sign in sign up
<<
Home , Cytidine monophosphate, Dihydrothymine, DPYS, Guanine deaminase, Guanosine, Guanosine monophosphate

Biochemistry of Salivary Metabolites and their Role in the Innate Immunity of Neonates

Saad Saeed Abdulwahab Al-Shehri

(MSc, Clinical , Griffith University)

A thesis submitted for the degree of Doctor of Philosophy at

The University of Queensland in 2013

The School of Pharmacy ABSTRACT

Saliva contains a variety of biochemical components, some of which may be used for the diagnosis of metabolic disorders, as cancer and heart disease markers, and the drug monitoring purposes. Saliva collection is particularly attractive as a non-invasive sampling method for infants and elderly patients. However, the use of saliva to evaluate metabolic status has not been widely studied; in particular, nucleotide metabolites in saliva have received little attention.

Nucleotides are divided mainly into and , and they play an essential role in many cellular processes involving , energy supply, DNA/RNA synthesis, essential coenzymes, and regulatory mechanisms. are synthesised in mammalian cells by de novo pathways, which are energy-consuming; alternatively they are recycled by energy-saving pathways from and bases that act as metabolites.

This project was initiated by the discovery of a wide variation in salivary concentrations of nucleosides and bases between individuals, raising questions about their biological function. A method was developed and validated for the collection of small volumes of saliva as well as the extraction of purines and pyrimidines from oral swabs. These were then analysed by HPLC with tandem mass spectrometry.

A survey of salivary nucleotide metabolites in 77 adults was performed, followed by a study of 60 neonates. Concentrations of the major salivary and nucleosides were significantly higher in neonatal compared to adult saliva. The median concentrations (µM) in neonates\adults respectively were: 7.3\<1.5, Uridine 12\0.4, Hypoxanthine 27\2.1, 19\1.7, 12\0.1, 11\0.2, 6.8\0.1. These nucleotide metabolites were not simply in equilibrium with plasma, but appeared to be actively secreted into saliva. The high levels of xanthine and in neonates were particularly interesting, because these are substrates for xanthine (XO), which occurs with high activity in breast milk.

These observations led to a general hypothesis that the role of these nucleotide metabolites in neonatal saliva is concerned with regulation of the oral microbiota of infants. XO is involved in the final stage of degradation of nucleotides, and it has been proposed to have a role in the innate immunity of babies arising from its ability to generate antibacterial ‘’. The hypothesis proposed that high concentrations of nucleotide metabolites in neonatal saliva, together with stimulation of antibacterial breast milk XO activity, may play significant roles in early innate immunity in neonates. As part of the project, it was decided to also screen saliva from a selection of domestic mammals, which were found to have large differences in metabolite concentrations.

ii

Investigation of the generation of (H2O2) by breast milk XO was conducted in three stages: (1) fresh breast milk samples were collected and the endogenous H2O2 concentration was determined to be 27.3±12.2 µM; (2) XO activity was then assayed in the same samples using a horseradish peroxidase-Ampliflu™ Red based reaction; the activity of breast milk XO was found to be 8.0±5.3 Units/L; (3) The ability of breast milk to generate H2O2 (40 µM) when mixed with neonatal saliva was successfully demonstrated. XO was confirmed to be inactive in milk formula and pasteurized bovine milk and human breast milk.

Investigation of the effect of exogenous H2O2 on the in vitro growth of 4 bacterial species was undertaken. The peroxide mechanism had a strong inhibitory effect on Staphylococcus aureus at low, physiologically relevant concentrations (>25 µM H2O2). Lactobacillus plantarum and Escherichia coli required significantly higher concentrations of peroxide (~200 µM) to inhibit growth, while the effect of peroxide on the growth of Salmonella species was intermediate. As a proof-of-principle experiment, the effects of a saliva and breast milk mixture on bacterial growth in vitro were then tested. The reaction between salivary xanthine/hypoxanthine and breast milk XO was shown to be bactericidal against S. aureus and Salmonella spp., (p=0.001, p=0.007 respectively) whereas the effects on L. plantarum and E. coli were not statistically significant. A dose-dependent bactericidal effect of hypoxanthine/xanthine with milk on the 4 bacterial species was also demonstrated.

Finally, to test whether breast-feeding (i.e. active XO mechanism), compared with formula-feeding (inactive XO mechanism), produces an in vivo effect on the oral microbiota, a survey of oral microbiota in infants was undertaken. Oral swabs were collected from both groups and the 16S rRNA bacterial was ‘deep sequenced’ using Roche 454 platform. Bioinformatic analysis software was used to determine bacterial phyla and families (and is continuing to species level). Significant variation between the oral microbiota of breast-fed and formula-fed infants was found. Although other components of breast milk such as antibodies are presumed to have an effect on the infant’s microbiota, it was considered essential to show that an effect existed, and that it could be measured in the mouth – this has not previously been demonstrated. Teasing apart the ‘innate’ (peroxidative mechanism) immune system from the specific (antibody mechanism) system is a task for future studies.

iii

DECLARATION BY AUTHOR

This thesis is composed of my original work, and contains no material previously published or written by another person except where due reference has been made in the text. I have clearly stated the contribution by others to jointly-authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole, including statistical assistance, survey design, data analysis, significant technical procedures, professional editorial advice, and any other original research work used or reported in my thesis. The content of my thesis is the result of work I have carried out since the commencement of my research higher degree candidature and does not include a substantial part of work that has been submitted to qualify for the award of any other degree or diploma in any university or other tertiary institution. I have clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.

I acknowledge that an electronic copy of my thesis must be lodged with the University Library and, subject to the General Award Rules of The University of Queensland, immediately made available for research and study in accordance with the Copyright Act 1968.

I acknowledge that copyright of all material contained in my thesis resides with the copyright holder(s) of that material. Where appropriate I have obtained copyright permission from the copyright holder to reproduce material in this thesis.

iv

PUBLICATIONS DURING CANDIDATURE

Journal publications:

Al-Shehri, S., M. Henman, B.G. Charles, D. Cowley, P.N. Shaw, H. Liley, A. Tomarchio, C. Punyadeera, and J.A. Duley, Collection and determination of nucleotide metabolites in neonatal and adult saliva by high performance liquid chromatography with tandem mass spectrometry. J Chromatogr B, 2013. 931C: p. 140-147, (see Appendix 9)

In preparation:

1- Role of purines and pyrimidines and in regulating oral microbiota 2- Oral Bacterial 16S rRNA deep sequencing in breast-fed versus formula-fed infants

Published abstracts and posters:

Al-Shehri S., Shaw PN, Cowley D, Liley H, Henman M, Tomarchio A, Charles B, Duley JA. Nucleotide precursors in neonatal and adult saliva. J. Inherit. Metab. Dis. 2011, 34:S73 (Poster was presented at SSIEM, 2011, Geneva), (see Appendix 10 & 11)

Al-Shehri S., Henman M, Cowley D, Charles B , Shaw PN, Liley H, Duley JA. Longitudinal study of nucleotide metabolites in infants’ saliva compared to adults and mammals. (Poster at ComBio, 2012, Adelaide), (see Appendix 12 & 13)

Al-Shehri , SS., Liley, H., Knox, CL., Henman, M., Cowley, DM., Charles, B G., and Duley, JA. Human milk xanthine oxidase and neonatal salivary nucleotide precursors generate hydrogen peroxide; a novel pathway with potential role in regulating oral microflora. Journal of Paediatrics and Child Health. 2013, 49 (S2) p. 121. ISSN 1034-4810 ERA C, IF 1.281. (Poster was presented at PSANZ Annual Congress, 2013, Adelaide), (see Appendix 14 & 15)

Seminar oral presentations:

Al-Shehri S & Duley J. Baby Saliva and Breast Milk. Dept. of Pathology, Mater Hospital, Brisbane 25/11/2011

Al-Shehri S & Duley J. Baby Saliva and Breast Milk – What’s in the Mix? PSANZ meeting, Brisbane 15/12/2011

Al-Shehri S. Salivary nucleotide metabolites: their role in the innate immunity of babies, UQ 3MT competition, 2012

v

PUBLICATIONS INCLUDED IN THIS THESIS

No publications included

vi

CONTRIBUTIONS BY OTHERS TO THE THESIS

My principal advisor, Dr John Duley, as well as my associate supervisors; A/Prof. Bruce Charles, Prof. Nick Shaw, and Dr. David Cowley contributed to this thesis by providing input on experimental design, data analysis and critically revising and editing this document. Dr Christine Knox introduced me to the field of microbiology and 16S rRNA sequencing, arranged for me to work at the Institute of Health and Biomedical Innovation, QUT, as a visiting student, helped with data analysis, and sequencing bioinformatics. Australian Genome Research Facility (AGRF) and Purnika Ranasinghe (QUT) have contributed to this thesis by providing essential input in 16S rRNA sequence analyses. Mr Michael Henman introduced me to the field of HPLC-mass spectrometry and greatly aided me with method validations and sample analyses. Dr John Wright from the Veterinary School, UQ, contributed to this project by providing animal samples.

STATEMENT OF PARTS OF THE THESIS SUBMITTED TO QUALIFY FOR THE AWARD OF ANOTHER DEGREE

None

vii

ACKNOWLEDGEMENTS

Firstly, I thank Allah for this achievement.

I would like to express my deepest gratitude to my principal supervisor, Dr John A Duley for his encouragement, guidance, support and enduring patience. I also extend my sincere gratitude to my associate supervisors at both the University of Queensland and at the Mater Hospital, Brisbane, Australia: A/Prof. Bruce Charles, Prof. Nick Shaw, and Dr. David Cowley, for their invaluable guidance which had a remarkable influence on my knowledge. I also wish to express my warm and sincere gratitude to Dr Christine Knox from Queensland University of Technology for her wonderful supervision and access to the research facilities at Institute of Health and Biomedical Innovation (IHBI).

I thank all those acknowledged in the sections on publications, presentations and thesis contributions above. I thank Professor Deon Venter for his provision and access to the research space, and Dr. David Cowley for allowing me to utilise the biochemistry facilities, and Dr Helen Liley for helping with the ethics applications, organising access to the neonatal samples, and her advice on neonatal biology.

During this work I have also collaborated with many colleagues at PACE, QUT, and the Mater Hospital, for whom I have great regard and in particular I would like to thank Emma Sweeney, Angelo Tomarchio, and Michael Henman who devoted their time and efforts to assist me.

I would also like to thank the Golden Casket for funding this project, and my sponsor Taif University who provided me with the scholarship as well as the funds for my research.

I have been indebted to my loving wife Amal, my son Zyad, and my daughter Lara: without your encouragement, love and understanding it would have been impossible for me to finish this work. To my parents, who have been a constant source of support during my research period abroad, this thesis would certainly not have been existed without you both, and it is for you that this thesis is dedicated.

viii

KEYWORDS

Nucleotide metabolites, Neonates, Saliva, HPLC, Mass spectrometry, Xanthine oxidase, Peroxide, Microbiota, 16S rRNA

AUSTRALIAN AND NEW ZEALAND STANDARD RESEARCH CLASSIFICATIONS (ANZSRC)

ANZSRC code: 060101 Analytical Biochemistry, 50%

ANZSRC code: 060107 , 30%

ANZSRC code: 060104 Cell , 20%

FIELDS OF RESEARCH (FOR) CLASSIFICATION

FoR code: 0601 Biochemistry and Cell Biology, 80%

FoR code: 0605 Microbiology 10%

FoR code: 0699 Other Biological Sciences, 10%

ix

TABLE OF CONTENTS

Table of contents ...... x List of figures ...... xvi List of tables ...... xix List of appendices ...... xx List of abbreviations...... xxi

Chapter 1: Introduction ...... 1

1.1 Thesis structure ...... 2

1.2 General literature review ...... 3 1.2.1 Purine and bases and nucleosides ...... 3 1.2.2 Purine and pyrimidine nucleotides ...... 4 1.2.3 Coenzyme nucleotides ...... 5 1.2.4 Nucleotide cellular functions ...... 6 1.2.5 Metabolism of purine nucleotides ...... 9 1.2.5.1 ...... 9 1.2.5.2 Synthesis of ...... 9 1.2.5.3 Salvage of purines...... 11 1.2.5.4 of purines ...... 11 1.2.6 Metabolism of pyrimidine nucleotides ...... 13 1.2.6.1 De novo synthesis ...... 13 1.2.6.2 Salvage of pyrimidines ...... 15 1.2.6.3 Catabolism of pyrimidines ...... 15 1.2.7 Bacterial uptake and metabolism of purines and pyrimidines ...... 17 1.2.8 Xanthine oxidase ...... 19 1.2.8.1 General characteristics ...... 19 1.2.8.2 Structure and reaction mechanism ...... 19 1.2.8.3 Antibacterial properties of xanthine oxidase ...... 21 1.2.9 Physiology of reactive oxygen species ...... 23 1.2.10 Physiology of saliva ...... 25 1.2.10.1 Functions of saliva ...... 25 1.2.10.2 components of saliva ...... 25 1.2.10.3 Electrolyte components of saliva ...... 27 1.2.10.4 Oral microbiota ...... 27

x

1.2.10.5 Salivary production ...... 29 1.2.10.6 Diagnostic and metabolic roles of saliva ...... 30 1.2.11 Human milk physiology and composition ...... 31 1.2.11.1 Physiology of human milk secretion ...... 31 1.2.11.2 Human milk ...... 31 1.2.11.3 Human milk ...... 32 1.2.11.4 Human milk ...... 32 1.2.11.5 White blood cells in human milk ...... 32 1.2.11.6 Human milk minerals ...... 32 1.2.11.7 Human milk vitamins ...... 33 1.2.11.8 Human milk nucleotides ...... 33 1.2.11.9 Mastitis ...... 33

1.3 Overall project hypothesis ...... 35

1.4 Overall project aim ...... 35

1.5 Ethics ...... 35

1.6 Funding for the project ...... 35

Chapter 2: Determination of nucleotide metabolites in human and mammalian saliva by high performance liquid chromatography with tandem mass spectrometry: changes during early human infancy and a survey of ranges in adult humans and other mammals… ...... 36

2.1 Introduction ...... 37 2.1.1 Roles of purines and pyrimidines...... 37 2.1.2 Salivary purines and pyrimidines ...... 37

2.2 Specific hypotheses and rationale ...... 39

2.3 Specific aims ...... 39

2.4 Materials and methods ...... 41 2.4.1 Subjects and sample collection ...... 41 2.4.1.1 Adults for saliva survey ...... 41 2.4.1.2 Neonates for saliva survey ...... 41 2.4.1.3 Subjects for infant longitudinal study ...... 42 2.4.1.4 Survey of other mammalian species ...... 42 2.4.2 Neonatal saliva extraction and validation ...... 43

xi

2.4.2.1 Investigation for presence of interfering compounds ...... 43 2.4.2.2 Validation of recovery of extraction ...... 43 2.4.2.3 Validation of linearity of extraction ...... 44 2.4.3 Analysis of purines and pyrimidines ...... 44 2.4.3.1 Sample preparation ...... 44 2.4.3.2 Initial analysis using HPLC ...... 44 2.4.3.3 LC-MS/MS conditions ...... 45 2.4.3.4 Within run imprecision ...... 47 2.4.3.5 Limit of detection ...... 47 2.4.4 Statistical analysis ...... 47

