beijing, China July 2010 B.E., Food Science and Engineering Tianjin University, Tianjin, China July 2007 Publications

Ba Z., Y. Lee, H. Meng, E.J. Furumoto, J.A. Fleming, C.J. Rogers, P.M. Kris-Etherton, R.F. Roberts. The Effect of Bifidobacterium animalis subsp. lactis BB-12 and different delivery vehicles on gut transit time and fecal short chain fatty acid production in healthy adults. (In preparation) Tan T.P., Z. Ba, F. D’Amico, R.F. Roberts, K.H. Smith, D.J. Merenstein. Safety of Bifidobacterium animalis subsp. lactis (B. lactis) strain BB-12-supplemented yogurt in healthy children: a phase I safety study. (In preparation) Meng H., Z. Ba, Y. Lee, J.A. Fleming, E.J. Furumoto, R.F. Roberts, P.M. Kris-Etherton, C.J. Rogers. (2015). Consumption of Bifidobacterium animalis subsp. lactis BB-12 in yogurt reduced expression of TLR-2 on peripheral blood-derived monocytes and pro-inflammatory cytokine secretion in young adults. European Journal of Nutrition. (Accepted) Tang W., X. Zhu, Z. Ba. (2015). Bioactive substances of microbial origin. In Cheung, P.C.K. & Mehta, B.M. (Eds.). Handbook of Food Chemistry. Berlin: Springer, pp. 1035-1060. Meng H., Y. Lee, Z. Ba, J.A. Fleming, E.J. Furumoto, R.F. Roberts, P.M. Kris-Etherton, C.J. Rogers. (2015). In vitro production of IL-6 and IFN-γ is influenced by dietary variables and predicts upper respiratory tract infection incidence and severity respectively in young adults. Frontiers in Immunology, 6: 94. Selected Awards

Donald V. Josephson and Stuart Patton Mentorship Award in Dairy and Food Science 2013, 2015 Janet G. and Frank J. Dudek Graduate Scholarship in Food Science 2014 IFT Feeding Tomorrow Travel Grant 2014 Select Experience

Food Science 414 (Dairy Science and Technology), Teaching Assistant 2011, 2012, 2015 Food Science 497C (Topics in Dairy Products Processing), Teaching Assistant 2014 Ice Cream Short Course, Teaching Assistant 2011 - 2015 Ice Cream 101, Teaching Assistant 2011 - 2015 Cultured Dairy Products Short Course, Teaching Assistant 2010 - 2015 NDC New Product Development Competition, Team Leader 2015