2.5 Results ...... 48 2.5.1 Neonatal saliva extraction and validation ...... 48 2.5.1.1 Investigation of the presence of interfering compounds...... 48 2.5.1.2 Validation of recovery of extraction ...... 50 2.5.1.3 Validation of linearity of extraction ...... 54 2.5.2 HPLC-MS/MS determination of nucleotide metabolites ...... 56 2.5.2.1 Within run imprecision ...... 56 2.5.2.2 Limit of detection and condition parameters ...... 56 2.5.3 Survey of adult salivary nucleotide metabolites ...... 60 2.5.4 Survey of neonatal salivary nucleotide metabolites ...... 60 2.5.5 Infant longitudinal study ...... 63 2.5.6 Animal salivary nucleotide metabolites ...... 66

2.6 Discussion ...... 69 2.6.1 Collection of saliva samples ...... 69 2.6.2 HPLC-MS/MS determination of nucleotide metabolites ...... 70 2.6.3 Adult saliva ...... 70 2.6.4 Infant saliva ...... 71 2.6.5 Transitioning from infant to adult: The longitudinal study ...... 72 2.6.6 Salivary patterns in other mammals ...... 73 2.6.7 Conclusions and further hypotheses ...... 74

Chapter 3: Human milk xanthine oxidase and neonatal salivary nucleotide metabolites: Their role in generating hydrogen peroxide in the neonatal mouth ...... 75

3.1 Introduction ...... 76

xii

3.2 Specific hypothesis and rationale ...... 77

3.3 Specific aims ...... 77

3.4 Materials and Methods ...... 78 3.4.1 Clinical samples ...... 78 3.4.2 Reagent preparations ...... 78 3.4.3 Xanthine oxidase assays...... 79 3.4.4 Measurement of endogenous hydrogen peroxide in human milk ...... 80 3.4.5 Generation of hydrogen peroxide by mixing human milk with neonatal saliva ...... 81

3.5 Results ...... 82 3.5.1 Breast milk xanthine oxidase kinetics ...... 82 3.5.2 Endogenous hydrogen peroxide in breast milk ...... 88 3.5.3 Hydrogen peroxide generation ...... 88

3.6 Discussion ...... 91 3.6.1 XO activity and kinetic parameters ...... 91 3.6.2 XO activity in infant formula and pasteurised milk ...... 92 3.6.3 Endogenous breast milk hydrogen peroxide ...... 93 3.6.4 Generation of hydrogen peroxide in neonatal mouth ...... 94

Chapter 4: An innate immunity mechanism: Regulation of the neonatal microbiome by breast milk and saliva ...... 96 4.1 Introduction ...... 97 4.1.1 The role and regulation of commensal bacteria ...... 97 4.1.2 Effects of formula versus breast feeding on neonatal gut microbiota ...... 98 4.1.3 Oxidative mechanisms for oral microbial regulation in neonates ...... 98 4.1.4 Oxidative mechanisms for mammary glands innate immunity ...... 99 4.1.5 Growth stimulation mechanisms for microbial regulation ...... 100

4.2 Specific hypotheses and rationale ...... 101

4.3 Specific aims ...... 101

4.4 Materials and methods ...... 102 4.4.1 Nucleotide metabolite preparations ...... 102 4.4.2 Hydrogen peroxide standardization ...... 102 4.4.3 Preparation of biological samples ...... 102 4.4.4 Bacterial preparations ...... 104

xiii

4.4.5 Determination of purine and pyrimidine metabolites in bacterial media using LC-MS . 104 4.4.6 In vitro oral microbiota studies ...... 105 4.4.6.1 Hydrogen peroxide titration against bacteria ...... 105 4.4.6.2 Effects on bacterial growth of saliva and breast milk...... 105 4.4.7 In vivo oral microbiota studies of breast-fed versus formula-fed neonates ...... 106 4.4.7.1 Subjects and specimen collection ...... 106 4.4.7.2 DNA extraction and purification ...... 106 4.4.7.3 PCR amplification of the bacterial 16S rRNA ...... 107 4.4.7.4 Sample QC post PCR purification ...... 108 4.4.7.5 16S rRNA gene sequencing ('deep sequencing') ...... 108 4.4.7.6 16S rRNA gene sequence analysis ...... 108 4.4.7.7 Statistical analysis ...... 109

4.5 Results ...... 111 4.5.1 Supplemented saliva preparation ...... 111 4.5.2 Determination of purine and pyrimidine metabolites in bacterial growth media ...... 113 4.5.3 In vitro oral microbiota studies ...... 114 4.5.3.1 Hydrogen peroxide titration against bacteria ...... 114 4.5.3.2 Bacterial growth with saliva and breast milk ...... 116 4.5.4 In vivo oral microbiota studies of breast-fed versus formula-fed neonates ...... 120 4.5.4.1 Participants for in vivo studies of oral bacteria ...... 120 4.5.4.2 Breast-fed and formula-fed associated oral taxa detected using 16S rRNA 'next- generation deep sequencing' ...... 121 4.5.4.3 Phylum level comparison of breast-fed and formula-fed oral bacterial communities ...... 121 4.5.4.4 Family level comparison of breast-fed and formula-fed oral bacterial communities ...... 123

4.6 Discussion ...... 125 4.6.1 Hydrogen peroxide titration against bacteria ...... 125 4.6.2 Bacterial growth with saliva and breast milk ...... 126 4.6.3 In vivo oral microbiota studies of breast-fed versus formula-fed neonates ...... 128

Chapter 5: Conclusions and future directions ...... 131

5.1 Conclusions ...... 132

5.2 Future studies ...... 133

xiv

5.3 Summary of research rationale, outcomes and benefits ...... 135 Bibliography ...... 136 Appendices ………………………………………………………………………………………..158

xv

LIST OF FIGURES

Figure 1.1‎ Purine bases and pyrimidine bases...... 3 Figure 1.2 Examples of nucleosides...... 4 Figure 1.3 Examples of nucleotides...... 5 Figure 1.4 Examples of coenzyme nucleotides...... 6 Figure 1.5 Dietary metabolism of purines and pyrimidines...... 8 Figure 1.6 De novo synthesis of purine nucleotides...... 10 Figure 1.7‎ Catabolism and salvage pathway of purines...... 12 Figure 1.8‎ Pyrimidine de novo synthesis...... 14 Figure 1.9 Overview of pyrimidine metabolism...... 16 Figure 1.10 The salvage pathways of external nucleosides and nucleobases in microorganisms with (e.g. L. lactis)...... 18 Figure 1.11 The three structural domains of xanthine protein...... 19 Figure 1.12 Reactions catalysed by XDH and XO forms of xanthine oxidoreductase...... 20 Figure 1.13 Positive reaction for mouse monoclonal anti-human XO antibodies in small intestine...... 21 Figure 1.14 Xanthine oxidase catalyses production of NO and peroxynitrite...... 22 Figure 1.15 Metabolism of ROS in normal state...... 24 Figure 1.16 ROS generation and health...... 24 Figure 1 .17 The human salivary glands ...... 25 Figure 1.18 Saliva composition...... 26 Figure 1.19 Glandular contribution to unstimulated salivary flow ...... 30 Figure 2.1‎ Gradient profile of mobile phase B that was used for reversed phase HPLC elution of purine and pyrimidine nucleosides and bases followed by MS/MS analysis...... 47 Figure 2.2 Chromatograms show UV-absorbing peaks at 268 nm present in 100 µL distilled water extracted from: (a) a commercial cotton tip, (b) a Sorbette sponge, and (c) a commercial cotton tip following washing with water, and ...... 49 Figure 2.3 The “best-fit” curve for the percent recovery of major nucleotide metabolites using extraction with water followed by centrifugation (Method #1)...... 52 Figure 2.4‎ Percent recovery of major nucleotide metabolites using extraction with solvent and evaporation followed by centrifugation (Method #2)...... 53 Figure 2.5‎ The results of linear regression analysis of the calibration data in the measurement of nucleosides and bases extracted from cotton swabs using the solvent and evaporation method (Method #2)...... 54 xvi

Figure 2.6‎ An example of LC-MS/MS chromatograms for purine and pyrimidine metabolites and creatinine in neonatal saliva...... 58 Figure 2.7‎ Distribution and median values of nucleotide metabolites in adults and neonatal whole saliva...... 61 Figure 2.8‎ Median concentrations of nucleotide metabolites in whole saliva of full-term infants 1-4 days (n=60), 6 weeks (n=20), 6 months (n=19), 12 months (n=14)...... 63 Figure 2.9 Median concentrations of the major nucleotide metabolites in whole saliva of 10 followed up full-term infants at 1-4 days, 6 weeks, 6 months, 12 months...... 64 Figure 2.10‎ Mammalian salivary nucleotide metabolite concentrations...... 67 Figure 3‎ .1 The lactoperoxidase system...... 76 Figure 3.2‎ Principle of coupled enzymatic assays of XO using horseradish peroxidase-Ampliflu Red reagent...... 80

Figure 3.3‎ Standard curve for H2O2 concentration versus fluorescence...... 83

Figure 3.4‎ Kinetics of H2O2 production by XO during conversion of hypoxanthine to . ... 83 Figure 3.5‎ Hyperbolic nonlinear regression curve showing XO velocity as a function of increasing hypoxanthine concentration...... 84 Figure 3.6‎ Distribution of XO activity in breast milk...... 85 Figure 3.7‎ Scatterplots of XO activity versus the gestational and postpartum age in hour...... 86 Figure 3.8‎ Dose dependent inhibition of breast milk XO by oxypurinol...... 87

Figure 3.9‎ Distribution of concentrations of H2O2 in fresh breast milk samples...... 89

Figure 3.10‎ Scatterplot of XO activity versus endogenous H2O2 concentration in 21 of 24 breast milk samples...... 89

Figure 3.11‎ Kinetics of H2O2 generation after mixing 33 µL of diluted breast milk, 33 µL of 1/3 diluted neonatal saliva and 34 µL peroxidase assay working solution (green circles). .... 90

Figure 3.12‎ Assessment of H2O2 consumption by breast milk...... 90 Figure 4.1‎ Schematic diagram summarises the overall bacterial experimental plan...... 110 Figure 4.2‎ Chromatograms showing UV-absorbing peaks at 268 nm present in “supplemented simulated neonatal saliva” preparations...... 112

Figure 4.3‎ Effect of increasing H2O2 (0-400 µM) concentration on bacterial growth...... 115 Figure 4.4‎ The effect of breast milk plus simulated neonatal saliva (with and without supplements) on bacterial viability...... 118

Figure 4.5 Bactericidal effects of H2O2 generated by breast milk (XO) and increasing concentrations of salivary xanthine and hypoxanthine (0-50 µM)...... 119 Figure 4.6 Bar graph showing the relative distributions of the major oral phyla operational taxonomic units (OTUs)...... 122 xvii

Figure 4.7 Relative abundances of the most common Firmicutes, Actinobacteria, and Bacteroidetes families in buccal swabs of breast-fed vs formula-fed infants...... 124

xviii

LIST OF TABLES

Table 1.1 Electrolyte and total protein concentrations in whole human saliva and plasma...... 27 Table 2.1‎ Physical information about the participated neonates...... 42 Table 2.2‎ Percent recovery of nucleosides and bases from cotton swabs using centrifugation (Method #1)...... 51 Table 2.3‎ Percent recovery of nucleosides and bases from cotton swabs using the solvent and evaporation (Method #2)...... 51 Table 2.4‎ The results of linear regression analysis of the calibration data...... 55 Table 2.5‎ Within-run precision data...... 56 Table 2.6‎ Limit of Quantitation (LOQ) and Limit of Detection (LOD) ...... 57 Table 2.7‎ HPLC-MS/MS conditions and parameters for optimal separation of nucleosides and bases in saliva...... 59 Table 2.8‎ Salivary levels (median with range, µM) of nucleotide metabolites in neonates compared to adults ...... 62 Table 2.9‎ Salivary levels (median with range, µM) of nucleotide metabolites in a longitudinal study of infants’ saliva...... 65 Table 2.10‎ Salivary levels (median with range, µM) of nucleotide metabolites in a selection of mammals...... 68 Table 3.1‎ Information about breast milk samples and donor mothers ...... 78 Table 3.2‎ XO kinetic parameters of breast milk samples (n=6), using hypoxanthine as . .. 84 Table 4.1‎ Saliva experimental solutions...... 103 Table 4.2‎ Fusion primers design (5' to 3') for 16S rRNA amplification...... 107 Table 4.3 Concentration of nucleotide metabolites in Nutrient and MRS broth (µM) as determined by LC-MS ...... 113 Table 4.4‎ Infants’ mean gestation and weight at delivery and the infants’ mean age at the time of collection of the oral bacteria samples by buccal swabs ...... 120

xix

LIST OF APPENDICES

Appendix 1 Mater Health Services Human Research Ethics Committee Approval...... 159 Appendix 2 University of Queensland Human Research Ethics Committee Approval...... 161 Appendix 3 Mater Health Services Human Research Ethics Committee Approval...... 162 Appendix 4 University of Queensland Human Research Ethics Committee Approval...... 164 Appendix 5 Mater Health Services Human Research Ethics Committee Approval...... 165 Appendix 6 University of Queensland Human Research Ethics Committee Approval...... 167 Appendix 7 Golden Casket research fund approval...... 168 Appendix 8 NHMRC research fund approval...... 170 Appendix 9 Paper published in J Chromatogr B, 2013. 931C: p. 140-147...... 172 Appendix 10 Abstract Published in Journal of Inherited Metabolic Disease...... 180 Appendix 11 A poster presented at SSIEM, 2011, Geneva...... 181 Appendix 12 An abstract and a poster presented at ASBMB ComBio2012...... 182 Appendix 13 A poster presented at ASBMB ComBio, 2012, Adelaide...... 183 Appendix 14 Abstract Published Journal of Paediatrics and Child Health ...... 184 Appendix 15 A poster presented at PSANZ, 2013, Adelaide...... 185 Appendix 16 Raw data set of adult salivary nucleotide precursors concentrations (µM)...... 186 Appendix 17 Raw data set of neonatal salivary nucleotide precursors concentrations (µM)...... 189 Appendix 18 Raw data set of cow salivary nucleotide precursors concentrations (µM)...... 191 Appendix 19 Raw data set of sheep salivary nucleotide precursors concentrations (µM)...... 192 Appendix 20 Raw data set of goat salivary nucleotide precursors concentrations (µM)...... 193 Appendix 21 Raw data set of horse salivary nucleotide precursors concentrations (µM)...... 194 Appendix 22 Raw data set of one camel salivary nucleotide precursors concentrations (µM)...... 195 Appendix 23 Raw data set of dog salivary nucleotide precursors concentrations (µM)...... 196 Appendix 24 Raw data set of cat salivary nucleotide precursors concentrations (µM)...... 197 Appendix 25 Gel electrophoresis of bacterial 16S rRNA gene amplicons ...... 198 Appendix 26 Summary of the de novo assembly statistics of 16S rRNA deep sequencing ...... 199 Appendix 27 16S rRNA sequence dataset ...... 200

xx

LIST OF ABBREVIATIONS

AICAR 5-aminoimidazole-4-carboxamide AIR 5-aminoimidazole ATP CAIR Carboxyaminoimidazole ribonucleotides cAMP Cyclic CDA deaminase CDP CDP Cytidine diphosphate CFU Colony forming unit cGMP Cyclic CMP CTP dCK dNTP Deoxynucleoside triphosphate dTDP Deoxy- diphosphate dTMP Deoxy- dTTP Deoxy- FAD Flavin dinucleotide FAICAR N-formyl-AICAR FGAM Formylglycinamidine ribonucleotides FGAR Formylglycinamide ribonucleotides FMP Flavin monophosphate GAR Glycinamide ribonucleotide GIT Gastrointestinal tract GMP Guanosine monophosphate GTP HPLC-PDA High performance liquid chromatography with photodiode-array IMP Inosine monophosphate LOD Limit of detection LOQ Limit of quantitation LPO Lactoperoxidase MFGM Milk fat globule membrane

xxi

Mo n.a. Not applicable n.s. Not significant NAD adenine dinucleotide NEC Necrotising enterocolitis NADP Nicotinamide adenine dinucleotide NTP Nucleotide triphosphate OMP Orotidine 5-monophosphate OTU Operational taxonomic unit PCR Polymerase chain reaction PNP Purine nucleoside PPi Inorganic phosphate PRA Phosphoribosyl-amine PRPP 5-phosphoribosyl-1- RNR RNS Reactive nitrogen species ROS Reactive oxygen species SAICAR N-succinylo-5-aminoimidazole carboxamide ribonucleotide UDP diphosphate UDP UK UMP Uridine 5-monophosphate UMPH Uridine 5’ phosphate UTP XDH XO Xanthine oxidase

xxii

Chapter 1

Introduction

1

1.1 THESIS STRUCTURE

This thesis is divided into five chapters, followed by References and Appendices. The introductory chapter presents general information and a literature review of the research project. Overall research Hypothesis and Aim, Funding sources and Ethical approvals are summarised at the end of this chapter. Chapters 2 to 4 present the three distinct but inter-related research studies that were undertaken. Each has an additional short Introduction to introduce the study and to cover the variety of topics in this project. Specific Hypotheses and Aims of each study are also presented, followed by Materials and Methods used, then Results, and finally a Discussion has been included in each chapter separately. Chapter 5 presents general Conclusions summarising the important findings of this project as well as a focus on the future directions for this research.

2

1.2 GENERAL LITERATURE REVIEW

1.2.1 PURINE AND PYRIMIDINE BASES AND NUCLEOSIDES Purines and pyrimidines are nitrogenous bases that are an essential part of nucleotide structures. The major purine and pyrimidine bases are shown in (Fig. 1.1). Purine bases have a double heterocyclic ring structure including 4 nitrogens. Adenine and are the major purine bases in nucleic acids (DNA, RNA), while hypoxanthine, xanthine and uric acid are components of the catabolic pathway. On the other hand, pyrimidine bases have a single heterocyclic ring containing 2 nitrogen atoms, with the major pyrimidine bases being uracil, and [1, 2].

Figure ‎1.1 Purine bases (left) and pyrimidine bases (right).

Purine and pyrimidine bases attached to either or 2- are called nucleosides or deoxynucleosides respectively. Purines link to the moiety by β-N-glycosidic bonding at position N-9, while pyrimidines have a at position N-1 (Fig. 1.2). Adenosine, guanosine and are examples of purine nucleosides, while uridine, cytidine, and deoxythymidine are example of pyrimidine nucleosides [3].

3

Figure 1 .2 Examples of nucleosides. Purine nucleosides: (1) adenosine; (2) guanosine; (3) deoxyadenosine. Pyrimidine nucleosides: (4) cytidine; (5) uridine; (6) (deoxy-) thymidine.

1.2.2 PURINE AND PYRIMIDINE NUCLEOTIDES Nucleotides are nucleosides with one or more phosphate groups attached to the sugar moiety at the 5’-position (Fig.1.3). In normal cells, triphosphate and diphosphate nucleotides exist in higher concentrations than monophosphate nucleotides, nucleosides, and free bases [2]. Similarly, the concentration of (NTP) is higher than deoxynucleoside triphosphate (dNTP) in non-dividing cells, as the latter is needed only for the synthesis of DNA [4]. Nucleic acids (RNA or DNA) are polymers of nucleoside or deoxy-nucleoside monophosphates, respectively.

In addition to the role of purine nucleotides in the synthesis of genetic material these compounds have central roles in many metabolic processes, e.g. adenosine 5-triphosphate (ATP) functions as a carrier of cellular and mitochondrial chemical energy [5]. Guanosine triphosphate (GTP) is important for cellular synthetic reactions, as well as regulating with G-proteins to drive transmission 4 of stimuli across cell membranes; many hormones, neurotransmitters, local mediators exert their function by coupling to G-proteins [6]. Cyclic nucleotides such as cAMP and cGMP are other forms of nucleotides which function as second messengers in numerous cellular metabolism [7].

Besides their function in nucleic acids, pyrimidine nucleotides play an important role in synthetic cellular metabolism. For example, UDP- are required for protein and synthesis, while CDP conjugates with various lipids, e.g. CDP- and CDP diacylglycerol phosphoglyceride, for synthesis of cell membranes [8, 9].

Figure 1 .3 Examples of nucleotides.

1.2.3 COENZYME NUCLEOTIDES The coenzymes nicotinamide adenine dinucleotide (NAD), nicotinamide adenine dinucleotide phosphate (NADP), flavin mononucleotide (FMN) and flavin adenine dinucleotide (FAD) are specialised electron-transfer nucleotides, all derived from ATP (Fig. 1.4) [10, 11].

5

Figure 1.4 Examples of coenzyme nucleotides. Nicotinamide adenine dinucleotide (NAD); Flavin adenine dinucleotide (FAD).

1.2.4 NUCLEOTIDE CELLULAR FUNCTIONS Purine and pyrimidine nucleotides are formed by either de novo synthesis pathways, or by salvage of catabolised bases and nucleosides. Because nucleotides are so central to cellular function, defects in the metabolic pathways of nucleotides can lead to severe metabolic diseases, resulting in immunodeficiency, anaemia, neuropathologies, and renal disease [12, 13].

The concentrations of nucleotides in tissues vary widely reflecting specialised cellular function. The nucleotides in red blood cells rely on the salvage paths and are primarily concerned with synthesising ATP, NAD and NADP [14]. In contrast, ADP occurs in high concentrations in platelets, being part of the crucial mechanism mediating platelet aggregation [15]. In , a variety of nucleotides and their sugar and derivatives are present for use in biosynthetic processes. In non-dividing cells the concentration of nucleotides in the form of ribonucleotides is higher than for 6 deoxyribonucleotides, but when DNA replicates before cell division the concentrations of deoxy- ribonucleotides are highly elevated to supply the cell with the required substrates for DNA synthesis [16].

Dietary DNA and RNA are fully hydrolysed to mono-nucleotides in the small gut, then converted to nucleosides by the action of in the enterocyte. Dietary purines and pyrimidines are then handled differently by the gut (Fig. 1.5). Purines are mainly degraded via mucosal xanthine oxidase to uric acid before entry into the circulation, although the gut mucosal cells may use some dietary purines for their own metabolism [17]. In contrast, dietary pyrimidines are directly absorbed by the gut and passed into the circulation; pyrimidine salvage – or further catabolism – occurs mainly in liver [12]. Purine and pyrimidine nucleotides are also metabolised by the intestinal bacteria either for incorporation into their own nucleotides and nucleic acids, or for energy- releasing catabolism (see Section 1.2.7).

7

Figure 1.5 Dietary metabolism of purines and pyrimidines. Abbreviations: XO, xanthine oxidase; DPD, dihydropyrimidine dehydrogenase; DHP, ; UP, Ureidopropionase. Diagram from Loffler et al. [12] with modification.

8

1.2.5 METABOLISM OF PURINE NUCLEOTIDES

1.2.5.1 De novo synthesis De novo synthesis of purine nucleotides is initiated when 5-phosphoribosyl-1-pyrophosphate (PRPP) is synthesised from ribose 5-phosphate, by the PRPP synthetase (PRPS1: OMIM *603762). PRPP is also important for biosynthesis of pyrimidine nucleotides as well as nicotinamide coenzymes and the amino acids histidine and tryptophan [18].

The committed step in purine synthesis is the conversion of PRPP to 5- in the presence of , catalysed by glutamine PRPP amidotransferase (PPAT: OMIM *172450) [19]. A 10-step pathway then leads to formation of the nucleotide inosine monophosphate (IMP), which is then converted either to AMP or GMP and subsequently to ATP and GTP by [2]. Note that the de novo synthesis of purine is expensive in term of energy utilization, since 6 moles of ATP is consumed per mole of purine formed [20]. See Figure 1.6 for full purine de novo synthesis steps including enzyme nomenclature.

1.2.5.2 Synthesis of deoxyribonucleotides An important property of purines and pyrimidines is their ability to form hydrogen-bonded pairs composed of one purine and one pyrimidine, such as guanine-cytosine and adenine-thymine, which is essential for double-stranded DNA. The substrates for both purine and pyrimidine deoxyribonucleotides are 5-diphosphates (NDP). Reduction at the 2-position of the ribose ring in NDPs leads to 2- formation, catalysed by ribonucleotide reductase (RRM2B: OMIM *604712). Deoxy-NDPs are then converted to the triphosphates by kinases, for use as substrates for DNA [16].

9

Figure 1.6 De novo synthesis of purine nucleotides. Enzymes: (1) Phosphoribosylpyrophosphate synthetase; (2) Glu:PRPP amidotransferase; (3) GAR synthetase; (4) GAR transformylase; (5) FGAM synthetase; (6) AIR synthetase; (7) AIR carboxylase; (8) SAICAR synthetase; (9) ; (10) AICAR transformylase; (11) IMP synthase; (12) adenylosuccinate synthase/ lyase; (13) IMP dehydrogenase/GMP synthetase; (14) nucleoside monophosphate kinase; (15) nucleoside diphosphate kinase; (16) ribonucleotides reductase; (17) AMP deaminase. Abbreviations: PRPP, 5-phosphoribosyl-1-pyrophosphate; PRA, phosphoribosyl-amine; GAR, glycinamide ribonucleotide; FGAR, formylglycinamide ribonucleotides; FGAM, formyl- glycinamidine ribonucleotides; AIR, 5-aminoimidazole ribonucleotides ; CAIR, carboxy- aminoimidazole ribonucleotides; SAICAR, N-succinylo-5-aminoimidazole carboxamide ribonucleotide; AICAR, 5-aminoimidazole-4-carboxamide ribonucleotide; FAICAR, N-formyl- AICAR; IMP, inosine monophosphate; (-) inhibition. Diagram from Camici et al. [19] with modification.

10

1.2.5.3 Salvage of purines The salvage of purine nucleotides is important to minimise energy consumption, as the salvage pathway consumes only 1 mole of ATP [21]. The salvage of purine nucleotides occurs principally from the bases via phosphoribosyltransferases (Fig.1.7). Hypoxanthine-guanine phosphoribosyltransferase (HPRT1: OMIM *308000) converts guanine and hypoxanthine to GMP and IMP, and adenine phosphoribosyltransferase (APRT:OMIM +102600) salvages adenine to AMP [22]. These salvage reactions are highly regulated by their nucleotide end-products which act as competitive inhibitors. AMP is also produced by of adenosine by adenosine kinase (ADK: OMIM *102750) [19]. Significantly, dietary purines are not used in the body; apart from purine bases salvaged by gut cells, all other purines in the diet are degraded to uric acid by the small intestine.

Purine salvage has a significant role in controlling the de novo synthesis pathway as the recycling of AMP and GMP inhibits the de novo PRPP amidotransferase step. Furthermore, consumption of PRPP, an activator for purine de novo synthesis, by the salvage pathway, reduces the activity of the synthetic pathway [19].

1.2.5.4 Catabolism of purines The turnover of cells includes the breakdown of nucleotides and nucleic acids. ATP and GTP are degraded to their mononucleotides, AMP and GMP. DNA and RNA are degraded by nucleases to monophosphates. Purine convert nucleotides into nucleosides, then purine nucleoside phosphorylase (PNP: OMIM *164050) converts these to purine bases (hypoxanthine and guanine) (Fig. 1.7). Importantly, adenosine is not a substrate for PNP; it is first converted to inosine by (ADA: OMIM *608958) [23]. Guanine is converted to xanthine by (GDA: OMIM *139260), while hypoxanthine and xanthine are converted to uric acid by xanthine oxidase [19].

The last step of purine degradation by xanthine oxidase is characterized by uric acid production associated with hydrogen peroxide (H2O2) generation [24]. In humans and higher apes, uric acid is the final of purine catabolism.

11

Figure ‎1.7 Catabolism and salvage pathway of purines. Abbreviations: IMP: inosine-5′-monophosphate; PRPP: phosphoribosylpyrophosphate; PPi: inorganic phosphate; APRT: adenine phosphoribosyltransferase; HGPRT: hypoxanthine-guanine phosphoribosyltransferase; AMPD: adenosine monophosphate deaminase; ADA: adenosine deaminase; 5’-NT: 5'-; PNP: Purine nucleoside phosphorylase; GMP: guanosine monophosphate; GDA: guanine deaminase; XO: xanthine oxidase. Note conversion of adenosine to AMP is catalysed by adenosine kinase. Diagram was adapted from http://themedicalbiochemistrypage.org/nucleotide-metabolism.html (accessed 14/3/2010) with modification.

12

1.2.6 METABOLISM OF PYRIMIDINE NUCLEOTIDES

1.2.6.1 De novo synthesis The de novo synthesis of pyrimidine nucleotides begins with 6 metabolic steps, shown in Figure 1.8. Initially, is synthesised by carbamoyl phosphate synthetase (CPS1: OMIM *114010) via glutamine, bicarbonate and 2 ATP molecules [12, 25]. Carbamoyl phosphate may then enter either the or the pyrimidine pathway, however some studies suggest that this enzyme works independently in both pyrimidine synthesis and urea cycle [26, 27]. The conversion of carbamoyl phosphate to orotate is catalysed by a tri-functional enzyme (CAD: OMIM *114010): carbamylphosphate synthase-aspartate transcarbamoylase-. Note that the next step, dihydroorotate dehydrogenase (DHODH: OMIM*126064) is the only mitochondrial enzyme in the de novo synthesis of pyrimidines; all other enzymes are located in the cytosol [9, 12].

Uridine 5-monophosphate (UMP) is the central pyrimidine nucleotide synthesised from orotate by a bi-functional enzyme orotate phosphoribosyltransferase – orotidine-5-monophosphate (OMP) decarboxylase, defined as UMP synthase (UMPS: OMIM +258900) [25]. In the first step, the base orotate is converted to the nucleotide OMP using PRPP as the ribose phosphate donor, then OMP is decarboxylated to uridine 5-monophosphate (UMP). Nucleotide kinases convert UMP to UTP. The latter can be converted to CTP by the action of CTP synthetase (CTPS: OMIM *123860) [9, 12, 28- 30]. The thymidine deoxy-nucleotides are synthesised via (TYMS: OMIM*188350) from UMP via a -dependent path. This latter reaction has become a target for the inhibitory drugs of DNA synthesis such as , , and [31].

13

Figure ‎1.8 Pyrimidine de novo synthesis. (1) carbamyl-phosphate synthetase; (2) aspartate transcarbamylase; (3) dihydroorotase; (4) di- hydroorotate dehydrogenase; (5) orotate phosphoribosyltransferase; (6) orotidine 5'-monophosphate decarboxylase; (7) UMP kinase; (8) UDP kinase; (9) CTP synthetase. Diagram from biochem4.okstate.edu/.../nucleotides.htm (accessed 25/4/2013).

14

1.2.6.2 Salvage of pyrimidines The cellular requirement for pyrimidine nucleotides is also met by the recycling (‘salvage’) of pyrimidine nucleosides [32]. Mitochondrial and cytosolic nucleoside kinases catalyse the phosphorylation of uridine, deoxycytidine and deoxythymidine to their corresponding nucleotides [12] (Fig. 1.9). Uridine is central to the salvage of normal nucleotides, while deoxy-nucleotide salvage is reserved for few specialised functions.

1.2.6.3 Catabolism of pyrimidines Similar to purines, pyrimidine nucleotides and nucleic acids are initially broken down to mononucleotides which are then dephosphorylated to nucleosides by pyrimidine nucleotidases. Cytidine is converted to uridine by the action of (CDA: OMIM *123320). The degradation of uridine by (UPP1: OMIM *191730) to uracil and the degradation of thymidine by (TYMP: OMIM *131222) to thymine are the first committed steps of the catabolism of pyrimidines in humans [12]. Dihydropyrimidine dehydrogenase (DPYD: OMIM*612779) then catalyses the reduction of thymine and uracil to and respectively. Dihydropyrimidinase (DPYS: OMIM *613326) catalyses the hydrolysis of these dihydropyrimidines to ureidoisobutryic and ureidopropionic acid, respectively. Finally, these are broken down by ureidopropionase (UPB1: OMIM *606673) to form

β-aminoisobutyrate and β-alanine, releasing ammonia and CO2 [29]. As the end-product of thymine degradation, β-aminoisobutyric acid has been measured in urine to estimate DNA turnover [33, 34]. The catabolism of pyrimidines mainly occurs in liver; however Van Kuilenburg and co-workers (2006) [29] suggested that it may also involve other organs such as kidney and spleen. Pyrimidine metabolic pathways are illustrated in Figure 1.9.

15

Figure 1.9 Overview of pyrimidine metabolism. Abbreviations: UMP: ; UDP: Uridine diphosphate; UTP: Uridine triphosphate; CMP: Cytidine monophosphate; CDP: Cytidine diphosphate; CTP: Cytidine triphosphate; dTMP: Deoxy- thymidine monophosphate; dTDP: Deoxy- ; dTTP: Deoxy- thymidine triphosphate. RNR: Ribonucleotide reductase; CDA: Cytidine deaminase; UMPH: uridine 5’ phosphate hydrolase; UK: uridine kinase; dCK: ; URH: Uridine nucleosidase; TP: Thymidine phosphorylase; TK: ; DPD: dihydropyrimidine dehydrogenase; DPA: di- hydropyrimidinase; UP: Ureidopropionase. Diagram from Loffler et al. [12], Nyhan [13], and Zrenner et al. [35] with modification.

16

1.2.7 BACTERIAL UPTAKE AND METABOLISM OF PURINES AND PYRIMIDINES While most bacteria can synthesise nucleotides de novo, some, including lactic acid bacteria (e.g. Lactococcus lactis), require external sources of purines or pyrimidines to enhance their growth in synthetic media [18, 36]. These auxotrophic bacteria utilise salvage pathways to convert exogenous bases or nucleosides to their corresponding nucleotides (Fig. 1.10). In addition, exogenous nucleotides can be used by some types of bacteria as a source for purines or pyrimidines, with these nucleotides being dephosphorylated by extracellular nucleotidases prior to entering the cell [18].

However, prototrophic bacteria can also salvage nucleobases and nucleotides; the de novo pathways in these bacteria are usually silenced in the presence of excess external purine or pyrimidine sources [18]. For example, exogenous uracil, uridine, , cytosine, cytidine, and deoxycytidine, thymine and thymidine are taken up by Escherichia coli and Salmonella typhimurium from their host.

Uptake mechanisms show great diversity. L. lactis has a high affinity uptake system for uracil but not cytosine and thymine [18, 37-39], as with Neisseria meningitidis which utilises only uracil but not uridine, deoxyuridine, cytosine, cytidine, or deoxycytidine suggesting this bacterium lacks the appropriate nucleotide salvage pathways [40].

Many other bacteria such as Lactobacillus species have additional requirements for deoxynucleotides, as they lack the ability to reduce ribonucleotides to deoxyribonucleotides, which is crucial for DNA synthesis. Rather, they use a salvage enzyme, trans-N-deoxyribosylase [18], which catalyses the exchange reaction of the base moiety of 2′- [41].

Importantly, some strains of Lactobacillus bulgaricus and Lactobacillus jugurti were found to be stimulated by bovine milk which contains high concentrations of that is an important intermediate in pyrimidine synthesis. Growth of these same bacteria was inhibited in orotate-free milk suggesting that orotic acid is used for pyrimidine nucleotide synthesis [42]. It has been found that pyrimidine de novo synthesis was inhibited as a result of long habitation of these bacteria in milk containing orotic acid, while this does not support other auxotrophic bacteria such as L. lactis and Streptococcus thermophilus, which were characterised by inability to grow in milk [18].

In addition to purine and pyrimidine salvage by bacteria, some bacteria such as E. coli use nucleotides very efficiently as carbon and energy sources, for example degradation of the ribose sugar of the nucleosides. Bacteria may also utilise the amino group in cytidine and adenosine as a nitrogen source without degradation of the purine and pyrimidine rings [18].

17

Figure 1.10 The salvage pathways of external nucleosides and nucleobases in microorganisms with nucleoside phosphorylases (e.g. L. lactis). Diagram from Kilstrup, M et al. [18] with modification.

18

1.2.8 XANTHINE OXIDASE

1.2.8.1 General characteristics Xanthine oxidase (XO; OMIM-607633) is widely distributed in many species of mammals and bacteria [43, 44], but in humans XO distribution is restricted. The enzyme shows highest activity in intestine and liver [45], with low levels of XO RNA reported in the human heart, brain, lung and kidney [46]. Surprisingly, XO is present in high activity in the milk of mammals where it is associated with the milk fat globule membrane (MFGM) [24, 47, 48].

1.2.8.2 Structure and reaction mechanism Xanthine oxidoreductase (XOR) exists in two protein forms of the same gene product. After proteolysis by trypsin digestion, the protein has been shown to comprise a complex of three linked structural domains [43] (Fig. 1.11).

Figure 1.11 The three structural domains of xanthine oxidoreductase protein [49]. Domains of 20, 40, and 85 kDa, associated with an FeS domain containing two Fe2S2 groups, a flavin domain containing the FAD , and a molybdenum binding domain with the respectively. Abbreviations: Mo, molybdenum; FAD, flavin adenine dinucleotide; Fe-S, sulfate.

The two forms of the enzyme are referred to as XO and xanthine dehydrogenase (XDH). Degradation of xanthine and hypoxanthine to uric acid is accompanied by reduction of oxygen to + H2O2 for XO, while XDH reduces NAD to NADH [50]. A summary of the reaction is illustrated in Figure 1.12. The XO gene is located in human 2p22 and is large at 60kb in length containing 36 exons and 35 introns [51] corresponding to 1330-1335 amino acids.

The oxidation reaction occurs at the molybdenum (Mo) centre where electrons are donated from the substrate accompanied by the reduction of Mo (VI) to Mo (IV) [52, 53]. Electrons are then passed + through the Fe2S2 cluster to either NAD (for XDH), or to oxygen at the FAD site (for XO) [24]. Unlike cytochrome P450 oxidation reactions, the oxygen that accepts electrons from xanthine and hypoxanthine substrates comes from a water and not from molecular oxygen [53, 54]. 19

It has been shown that the XDH form exists predominant intracellularly, but it is converted to XO by extracellular oxidation of the free sulfhydryls to disulfides, with the process being reversible by sulfhydryl-reactive reagents suggesting that the disulfide bond in cysteine is associated with this phenomenon. Cysteine oxidation leads to modification of the flavin site and loss of NAD binding [43, 55]. Nishino and co-workers reported that rat liver XDH is converted to XO by modification of Cys535 and Cys992 residues [56]. XDH can also be converted irreversibly to XO by proteolysis [44, 57].

The enzyme present in milk appears to exist exclusively in the XO form, presumably following proteolysis of XDH before secretion. It has been shown that XO from human breast milk has activity of only about 5% that of purified bovine milk XO, due to low molybdenum (Mo) content of the human form. Purified human milk XO was reported to comprise at least 95% demolybdo- enzyme, compared with 30–40% for the bovine milk enzyme [58].

Figure 1.12 Reactions catalysed by XDH and XO forms of xanthine oxidoreductase.

20

1.2.8.3 Antibacterial properties of xanthine oxidase Xanthine oxidase (XO) generates reactive oxygen species (ROS) and reactive nitrogen species (RNS) [59], which are known to act as destructive agents in many situations, especially in ischemia- [60-63]. However, ROS also have been long recognised as antiseptic agents, with peroxide being used also for sterilising and bleaching. XO is highly expressed in the upper small intestine, especially in basal and apical layers [64, 65], as evidenced by assays of the intestinal villi and positive immunohistochemistry staining using monoclonal anti-human milk XO antibodies (Fig. 1.13). The presence of XO in the upper small intestine particularly in the surface epithelial cells suggests an antimicrobial role by the production of peroxide. This hypothesis was supported from a study by Van Den Munckhof who found bacteria were killed by XO in the rat digestive tract [66]. Furthermore, XO has been shown to be present at the luminal surface of bile duct and in biliary secretions where it may also act as an antibacterial agent to protect the biliary system [24, 47]. Gut XO also appears to function as a barrier against dietary purines, which are converted to uric acid when crossing the small intestine.

Figure 1.13 Positive reaction for mouse monoclonal anti-human XO antibodies in small intestine. Arrows show that XO is localised predominantly from mid-villi to the tip [67].

In mammalian milk XO is a major protein component of milk fat globule membrane (MFGM). MFGM is synthesised from fat droplets, which originate in the endoplasmic reticulum of secretory mammary gland cells. The fat droplets are then enveloped by the apical cell membrane before release [68, 69]. Interestingly, mammary glands in mice showed an increase in XO mRNA transcription during pregnancy and lactation [70].

There are also anecdotal but widespread “mid-wives’ tales” alluding to the antiseptic properties of breast milk, for example in treating eye infections in babies or for sores on the breast (from: http://thetruthaboutbreastfeeding.com/2009/10/11/can-breast-milk-cure-an-eye-infection/, accessed

21

1/8/2013). Purified bovine milk XO has been shown to inhibit the growth of several bacteria species, including the clinically important and ubiquitous Staphylococcus aureus [67]. Stevens and co-workers found that breast milk inhibited the growth of E.coli and Salmonella enteritidis, particularly following the addition of hypoxanthine (a substrate for xanthine oxidase) to milk- supplemented media [71].

- Under anaerobic conditions, XO can also reduce (NO2 ) at the molybdenum site [72, 73]. The product, nitric oxide (NO), in low concentration has important physiological roles of vasodilation, neurotransmission and platelet aggregation [74]. But in high concentration NO is toxic and inactivates many enzymes [75]. generated in the presence of molecular oxygen can react rapidly with NO to yield peroxynitrite ONOO-, which also is a powerful antibacterial agent

[24]. Even though superoxide and H2O2 generation is markedly stimulated above pH 7 in the presence of oxygen, Stevens and co-workers [71] found that neonatal gastric conditions (pH 4 to pH 6, with an oxygen tension less than 5%) are sufficient for XO to generate NO and consequently peroxynitrite (Fig. 1.14).

Figure 1.14 Xanthine oxidase catalyses production of NO and peroxynitrite. Diagram was adapted from [24]. Abbreviations: Mo, molybdenum; Fe2S2, iron sulfate; FAD, flavin adenine dinucleotide; NAD, nicotinamide adenine dinucleotide; NO2-, nitrite; NO, nitric oxide; ONOO-, peroxynitrite.

Furthermore, peroxide-free radicals can combine with thiocyanate (SCN-), catalysed by the enzyme lactoperoxidase (LPO), which is also present in milk and saliva, to produce a more potent anti- bacterial free hypothiocyanite (OSCN-); this is known as the 'lactoperoxidase system' [76- 78]. In addition, the role of XO in milk secretion is considered to be primarily to produce ROS, which then act as an antibacterial agent in both the milk glands to prevent mastitis, and in the milk itself to inhibit microbial growth [79].

22

1.2.9 PHYSIOLOGY OF REACTIVE OXYGEN SPECIES ROS are mediated metabolically in the human body by enzymes (e.g. NADPH , XO, nitric oxide synthetase, glutathione oxidase, cytochrome P450) as well as non-enzymatically by reactive compounds such as the semi-ubiquinones of mitochondria produced as a consequence of oxidative phosphorylation processes, and by peroxisomes as metabolic by-products [80]. Mitochondria are the major source of cellular ROS, especially during pathological conditions, when electrons leak.

.- In cells the initial reaction of ROS is an electron transfer to oxygen to form superoxide (O2 ). (SOD) then converts superoxide to long-lasting and membrane-diffusible

H2O2 [81]. In the presences of reduced metals such as ferrous ions, H2O2 may split into two . molecules of a highly reactive hydroxyl radical ( OH); otherwise H2O2 is converted to water by catalase or glutathione peroxidase (Fig 1.15) [82]. Free radicals have been described as destructive agents for tissues and organs causing pathological conditions [83, 84]. At high concentrations, ROS can react with cellular proteins, lipids, and nucleic acids thereby inducing functional alterations, or complete destruction.

However, it is important to note that ROS have an essential role in human normal physiology. At moderate (micromolar) concentrations, ROS are critical for cellular proliferation, differentiation, growth regulation, apoptosis, and other second messenger signalling pathways [82, 85, 86] (Fig 1.16). In addition, they have vital roles in the innate immune system where they directly fight pathogens [59]. When exposed to pathogens, activated phagocytic white blood cells present in localised inflammation within tissues produce an oxidative burst of excess ROS as a part of the first line defence against microbial invasion [85]. Recently, it was suggested that cryptotanshinone, (a compound isolated from some medicinal plants) exhibits its antimicrobial activity against S. aureus strains by generating ROS [87]. Importantly, this molecule targets S. aureus isolates of which more than 60% are resistant to methicillin. Research in this area of utilising ROS activity is promising and aims to discover novel oxidative molecules that target antibiotic resistant bacterial strains, a serious clinical problem that is increasingly encountered.

Therefore, the term “oxidative regulation” might prove to be more precise than “” to describe the role of ROS in the body. An example of a revised role for ROS is illustrated by the interesting question, “Can promote diseases?” addressed by Perera and Bardeesy in their 2011 Nature paper: “Cancer: when antioxidants are bad?” [88]. This is an area of research finding increasing applications in cancer and other disease-related studies.

23

Figure 1.15 Metabolism of ROS in normal state. Abbreviations: XO, xanthine oxidase; SOD, superoxide dismutase; GSH, glutathione; GSSG, glutathione disulfide; NADPH, nicotinamide adenine dinucleotide phosphate. Diagram from Droge [82] with modification.

Figure 1.16 ROS generation and health. At micromolar concentrations, ROS have crucial roles in cells through mechanisms including signalling, biosynthetic processes, and host defence. When ROS levels are too low, the condition may result in decreased antimicrobial and anticancer defence and inappropriate signalling. When levels are too high there is increased non-specific tissue damage, resulting in a higher risk of cardiovascular disease, neurological damage, and chronic inflammation. Diagram from Brieger et al. [80].

24

1.2.10 PHYSIOLOGY OF SALIVA

1.2.10.1 Functions of saliva Saliva is a complex fluid secreted by 3 pairs of salivary glands (parotid, submandibular, sublingual) which are located under and to the side of the oral cavity, and which secrete saliva into the mouth via a series of short ducts (Fig. 1.17) [89]. Saliva has a variety of functions; see Figure 1.18 for a summary of salivary components and functions. Saliva lubricates and partially digests before swallowing and plays an important role in buffering the pH of the oral cavity [90].

The pH of saliva can range from 5.3 during low flow to 7.8 during peak salivary flow. Bicarbonate and phosphate are major buffering ions that neutralise acids formed by bacterial of sugar, which causes dental caries. Urea also plays a significant role in saliva buffering, as some oral bacteria can hydrolyse urea to ammonia and carbon dioxide, which consequently increases the pH, although elevation of ammonia in saliva can initiate gingivitis [91].

Figure 1.17 The human salivary glands – from http://www.oral-cancer.info/ (accessed 30/3/2013).

1.2.10.2 Protein components of saliva Saliva contains a variety of specialised proteins and enzymes. The immunologic proteins of saliva are IgA, IgM, and IgG, with secretory IgA having the highest concentration [92]. These immunoglobulins bind the bacteria and inhibit their adhesion to oral tissues. Lactoferrin and cystatines also play a role in inhibiting microbial activity. Non-immunologic antibacterial components of saliva include several enzymes, in particular peroxidase (which has a significant role in ROS generation), and lactoferrin. Salivary enzymes also include , lipase, protease, DNAse and RNAse [90, 91]. 25

Proline-rich proteins (PRPs) comprise approximately 70% of the total protein content of human parotid saliva [90]. PRPs inhibit spontaneous precipitation of phosphate salts on the tooth surface, preventing the formation of dental caries. These proteins also improve oral lubrication and selectively enhance bacterial adhesion to teeth surfaces [91]. Histatins, a class of protein rich in histidine and found in both parotid and submandibular secretions, reportedly show antimicrobial properties against the yeast Candida albicans [93]. These anti-Candidal properties of histatins were demonstrated when a recombinant adenovirus vector containing histatin-3 DNA was transferred into the rat parotid gland, histatin was secreted into rat saliva and demonstrated increased inhibitory activity against Candida compared to control animals [94].

Proteins of submandibular and sublingual saliva also include the mucins, which are large viscous glycoproteins comprising 2 groups: highly-glycosylated MG1 and single-glycosylated MG2 [90]. Mucins are characterised by low solubility, high viscosity, high elasticity, and strong adhesiveness [95]. They have a lubricating function when food is swallowed as well as protecting oral surfaces against friction, and similar to the role within the gut, they form a barrier between bacteria and epithelial cells in the mouth [91].

Figure 1.18 Saliva composition. Diagram from [96].

26

1.2.10.3 Electrolyte components of saliva The major component of saliva is water (99% w/w), with dissolved inorganic salts containing sodium, potassium, calcium, , bicarbonate, phosphate, and chloride. Table 1.1 contains summary data for electrolytes and total protein concentrations in whole human saliva compared with plasma. Note that while plasma ultrafiltrates are isotonic in the acinar cells of salivary glands, the energy-dependent reabsorption of Na+ and Cl- from the salivary ductal system results in saliva being a hypotonic secretion. This hypotonic state facilitates the function of and also allows hydration and expansion of mucin glycoproteins [91, 95, 97]. The thiocyanate ion (SCN-) is secreted in high concentrations in saliva - it is negligible in plasma (Table 1.1). Thiocyanate plays a role in combination with H2O2 and lactoperoxidase to produce the bacteriostatic nitrogen species, hypothiocyanate.

Table 1.1 Electrolyte and total protein concentrations in whole human saliva and plasma. From Aps - - and Martens, 2005 [97]. (NH2)2CO2, urea; SCN , thiocyanate; HCO3 , bicarbonate; NH3, ammonia.

Plasma Whole human resting Whole human stimulated saliva saliva Na+ (mM) 145 5 20–80 K+ (mM) 4 22 20

Ca2+ (mM) 2.2 1–4 1–4

Cl− (mM) 120 15 30–100

− HCO3 (mM) 25 5 15–80

Phosphate (mM) 1.2 6 4

Mg2+ (mM) 1.2 0.2 0.2

SCN− (mM) <0.2 2.5 2

NH3 (mM) 0.05 6 3

(NH2)2CO (mM) 2–7 3.3 2–4

Protein (g/l) 70 3 3

1.2.10.4 Oral microbiota The human body is host to many commensal microorganisms, with distinct communities at different anatomical sites [98, 99]. Recent studies have highlighted the importance of the gut microbiota to the host. These microorganisms contribute to the metabolism and supply of essential nutrients, they stimulate the immune system assisting in its development and the immune response, and these colonizers protect the host from invasive pathogens (see Section 4.1.1). The healthy oral cavity is constantly moist with a temperature range of 34oC to 36oC and an approximately neutral 27 pH [100]. These conditions provide an ideal environment for the growth of many microorganisms. Early studies reported that the oral cavity contained at least 300 cultivable bacterial species including mutans streptococci (Streptococcus mutans, S. sobrinus, S. cricetus, and S. rattus); S. sanguis, which are found in larger numbers on teeth than on soft tissues; and S. salivarius, which is located chiefly on the surface of the tongue [101, 102]. However, the use of culture-independent methods for determining the composition of the oral microbiota, allied with next generation DNA sequencing methods is providing a far deeper analysis, as a majority of bacteria have yet to be cultivated [103]. Recent studies using DNA-based methods have demonstrated the presence of non- cultivable bacterial strains, which indicates that the oral microbiome is far more diverse than originally thought. Most of these molecular studies are based on sequencing of the small subunit (16S) ribosomal RNA (rRNA) gene because of its universal presence in bacterial organisms, the presence of conserved regions and its reliability for phylogenetic analysis [104, 105]. Using 16S rRNA gene-based techniques, it is estimated that the bacterial communities found in the human mouth are highly complex with around 1000 species present [106]. The oral microbiome has also been shown to be the second most complex in the body, after the colon [98].

Recent molecular genetic studies of bacterial communities in the mouth have shown that the phyla Firmicutes, Bacteroidetes, Proteobacteria, Actinobacteria, Spirochaetes and Fusobacteria [106] are the most predominant; and the most predominant bacterial genera in the human adult oral cavity include Streptococcus, Gemella, Abiotrophia, Granulicatella, Rothia, Neisseria, Prevotella, Haemophilus and Lautropia. However, large differences within the community composition were reported between individuals [107]. Most oral diseases are caused by multiple bacteria, e.g. dental caries are caused by acid-tolerant bacteria: streptococcal species, lactobacilli, and the actinomyces [103]. Lesions of oral tissues thus may contain hundreds of bacterial species [108]. The oral cavity also accommodates other non-bacterial microbiome including yeasts, e.g. Candida albicans, and protozoans, such as Entamoeba gingivalis and Trichomonas tenax [103].

The gut microbiome has been intensively studied in infants (Section 4.1.1), however, there are few studies examining microbial communities in the mouths of neonates and the differences in the community structure between breast-fed and formula-fed infants. Bacteria from the mother’s vagina, gut, and skin may be transmitted to neonates during delivery. Importantly, breast milk affects the establishment of the microbiota in the mouth as well as in the gastrointestinal tract (GIT) by providing nutrition and bacteria, predominantly lactobacilli, bifidobacteria, and streptococci [109]. Interestingly, recent publications report that the oral microbiota in infants also differ depending on the mode of delivery. Significantly higher numbers of bacterial taxa were also 28 detected in the mouths of vaginally-delivered infants compared with infants delivered by Caesarean section [110]. The mode of delivery therefore influences the types of bacteria acquired by the neonate and the communities established [111].

1.2.10.5 Salivary production Healthy adults produce approximately 500-1500 mL of saliva per day. However, daytime salivary flow rates vary greatly between individuals [92]. The normal flow rate of stimulated saliva is 1-3 mL/min and up to 6 mL/min. The low range is considered to be 0.7-1.0 mL/min, with hyposalivation usually defined as flow rates less than 0.7 mL/min. By contrast, normal unstimulated saliva secretion ranges from 0.25-0.35 mL/min, with low values ranging from 0.1-0.25 mL/min, and hyposalivation defined as less than 0.1 mL/min [91]. Circadian rhythms greatly influence the salivary flow rate, with the salivary flow reaching its highest peak at the end of the afternoon then declining to almost zero during sleep [95].

However, several factors which vary greatly among individuals influence salivary flow. The degree of hydration is the most potent factor. Irritation of salivary glands by tobacco components stimulates salivary secretion, thus smokers have higher salivary flow than non-smokers [112], while intake significantly reduces the salivary flow [113]. Other factors such as various medications, thinking about food, and mechanical stimulation, can influence salivary flow [91, 95]. Saliva is stimulated by smells and , and mild food acids such as citric acid are considered to be the most potent stimulators [114].

The contribution by the various salivary glands to secretion during unstimulated flow is 20% by parotid glands, 65-70% by the submandibular gland, 7-8% by the sublingual glands, and less than 10% by the minor salivary glands (Fig. 1.19). However, during stimulation, the parotids contribute on average more than half of total salivary secretion [91, 95].

29

Figure 1.19 Glandular contribution to unstimulated salivary flow [91].

1.2.10.6 Diagnostic and metabolic roles of saliva Analysis of saliva for investigation of diseases is potentially valuable for children and for older adults, as sample collection is less invasive than blood collection. In addition, analysis of saliva may provide a cost effective approach for screening for some types of diseases [115]. It is used to assess the diagnosis of dental conditions such as caries, periodontal disease, salivary gland disease, and Candida infections, as well as for the diagnosis of viral diseases, tuberculosis, lymphoma, gastric ulcer and cancers, heart disease, and Sjogren’s syndrome [95, 116]. However, the use of saliva to diagnose metabolic status has not been widely studied. In particular, purines and pyrimidines have only been analysed in one short report about adult saliva [117]. Saliva has also been advocated for some forms of therapeutic drug monitoring [118-120].

30

1.2.11 HUMAN MILK PHYSIOLOGY AND COMPOSITION One aim of the present study was to investigate the interaction between neonatal saliva and breast milk. As discussed above, xanthine oxidase (XO) is found in an active form in mammalian milk [67], and it has been suggested that this plays an important role in establishing immunity in the newborn through its ability to generate ROS and RNS which have antibacterial properties [71]. Xanthine is excreted in the mammary glands coincident with milk, and this generates peroxide in the mammary milk sinuses, although this peroxidative power is considered to be depleted by the time the milk is exuded [24, 47, 73].

1.2.11.1 Physiology of human milk secretion There are 2 essential hormones controlling breast milk secretion, prolactin and oxytocin. Prolactin is produced by the anterior pituitary gland and functions as a stimulator for secretory cells in the mammary alveoli to produce milk. This hormone is inhibited during pregnancy by the high level of oestrogen and , which consequently inhibits lactation during pregnancy. Oxytocin is produced by the posterior pituitary and enhances the alveoli to contract and push milk from the secretory cells to the milk sinuses in the mammary alveolar area. Sucking plays an important part in the secretion of milk, as prolactin and oxytocin are both stimulated by this neonatal reflex [121]. The volume and contents of human milk are influenced by various factors such as genetic profile, maternal nutrition, and lactation [122]. The initial form of milk produced at the beginning of lactation is called “colostrum” and has differences in protein content including high concentrations of XO [123].

1.2.11.2 Human milk proteins Human milk contains the lowest total protein concentration of all mammalian milk, with approximately 0.8-0.9 g/dL, perhaps due to the slow growth of the child [122, 124]. Human milk proteins are classified as three types: casein, mucins and whey proteins, containing an adequate amount of amino acids except phenylalanine and tyrosine which are present in low levels [124]. Casein has highly glycosylated subunits [125] which bind to form micelles [126]. Casein also binds calcium and to facilitate their absorption, [124] and forms curds in the stomach [121] which reduce gastric emptying time in infants [124]. Furthermore, the casein κ-subunit shows some antibacterial activity, as casein can inhibit the adherence of Helicobacter pylori to gastric mucosa, and Streptococcus pneumoniae and Haemophilus influenzae to respiratory-tract epithelial cells in infants [125].

Mucins comprise the lowest fraction of human milk proteins and function mainly to surround the fat globules with membranes. Interestingly, the mucin protein fraction remains constant as a result of 31 stable milk fat secretion during the period of lactation [127]. Milk whey proteins comprise all other proteins including enzymes. These have functions in digestion, metabolism, immunity, proteolysis inhibition, and transportation. Examples of whey proteins include lactoferrin, albumin, immunoglobulins, lysozyme, lactoperoxidase, protease inhibitors and XO [126].

Human milk especially colostrum, is rich in immunoglobulins. Secretory IgA represents the highest immunoglobulin levels, while IgG and IgM occur in lower concentrations. These immunoglobulins may protect the mammary gland and infants from infection. Secretory IgA is resistant to proteolysis, a feature that may allow it to exert its function in the gastrointestinal tract of breast-fed infants [126].

1.2.11.3 Human milk lipids Fat is present in breast milk as triglycerides, phospholipids and cholesterol, and represents the main energy source for infants (35%-50% of total needs). Most human milk fats are present as globules [124]. Triglycerides, the main components of milk fats, [121] are digested by lipase into glycerol and free fatty acids, especially long-chain polyunsaturated fatty acids ( and arachidonic acid) which are important for growth and brain development [124].

1.2.11.4 Human milk carbohydrates Most in human milk is in the form of the disaccharide lactose. Lactose is degraded to monosaccharaides and , the former having an essential role in the development of brain white matter in infants [124]. Moreover, lactose facilitates the absorption of calcium and iron and enhances the growth of lactobacilli in the intestinal lumen, which acts to minimise the proliferation of pathogenic bacteria. On the other hand, glucose derived from lactose is an important energy source for the infant brain [121]. Other complex carbohydrates in human milk have antibacterial properties such as glycoproteins, glycolipids, and oligosaccharides [124]

1.2.11.5 White blood cells in human milk Leukocytes (white cells) occur in human milk and comprise macrophages (40%-50%), neutrophils (40%-50%) and lymphocytes (5%-10%). These occur in high numbers at the beginning of lactation in colostrum, but then decrease in the mature milk [124, 126]. Leukocytes in human milk predominantly exert their function by phagocytosis of foreign microorganisms, and this is enhanced when complement proteins C3b first coat the surface of microorganisms [124].

1.2.11.6 Human milk minerals Sodium, potassium, chloride, calcium, magnesium, and phosphorus are the major elements in human milk [124]. Interestingly, there is a weak correlation between calcium, magnesium, and 32 phosphorus in human milk and maternal serum. The infant serum concentrations of calcium and magnesium increase with lactation which is associated with a decline in phosphorus concentration as this element is absorbed with bone growth [122]

The trace minerals , iron, , selenium and manganese are also found in human milk. Breast-fed infants have adequate amounts of zinc compared to formula-fed. Copper and iron decrease rapidly in early lactation and are stable in mature milk, while zinc undergoes a continuous decline overtime even in mature milk [122]. Infants absorb iron from human milk at a five-times higher rate than from bovine milk [128]. The absorption of iron from human milk is enhanced by the presence of and lactose [124].

1.2.11.7 Human milk vitamins The vitamin content of human milk is markedly affected by the maternal nutritional status. All water soluble vitamins are found in breast milk, with vitamin C concentrations in maternal milk being 8 to 10 times higher than in plasma. Furthermore, fat-soluble vitamins are also present in human milk; vitamin D supplementation is essential especially in vegetarian mothers as concentrations of this vitamin critically decrease in maternal milk [124, 129].

1.2.11.8 Human milk nucleotides Human milk has high amounts of nucleotides compared to bovine milk. Human milk is reported to contain the mononucleotides CMP, UMP, AMP, GMP and IMP, and the dinucleotides CDP, UDP, ADP and GDP, with the pyrimidine nucleotides (of cytidine and uridine) representing the highest fractions [130, 131]. Interestingly, no trinucleotides have been detected in these studies. It has been suggested that human nucleotides may have roles in intestinal development, iron absorption, regulation of intestinal flora, , and immune function [132, 133]. More recently, some studies have shown a reduction in the incidence of diarrhoea in infants who are fed nucleotide-supplemented infant formula compared to those fed non-supplemented formula [134]. There appears to be no published analyses of the nucleoside and base composition of human milk, although bovine milk has been shown to have high levels of the pyrimidine base orotic acid [135] which appears to be absent in human milk.

1.2.11.9 Mastitis Mastitis is an inflammation of breast tissues brought about by infection with several species of microorganisms. The usual symptoms of mastitis are breast pain, redness, tenderness, and fever. The inflammation can develop into abscesses and septicaemia [136]. Many factors have been reported to predispose to mastitis such as stress, malnutrition [137], trauma, immunity, hygiene and genetic factors [138]. 33

Mastitis can occur at any stage of lactation but the infection rate is highest in the first few weeks postpartum at an incidence of 2%-33% [136, 139]. Some mastitis infections are asymptomatic (subclinical mastitis) but may be identified by low milk production and increased sodium concentration and bacterial counts [138].

The most common bacteria causing mastitis is S. aureus (especially hospital-acquired mastitis) and other Staphylococcal species as well as E. coli. Mastitis is caused rarely by some Mycobacterium species, and the yeast, Candida and Cryptococcus species [137-140]. As discussed above, an important role attributed to XO in milk, has been protection against mastitis by the production of

H2O2 within the mammary alveolar spaces and storage ducts.

34

1.3 OVERALL PROJECT HYPOTHESIS

It is hypothesised that the substrates for xanthine oxidase (XO), xanthine and hypoxanthine, as well as other nucleotide metabolites are present in higher concentrations in neonatal saliva, have important roles in generating H2O2 and in oral biology by regulating the microbiota of the mouth and the gut.

1.4 OVERALL PROJECT AIM

To study and establish a new research direction for the role of nucleotide metabolites in saliva, and to compare patterns of concentrations of these metabolites in mammals specially humans. In particular, neonatal salivary nucleotide metabolite concentrations and their possible interaction with breast milk will be studied. These metabolites are poorly described and their functions and level of excretion by salivary glands are still unknown.

1.5 ETHICS

Permission from the Human Research Ethics Committee (HREC) of two institutions was granted for the project:

1. Mater Health Services HREC (see Appendix 1, 3, 5). 2. The University of Queensland HREC (see Appendix 2, 4, 6).

1.6 FUNDING FOR THE PROJECT

1. Golden Casket Research Fund (see Appendix 7). 2. NHMRC (see Appendix 8).

35

Chapter 2

Determination of nucleotide metabolites in human and mammalian saliva by high performance liquid chromatography with tandem mass spectrometry: changes during early human infancy and a survey of ranges in adult humans and other mammals

The methodology component and relevant results of this chapter have been published in Journal of Chromatography B (see Appendix 9)

36

2.1 INTRODUCTION

2.1.1 ROLES OF PURINES AND PYRIMIDINES Purine and pyrimidine nucleotides are critical molecules in cells, being essential in many biochemical functions including biosynthesis, energy supply, units of DNA and RNA, essential coenzymes, and regulatory mechanisms. They are synthesised in mammalian cells by either de novo pathways that are energy consuming, or by salvage pathways that conserve energy [28, 141]. Purine nucleotides are usually salvaged from their nucleobases, whereas pyrimidines are recovered from their nucleosides. Non-salvaged purines are catabolised to uric acid by the enzyme xanthine oxidase (XO), while pyrimidines are catabolised to amino acids, ammonia and carbon dioxide [12, 19].

Dietary and endogenous nucleosides and bases are rapidly absorbed for salvage or catabolism, and their plasma concentrations are very low. With the exception of uridine and uric acid, nucleoside and base concentrations are less than 1 µM in plasma [4, 142-145]. Nucleotides are usually located intracellularly and are therefore regarded as indicators of cellular damage when they appear in the extracellular fluid. Nucleosides and bases (which are metabolites of nucleotides) occur extracellularly, and raised concentrations in plasma and urine are used as markers of nucleotide metabolic disorders [142, 146].

2.1.2 SALIVARY PURINES AND PYRIMIDINES Research interest in saliva has been heightened with the discovery that saliva contains a number of biochemical components, which may be useful for the monitoring of drugs and for early diagnosis of diseases such as cancer and heart disease [97, 116, 147, 148]. Moreover, for young children and the elderly the collection of saliva is much less invasive than blood sampling [115]. However, the use of saliva to evaluate metabolic status has not been widely studied; in particular, purine and pyrimidine metabolites in saliva have been examined only by one group, with this report not being followed up elsewhere [117].

In this chapter a method is described that was developed specifically for the collection of saliva in neonates and adults. This was validated and then applied for the determination of purine and pyrimidine metabolites in small volumes of saliva using high performance liquid chromatography (HPLC), initially with detection by photodiode-array (PDA) and then tandem mass spectrometry (MS/MS). Pilot data concerning the salivary concentrations of nucleotide metabolites were collected from a small sample of adults and neonates. It was discovered that the concentrations of 37 these metabolites were significantly raised in neonates compared with adults. Salivary analyses of 77 adults, and 60 neonates were conducted. A longitudinal study of infants over 12 months was then undertaken. In addition, a selection of domestic animals was chosen for analysis and the patterns of salivary purines and pyrimidines were found to vary greatly. The proposed function of these metabolites in saliva, especially from neonates, was subsequently discussed.

Initially, a pilot study on infants of ages ranging from newborn to 2 weeks to 3 years led to the finding that neonatal saliva had higher concentrations of nucleosides and bases compared to adults, and that the transition between infant and adult saliva patterns of purines and pyrimidines apparently occurs during the first year of life. These observations led to the formulation of the hypotheses listed below.

38

2.2 SPECIFIC HYPOTHESES AND RATIONALE

Hypothesis 1: That purine and pyrimidine bases and nucleosides are present in adult and neonatal saliva and that these nucleotide metabolites are present in higher concentrations in neonatal saliva compared to adults.

Hypothesis 2: That the presence of raised concentrations of purine and pyrimidine bases and their nucleosides in neonatal saliva return to adult levels in the first year of life.

Hypothesis 3: That purine and pyrimidine bases and their nucleosides are present in saliva of some other mammals.

Rationale: Low concentrations of some purine bases, nucleosides and nucleotides have been reported in adult saliva; but these have not been studied in infants or children. If they are different then they may relate to early regulation of the micro-environment of the infant mouth, and must change to the adult pattern at some stage. For ethical and organisational reasons at the Mater Hospital where this study was to be undertaken, this initial investigation was most easily performed over the first year of life. Similarly, the pattern of nucleotide metabolites in mammals other than humans has not been reported previously. This will provide species comparative data on salivary purines and pyrimidines. It may also provide clues to the roles of salivary nucleotide metabolites by relating them to known oral microbial compositions and dietary habits (i.e. grass-eating, omnivorous, carnivorous). Any such study will be dependent upon initial careful standardisation of methods for saliva collection and the development of sufficiently sensitive analytical methods.

2.3 SPECIFIC AIMS

1. To develop and validate a method for sampling saliva from adults and infants and extraction of purines and pyrimidines metabolites, including a method for collection of small volumes of saliva from neonates 2. To develop and validate an analytical method for the determination of purine and pyrimidine metabolites in small volumes of saliva using high performance liquid chromatography (HPLC) with tandem mass spectrometry (LC-MS/MS) 3. To conduct a survey of concentration ranges of nucleotide metabolites in adult saliva 4. To conduct a survey of concentration ranges of metabolites in neonatal saliva

39

5. To conduct a 12-month longitudinal study of the nucleobases and nucleosides metabolome of saliva of newborn babies through to weaning and early infancy 6. To conduct a survey of saliva from other mammalian species to investigate and compare the salivary nucleotide metabolite patterns between mammals and humans

40

2.4 MATERIALS AND METHODS

2.4.1 SUBJECTS AND SAMPLE COLLECTION Study protocols received prior written approval from Mater Health Services and The University of Queensland Human Research Ethics Committees (see Appendices 1-4).

2.4.1.1 Adults for saliva survey Unstimulated whole saliva samples were collected from 77 adult volunteers, 30 males and 47 females, aged 19-70 years, at the Mater Adult Hospital, Brisbane, Australia. The inclusion criteria for the study were healthy adults (>18 years old), males and females. Adults of ethnicity other than Caucasian and adults with metabolic disorders or oral pathology were excluded from the study. The reason for excluding non-Caucasian was to minimising genetic variance. Sampling was performed between 0800 and 0900, with subjects having no food or drink for at least 2 h before sampling. The participants were asked to rinse their mouth with water and wait for approximately 10 min before salivating ('passive drool') into sterile 50 mL polypropylene universal containers (Sarstedt Ply Ltd, Mawson Lakes, SA, Australia). Samples were immediately placed on ice, transferred to the laboratory, divided into small portions and stored at -80oC until analysed.

2.4.1.2 Neonates for saliva survey Neonatal saliva was collected using cotton oral swabs (Swisspers Baby Tips, McPherson’s Consumer Products, Kingsgrove, NSW, Australia) with plastic stick and double-ended cotton buds. Because of chromatographic interference, it was necessary to remove UV-absorbing compounds which were present in the cotton swabs. Therefore, before saliva collection all cotton swabs were washed sequentially with deionised water, ethanol, diethyl ether, then left to dry for at least 24 h at room temperature.

One cotton tip of each oral swab was removed to avoid confusion between “used” and “unused” ends. Two swabs were then placed in a 5 mL screw-cap plastic test tube and weighed to the nearest milligram to obtain the (pre-collection) dry weight. Unstimulated whole saliva samples were collected by a research nurse from 60 healthy neonates who had not been fed for at least 30 min, aged 1-4 days, at the Mater Mothers’ Hospital, South Brisbane, QLD, Australia. The inclusion criteria for the study were healthy full-term vaginally-delivered Caucasian neonates, males and females. Neonates of ethnicity other than Caucasian and those with complications, pathologies or infections and metabolic disorders that may affect the baby’s saliva production or quality were excluded. Two cotton swabs were placed at the same time under the tongue and moved gently around the buccal cavity for approximately 5 min; no infant experienced any obvious discomfort. 41

The swabs were immediately put back into the capped test tube to prevent evaporation then placed on ice. The test tube was re-weighed on the same balance used for the pre-collection weighing to calculate the volume of saliva on the swabs assuming a salivary specific gravity of 1.0 (i.e. assuming saliva density is similar to the density of water [149]. The samples were then processed as described below, or stored at -80°C at which temperature standard solutions of nucleosides and bases were found to be stable for at least 12 months. The subjects’ characteristics are shown in Table 2.1 below.

Table ‎2.1 Physical information about the participated neonates.

Male Female p value Total

N 22 38 n.s. 60

Gestational age (weeks) Mean±SD 40±1.4 40±1.3 0.5 (n.s.) 40±1.3 Range 38-42 37-42 37-42

Birth weight (g) Mean±SD 3751±467 3665±435 0.5 (n.s.) 3697±445 Range 2630-4685 2768-4446 2630-4685

Age (hours) Mean±SD 34±14 34±16 0.9 (n.s.) 34±14 Range 8-62 8-80 8-80 n.s.= not significant at p=0.05

2.4.1.3 Subjects for infant longitudinal study Of the 60 neonates studied, the parents of 20 of them agreed for their infants to participate in a longitudinal study involving follow-up sampling at 6 weeks (n= 20), 6 months (n= 19), and 12 months (n= 14) after the initial collection. Unstimulated samples were collected as described above using polystyrene plastic pipettes (Sigma-Aldrich Pty Ltd, Castle Hill, NSW, Australia) from infants who had not been fed for at least 30 min. Samples were immediately placed on ice, then transferred without delay to the laboratory for storage at -80oC.

2.4.1.4 Survey of other mammalian species Domesticated animals were chosen for saliva sampling. Selection was based essentially on a dietary rationale (which may be reflected by the oral microbiota) and the digestive tract characteristics: cat: obligatory carnivore; dog: omnivore; horse, cow, sheep, goat and camel: herbivores; cows, sheep and goats are ruminants; horses have pre-ruminant sacculation; camels are pseudo-ruminants. Saliva was collected from 8 cattle (5 of the 8 samples using cotton swabs, 3 collected from oral drool), 5

42 sheep (cotton swabs), 7 dogs (cotton swabs), 5 cats (cotton swabs), 5 horses (samples collected using polystyrene plastic pipettes), 4 goats (cotton swabs), and 1 camel (cotton swab). Sampling the herbivorous mammals was initially problematic because of contamination by vegetation, presumably by chlorophyll, which resulted in a strong green appearance of the saliva. For horses, this was overcome by placing them in a sand yard for one hour prior to sampling. Cows were sampled as they waited in bales before milking, during which time they do not tend to regurgitate their cud. Sheep, goats and the camel were penned before sampling, which discourages cud- chewing. Dogs and cats were sampled before feeding.

2.4.2 NEONATAL SALIVA EXTRACTION AND VALIDATION

2.4.2.1 Investigation for presence of interfering compounds Two types of swabs, cotton and Sorbette® sponge swabs (Stratech Scientific, Sydney, NSW, Australia) were initially investigated for the presence of interfering compounds. One hundred microliters of deionised water was added to each swab which was kept at 4°C for 30 min in a 5 mL capped plastic test tube, then sonicated in a bath (Unisonics Pty. Ltd, Brookvale, NSW, Australia) for 3 min. The tip of each swab was removed and transferred to a Salivette® (Sarstedt Pty Ltd, Mawson Lakes, SA, Australia) and the water was collected by centrifugation at 4500 g, for 5 min, then analysed by HPLC with photodiode array (PDA) UV detection at 268 nm which is the mean peak wavelength for most purines and pyrimidines. Two recovery methods were evaluated for the extraction of nucleotide metabolites from the cotton swabs as detailed below.

2.4.2.2 Validation of recovery of extraction Method #1 (Aqueous extraction): The aqueous extraction of nucleotide metabolites from cotton swabs was evaluated by spiking washed cotton swabs with 25, 50, 75, and 100 µL of a mixture of uracil, hypoxanthine, xanthine, adenosine, inosine, guanosine, and uridine (59, 55, 75, 47, 47, 41, 66 µM respectively) prepared in deionised water in 5 replicates. Each cotton swab was then maintained at 4°C for 30 min in a 5 mL capped Pyrex test tube (Sarstedt Pty Ltd, Mawson Lakes, SA, Australia), then transferred into another 5 mL test tube containing 100 µL of deionised water and sonicated in an ultrasonic bath for 3 min. The cotton tip was removed and transferred to a Salivette tube and the aqueous component (containing the analytes) was obtained via centrifugation at 4500 g, for 5 min at 20oC. Extracts were then stored at -80°C pending HPLC-MS/MS analysis.

Method #2 (Organic solvent extraction): Organic solvent extraction of nucleotide metabolites from cotton swabs was also evaluated as a potential means of improving recovery. Each cotton

43 swab was spiked in 5 replicates with 25, 50, 75, 100 µL of the standard mixture as detailed above. The mixture was kept at 4°C for 30 min in a capped test tube, then placed into Pyrex 5 mL test tubes, 400 µL of deionised water were added, then the vial was sonicated in the ultrasonic bath for 3 min. The aqueous extract was transferred to a clean Pyrex glass evaporating tube, and the extraction of the swab was repeated twice with 600 µL of acetonitrile, which were then combined with the aqueous extract in the evaporating tube. The cotton tip was removed and the swab was transferred to a Salivette tube, which was centrifuged at 4500 g for 5 min at 20oC to recover the remaining acetonitrile in each tip, which was also added to the combined extracts. The total extract was then evaporated to dryness in a heating block (Pierce Reacti-Therm III Model 18823, Thermo Fisher Scientific Australia Pty Ltd, Scoresby, VIC, Australia) at 40oC under a gentle stream of nitrogen. The residue was dissolved with vortex mixing in 100 µL of deionised water then stored at -80°C pending LC-MS/MS analysis.

2.4.2.3 Validation of linearity of extraction The linearity of the extraction method was evaluated using pooled standard solutions prepared in deionised water. Cotton swabs in 5 replicates were spiked with a range of concentrations of bases and nucleosides (5, 10, 20, 40, 80 µM) using the standard mixture described above. The bases and nucleosides were extracted as described for extraction method #2, then subjected to LC-MS/MS analysis. The data were subjected to linear regression analysis (unweighed).

2.4.3 ANALYSIS OF PURINES AND PYRIMIDINES

2.4.3.1 Sample preparation Prior to HPLC analysis, adult and neonatal saliva samples as well as extracted standards were ultra- filtered by microfuging in 0.5 mL, 50 kDa Amicon ultrafilters (Merck Pty. Ltd., Kilsyth, VIC Australia) at 9500 g, for 10 min at 20°C to remove proteins, bacteria and fibres. For HPLC-MS/MS analysis each salivary filtrate and the extracted standards were mixed with an isotope-labelled internal standard solution as described below (2.4.3.3).

2.4.3.2 Initial analysis using HPLC Initially, the presence of purines and pyrimidines in saliva was investigated using an Agilent 1100 HPLC system (Agilent Technologies Pty Ltd, Mulgrave, VIC, Australia) with auto-injector, column heater and PDA detection, scanning absorbance over a range from 230-350 nm. The initial analyses were conducted on a C18 column (Phenomenex Synergy Hydro RP-80A, 4 µm, 250 x 3 mm, Lane Cove West, NSW, Australia). Chromatography was performed at 25°C at a flow rate of 0.4 mL/min using a linear gradient elution profile of 100% v/v aqueous 40 mM ammonium acetate pH 5.0 44

(mobile phase A) (Sigma-Aldrich Pty. Ltd, Castle Hill, NSW, Australia) to 15% v/v acetonitrile (mobile phase B) (Sigma-Aldrich Pty. Ltd , Castle Hill, NSW, Australia) over 25 min; the composition was returned to 100% v/v A for 10 min equilibration before the next injection, as described previously [146]. Nucleosides and bases were identified by their retention times and UV spectra scanning from 230 nm to 350 nm. Subsequently, all HPLC analyses were conducted using HPLC with tandem mass spectrometry (LC-MS/MS), as detailed below.

2.4.3.3 LC-MS/MS conditions The LC-MS/MS method for measuring nucleotide metabolites in biological fluids was developed at the Pathology Department, and was a modification of that reported previously for purine and pyrimidine determination in urine [150], using a Shimadzu Prominence UFLC system (Shimadzu, Japan). Briefly, 10 µL of processed saliva sample was mixed with 35 µL of 0.1% (w/w) formic acid and 5 µL of a solution of deuterated standards comprising 2,3,7,8-[D4]-hypoxanthine, 5,6-[D2]- uridine-, 5,6-[D2]-2-deoxyuridine, 6-[D3]-methyl-thymine, 1,3-[15N2]-uric acid, [D3]-methyl- creatinine, and 1,3-[15N2]-orotic acid (Cambridge Isotope Laboratories Inc. Andover, MA, USA), (CDN, Canada), and (Sigma-Aldrich Pty Ltd, Castle Hill, NSW, Australia). Low and high concentration quality control standards were prepared similarly to the biological samples.

Samples were injected onto a 4 µm C18 column (Phenomenex Synergy Hydro RP-80A, 4 µm, 150 x 3.0 mm, Lane Cove West, NSW, Australia). Chromatography was performed at 30°C at a total flow rate of 0.3 mL/min, using a linear gradient elution from 98% v/v mobile phase A (aqueous 5 mM ammonium acetate, 0.05% w/v formic acid) to 15% v/v mobile phase B (0.05% w/v formic acid in methanol) over 11 min, then 60% v/v mobile phase B for 1 min, before returning to 2% v/v B for a further 6 min for equilibration. The gradient profile is shown in Figure 2.1. This method varied slightly from the HPLC-PDA method above, in that a shorter column and a slightly reduced flow rate were used. The addition of formic acid to the ammonium acetate buffer and methanol provided earlier elution of peaks and faster run time. There was some loss of peak resolution compared with the HPLC-PDA method, but this was compensated by the higher specificity and positive identification provided by tandem mass spectrometry.

Tandem mass spectrometric (MS/MS) detection was performed using an Applied Biosystems API 4000 QTRAP mass spectrometer (Applied Biosystems, Forster City, CA, USA), equipped with a Turbo V-Spray source with the gas temperature set at 500°C. The source operated an electrospray interface (ESI) with switching ionisation polarity (between +5000V and -4000V) during the run (18

45 min). The eluent was monitored by specific ion transitions (MRMs) for each compound and an internal standard. All data were quantified using Applied Biosystems Analyst version 1.5 software.

46

100

80

60 % B 40

20

0 0.0 5.0 10.0 15.0 18.2 Time (min)

Figure 2‎ .1 Gradient profile of mobile phase B that was used for reversed phase HPLC elution of purine and pyrimidine nucleosides and bases followed by MS/MS analysis. Chromatography was performed at a total flow rate of 0.3 mL/min, using a linear gradient elution from 98% v/v mobile phase A to 15% v/v mobile phase B over 11 min, then 60% v/v mobile phase B for 1 min, before returning to 2% v/v B for a further 6 min for equilibration.

2.4.3.4 Within run imprecision Within-run precision was assessed by spiking 2 adult salivary samples with low (30, 25, 32, 9, 14, 5, 14 µM) and high (127, 136, 131, 71, 90, 57, 115 µM) concentrations of uracil, hypoxanthine, xanthine, adenosine, inosine, guanosine, and uridine respectively. Samples were then ultra-filtered and run on LC-MS/MS; 8 replicates were run for each QC.

2.4.3.5 Limit of detection Limit of quantitation (LOQ) and limit of detection (LOD) were determined by injection of 10 µL of aqueous standards with varying low metabolite concentrations. The LOD was defined as the lowest concentration that gave a signal-to-noise (SN) ratio of 3, while the LOQ was defined by a SN ratio of 10.

2.4.4 STATISTICAL ANALYSIS Because the raw data were not normally distributed according to D’Agostino and Pearson omnibus normality test, non-parametric statistical approaches were used for the data analyses. Median values were compared between 2 different groups using the Mann-Whitney test and between more than 2 groups using the Kruskal-Wallis test. A p-value of 0.05 was the cut-off for statistical significance. All analyses were performed using GraphPad Prism 5.

47

2.5 RESULTS

2.5.1 NEONATAL SALIVA EXTRACTION AND VALIDATION Aim1: To develop and validate a method for sampling saliva from infants and extraction of purines and pyrimidines metabolites, including a method for collection of small volumes of saliva from neonates

2.5.1.1 Investigation of the presence of interfering compounds In this study, the neonatal saliva collection methodology was validated using 2 different types of swabs: Sorbette® sponges and cotton swabs. Before sampling, the presence of interfering compounds was first checked by spiking the swabs with 100 µL of distilled water, followed by extraction and centrifugation and subsequent HPLC-PDA using Method #1 (see section 2.4.2.2) as outlined for purine and pyrimidine extraction. Interestingly, the cotton swabs produced large UV- absorbing peaks in the chromatogram (Fig. 2.2 A). Sorbette sponge swabs have been used previously to collect saliva from infants for cortisol determination [151], but in this study the Sorbette sponge also produced UV-absorbing peaks (Fig. 2.2 B). However, these swabs could not be washed with solvents prior to saliva collection without destroying the sponge. The commercial cotton swabs were more robust and were not affected by the washing method.

Removal of UV-absorbing peaks was best achieved by washing the cotton swabs with excess volumes of water, ethanol, then diethyl ether followed by air-drying. This removed all but one of the UV-absorbing compounds (Fig. 2.2 C); this method was subsequently used for all neonatal and animal saliva collections. The remaining UV-absorbing compound was not one of the original peaks from the unwashed swabs but may have arisen from the diethyl-ether treatment, but it did not co- elute with any of the peaks of interest.

The mean±SD volume of neonatal saliva collected on cotton swabs was (135± 79 µL); these small volumes were sufficient since only 10 µL of ultrafiltered sample was needed for HPLC-MS/MS analysis. Any remaining sample was used for further studies (see Chapter 3 for more details).

48

A

B

C

Figure 2.2 Chromatograms show UV-absorbing peaks at 268 nm present in 100 µL distilled water extracted from: (a) a commercial cotton tip, (b) a Sorbette sponge, and (c) a commercial cotton tip following washing with water, ethanol and diethyl ether. The water was then extracted from the tips by centrifugation and run on HPLC-PDA. A wavelength of 268 nm was used for monitoring as this is an average value for purine and pyrimidine peak absorption.

49

2.5.1.2 Validation of recovery of extraction Experiments to determine the recovery of nucleotide metabolites from cotton swabs were conducted using 5 replicates for each analysis of the aqueous standard mixture of purine and pyrimidine metabolites. Extraction of the analytes from cotton tips using water only (Method #1) showed marked variability in the recoveries at low volumes (25 µL) of the standard mix (Table 2.2). For example, hypoxanthine, xanthine, adenosine, and guanosine recoveries were 66.3% ± 7.8 (mean ± SD), 56.9% ± 8.5%, 71.2% ± 10.3, and 75.0% ± 6.9 respectively. This was improved by increasing the standard mix volume to at least 100 µL (a volume greater than some neonatal saliva samples) (Fig. 2.3). Nonetheless, low overall recoveries persisted for some compounds; e.g. recoveries of hypoxanthine (77.8% ± 8.2), xanthine (68.6% ± 8.8) and adenosine (81.0% ± 7.3) were low compared to those for uracil (90.6% ± 7.0), inosine (97.3% ± 9.7), guanosine (92.0%±12.0), and uridine (109.2% ± 6.6). This may be explained by the use of water only for the extraction procedure, as xanthine and hypoxanthine are relatively non-polar compared to the other bases and nucleosides and hence are more likely to be adsorbed onto cotton fibres.

In comparison, the recoveries using the water/acetonitrile extraction (Method #2, section 2.4.2.2) for all purine and pyrimidine metabolites were consistently high (Table 2.3), even for low standard mix volumes, with uracil, adenosine, inosine, and guanosine having the highest overall mean recoveries which exceeded 95%. Acceptable mean recoveries (>90%) also were obtained for hypoxanthine, xanthine, and uridine. Method #2 thus showed consistently satisfactory recoveries of the analytes from cotton swabs, even with low volumes of the standard mixture (Fig. 2.4). This was important because neonatal saliva was frequently recovered in small sample volumes.

50

Table 2‎ .2 Percent recovery of nucleosides and bases from cotton swabs using centrifugation (Method #1). The concentration of the analyte standards are shown in parentheses. The results are shown as the mean±SD (n=5)

Uracil Hypoxanthine Xanthine Adenosine Inosine Guanosine Uridine Volume (µL) (59 µM) (55 µM) (75 µM) (47 µM) (47 µM) (41 µM) (66 µM) 25 83.6±15.1 66.3±7.8 56.9±8.5 71.2±10 85.2±10 75.0±6.9 100±4.9

50 89.2±16 80.4±7.1 67.4±7.1 81.9±5.6 95.2±5.5 92.6±7.8 109±7.6

75 89.4±6.0 78.8±2.2 73.2±5.4 82.1±5.5 100±3.4 98.1±5.1 112±2.9

100 100±3.9 85.5±5.4 77.0±3.8 88.9±4.6 108±4.5 102±2.7 115±4.1

Mean±SD 90.6±7.0 77.8±8.2 68.6±8.8 81.0±7.3 97.3±9.7 92.0±12 109±6.6

Table 2‎ .3 Percent recovery of nucleosides and bases from cotton swabs using the solvent and evaporation (Method #2). The concentration of the analyte standards are shown in parentheses. The results are shown as the mean±SD (n=5)

Uracil Hypoxanthine Xanthine Adenosine Inosine Guanosine Uridine Volume (µL) (59 µM) (55 µM) (75 µM) (47 µM) (47 µM) (41 µM) (66 µM) 25 101±5.5 90.9±5.0 92.7±3.7 103±6.5 109±4.7 99.3±3.9 90.4±2.4

50 98.6±3.0 88.0±3.9 88.9±4.8 101±4.3 113±2.7 99.3±3.7 91.4±1.9

75 108±3.5 98.4±4.0 97.6±3.4 112±6.3 111±5.5 106±4.8 95.9±1.2

100 108±2.4 93.9±2.8 91.7±3.3 112±2.4 109±3.8 107±4.3 98.2±3.8

Mean±SD 104±4.2 92.8±3.8 92.7±3.2 106.9±4.8 111±1.5 103±3.7 94.0±3.2

51

120

100

80

60

40 Hypoxanthine

(%) Recovery Xanthine 20 0

0 50 100 150 200 250 Volume (L)

120

100

80 60 Adenosine 40 Inosine (%) Recovery Guanosine 20

0 0 50 100 150 200 250 Volume (L)

Figure 2.3 The “best-fit” curve for the percent recovery of major nucleotide metabolites using extraction with water followed by centrifugation (Method #1). The graphs show the gradual increase of recovery with increasing volumes of standard mix loaded onto the cotton tips. The error bars represent mean±SD (n=5).

52

120 Uracil 100 Hypoxanthine Xanthine 80 Adenosine 60 Inosine Guanosine 40 Recovery (%) Recovery Uridine 20

0 0 20 40 60 80 100 120 Volume (L)

Figure 2‎ .4 Percent recovery of major nucleotide metabolites using extraction with solvent and evaporation followed by centrifugation (Method #2). The graph shows consistent recovery with all volumes of standard mix loaded onto the cotton tips. The error bars represent mean±SD (n=5).

53

2.5.1.3 Validation of linearity of extraction Calibration plots for the standard bases and nucleosides recovered from cotton swabs by the solvent and evaporation method demonstrated excellent linear relationships between area ratio and concentration (Fig. 2.5) for each compound. Results obtained from the linear regression data analysis are shown in Table 2.4.

1.5

Hypoxanthine

1.0 Inosine Uridine Uracil

Area ratio Area 0.5

0.0 0 20 40 60 80 100 Standard concentration (M)

1.5

Adenosine 1.0 Guanosine Xanthine

ratio Area 0.5

0.0 0 20 40 60 80 100 Standard concentration (M)

Figure 2‎ .5 The results of linear regression analysis of the calibration data in the measurement of nucleosides and bases extracted from cotton swabs using the solvent and evaporation method (Method #2). The error bars represent mean±SD (n=5).

54

Table 2‎ .4 The results of linear regression analysis of the calibration data (Fig. 2.5) in the measurement of nucleosides and bases extracted from cotton swabs using the solvent and evaporation method (Method #2). The experiment was conducted in 5 replicates.

2 Compound Regression equation x-Intercept2 r Uracil y1=0.001x+0.001 -0.876 0.998

Hypoxanthine y=0.018x+0.006 -0.351 0.999

Xanthine y=0.002x+0.004 -1.812 0.998

Adenosine y=0.016x+0.015 -0.956 0.999

Inosine y=0.016x+0.005 -0.283 0.999

Guanosine y=0.011x-0.020 2.052 0.998

Uridine y=0.009x+0.004 -0.466 0.998

1y: Peak area ratio 2x: standard concentrations

55

2.5.2 HPLC-MS/MS DETERMINATION OF NUCLEOTIDE METABOLITES Aim 2: To develop and validate an analytical method for the determination of purine and pyrimidine metabolites in small volumes of saliva using high performance liquid chromatography (HPLC) with tandem mass spectrometry (LC-MS/MS)

2.5.2.1 Within run imprecision Adult saliva was spiked with a “high-level” and a “low-level” purine and pyrimidine mix, then analysed for precision determination. Within-run precision data were calculated from saliva QC samples on a single occasion (8 replicates at each QC level). The data are shown in Table 2.5. Precision was expressed as coefficients of variation (CV %), which was less than 10 %.

Table 2‎ .5 Within-run precision data. Eight replicates of 2 levels (low and high) of spiked saliva with bases and nucleosides were measured for each compound within one run. The high concentrations were first prepared in saliva, then diluted 2.25 fold before the injection. Low Level (µM) High Level (µM) Compound Mean SD %CV Mean SD %CV Uracil 31.2 2.2 6.9 131 12 8.9

Hypoxanthine 24.3 0.6 2.6 135 6.0 4.5

Xanthine 30.6 2.0 6.4 128 7.2 5.6

Adenosine 8.40 0.4 4.4 68.8 4.8 7.0

Inosine 12.9 0.8 6.1 88.1 4.0 4.5

Guanosine 5.40 0.3 5.8 55.9 2.2 3.9

Uridine 13.0 0.8 5.9 113 3.8 3.4

2.5.2.2 Limit of detection and condition parameters The detection limits for each metabolite measured in this study and its corresponding internal standard are presented in Table 2.6. The limit of detection was based on a 10 µL injection and showed good sensitivity with a limit of detection (LOD) and limit of quantitation (LOQ) ranging from 0.01 to 1.5 µM and 0.02 to 5 µM, respectively. An example of a chromatogram of a saliva sample is shown in Figure 2.6. The HPLC-MS/MS conditions and parameters for optimal separation of nucleosides and bases in saliva are shown in Table 2.7.

56

Table ‎2.6 Limit of Quantitation (LOQ) and Limit of Detection (LOD)

Compound LOQ (µM) LOD (µM)

Creatinine 5.0 1.5

Pseudouridine 0.20 0.06

Uracil 1.50 0.50

Hypoxanthine 0.20 0.07

Xanthine 0.70 0.24

Adenine 0.10 0.04

Adenosine 0.03 0.01

Inosine 0.06 0.02

Guanosine 0.03 0.01

Dihydrothymine 1.5 0.50

Dihydrouracil 3.00 1.0

Uridine 0.02 0.01

Thymine 1.5 0.50

Orotic acid 0.50 0.15

Uric acid 0.30 0.10

Deoxyadenosine 0.02 0.01

Deoxyuridine 0.50 0.15

Deoxyinosine 0.40 0.10

Deoxyguanosine 0.05 0.01

Thymidine 0.32 0.10

57

Figure ‎2.6 An example of LC-MS/MS chromatograms for purine and pyrimidine metabolites and creatinine in neonatal saliva. The abundant peak eluting behind the inosine may result from the cross-talk from the more abundant adenosine as they have similar MS transitions. The mass spectrometry data of the identified bases and nucleosides are shown in Table 2.7. Shading indicates peak areas.

58

Table 2‎ .7 HPLC-MS/MS conditions and parameters for optimal separation of nucleosides and bases in saliva. Q1: molecular ion; Q3: daughter ion; Da, Dalton; DP: declustering potential (volts); CE: collision energy (volts).

Nucleosides/Bases Retention time Mode Molecular Transition(Da) Time (ms) DP (V) CE (V) (min) weight Q1/Q3 Creatinine 2.64 Positive 113.12 114.1/44.1 30 60 33

Pseudouridine 4.53 Positive 244.20 245.0/209.2 30 46 15 Uracil 4.70 Positive 112.10 113.1/70.1 30 50 25

Hypoxanthine 7.36 Positive 136.11 137.0/119.0 30 86 29

Xanthine 8.37 Positive 152.11 153.1/110.1 30 70 30

Adenine 6.59 Positive 135.13 136.1/119.1 30 71 33

Adenosine 12.7 Positive 267.20 268.1/136.1 30 50 15

Inosine 11.3 Positive 268.23 269.1/137.1 30 36 23

Guanosine 11.7 Positive 283.24 284.1/152.1 30 41 23

Dihydrothymine 8.92 Positive 128.13 129.1/112.1 30 66 13

Dihydrouracil 4.00 Positive 114.10 115.1/73.1 30 60 17

Succinyladenosine 14.3 Positive 383.31 384.1/252.1 30 66 29

Uridine 8.05 Positive 244.20 247.1/115.1 30 36 15

Thymine 9.70 Negative 126.11 125/42 30 -55 -26

Orotic acid 3.42 Negative 156.10 154.9/111.1 30 -40 -14

Uric acid 6.59 Negative 168.11 167.1/124.1 30 -70 -27

59

2.5.3 SURVEY OF ADULT SALIVARY NUCLEOTIDE METABOLITES Aim 3: To conduct a survey of concentration ranges of nucleotide metabolites in adult saliva The distributions of these metabolites in both adult and neonatal salivary samples are shown in Figure 2.7, while the significant differences between both groups are shown in Table 2.8 (see Appendix 16 for full data set). The HPLC-MS/MS method also identified creatinine in saliva, albeit at very low concentrations compared to plasma levels.

In adult saliva, the concentrations of nucleosides and bases were less than the LOQ, except for hypoxanthine (median=2.1 µM) and xanthine (median=1.7 µM). Uric acid was lower than typical adult plasma values (median=173 µM), but much higher than for the other metabolites. Adult saliva also contained low concentrations of creatinine (median=6.6 µM), which is far below normal plasma ranges for males (60-110 µM), and females (45-90 µM). Interestingly, uracil, hypoxanthine, xanthine and inosine concentrations exceeded 10 µM in a few adults; these values are the outliers presented in Figure 2.7.

2.5.4 SURVEY OF NEONATAL SALIVARY NUCLEOTIDE METABOLITES Aim 4: To conduct a survey of concentration ranges of metabolites in neonatal saliva The chromatograms in Figure 2.6 correspond to the major purine and pyrimidine nucleosides and bases identified in neonatal saliva using this HPLC-MS/MS method. The median concentrations of most purine and pyrimidine metabolites in neonatal saliva were much higher than for adults (Fig. 2.7). Some neonates showed low concentrations but in contrast to adults they were never below the LOQ except for thymine and orotate (Table 2.8) (see Appendix 17 for full data set). Adenine (p=0.15) and uric acid (p=0.07) were respectively similar in each of the two groups. Median thymine concentrations in adult and neonatal saliva were both below the LOQ, which was supported by commensurate low levels of its metabolite, dihydrothymine.

Creatinine concentrations in neonatal saliva (median=22 µM) were low compared to the typical ranges in plasma, but significantly higher than in adult saliva (median=6.6 µM, p=0.001), which could be a result of high plasma creatinine in newborns especially in the first 5 days of life [152] and where the smaller the birth weight, the higher the plasma creatinine level [153]. In contrast, uric acid concentrations in adults saliva were high, but not significantly different to those in neonatal saliva (p=0.9). The level of the deoxy- forms of nucleosides, orotidine, succinyladenosine, AICAR were negligible (<1 µM) in both neonatal and adult saliva (not shown).

60

A 120

100 M)  80

60 40

Concentration ( Concentration 20

0

Uracil Uridine XanthineAdenine Inosine Thymine Adenosine Guanosine Orotic acid Dihydrouracil PseudouridineHypoxanthine Dihydrothymine

120 B 100

M)  80

60

40

Concentration ( Concentration 20

0

Uracil Inosine Uridine XanthineAdenine Thymine Adenosine Guanosine Orotic acid Dihydrouracil PseudouridineHypoxanthine Dihydrothymine

Figure 2‎ .7 Distribution and median values of nucleotide metabolites in adults and neonatal whole saliva. (A) adults, n= 77 and (B) full-term neonates, n=60. Straight lines show median values (µM).

61

Table 2‎ .8 Salivary levels (median with range, µM) of nucleotide metabolites in neonates compared to adults

Neonates adults Nucleotide metabolites (n=60) (n=77) p value Median Range Median Range Creatinine 22 *<5.0-58 6.6 <5-12 <0.0001 Pseudouridine 2.1 0.5-11 0.3 <0.2-1.0 <0.0001

Uracil 7.3 <1.5-95 <1.5 <1.5-21 n.a

Hypoxanthine 27 4.1-113 2.1 0.2-22 <0.0001

Xanthine 19 1.2-92 1.7 0.8-33 <0.0001

Adenine 1.3 <0.1-14 0.7 <0.1-3.7 0.15(n.s.)

Adenosine 12 0.6-76 0.1 <0.03-1.2 <0.0001

Inosine 11 0.8-99 0.2 <0.06-51 <0.0001

Guanosine 6.8 0.9-35 0.1 <0.03-0.7 <0.0001

Dihydrothymine 2.0 <1.5-3.9 <1.5 <1.5-1.8 n.a

Dihydrouracil 5.2 <3.0-25 <3.0 <3.0-3.4 n.a

Uridine 12 2.9-54 0.4 <0.02-8.4 <0.0001

Thymine <1.5 <1.5 <1.5 <1.5-4.4 <0.0001

Orotic acid <0.5 <0.5-26 <1.5 <1.5 n.a

Uric acid 133 26-509 173 39-611 0.07(n.s.)

*<: less than LOQ n.s.: not significant n.a: not applicable

62

2.5.5 INFANT LONGITUDINAL STUDY Aim 5: To conduct a 12-month longitudinal study of the nucleobases and nucleosides metabolome of saliva of newborn babies through to weaning and early infancy

In this longitudinal study, saliva samples were collected from infants aged 1-4 days, 6 weeks, 6 months and 12 months. Figure 2.8 shows the pattern of salivary concentrations of purine and pyrimidine metabolites decreasing from birth until they attained adult values by 12 months of age. The median and range concentrations of each metabolite are shown in Table 2.9. In particular, the purine metabolites hypoxanthine, xanthine, inosine and guanosine gradually decreased from 6 weeks to adult levels at 6-12 months of age. In contrast, the pyrimidine metabolites uracil and uridine decreased sharply to adult levels by 6 weeks of age. The primary pyrimidine precursor orotic acid was unusual because while the median concentrations were less than 1 µM for all time periods (p=0.2), some neonatal saliva samples greatly exceeded 10 µM and which remained stable for all 4 time points examined. Uric acid levels were also exceptional, with median values of 133 µM at birth, decreasing to 81.55 µM at 6 weeks of age, then gradually increasing to 199 µM at 6 months and 379 µM at 12 months. The major nucleotide metabolites of 10 individual neonates out of 14 who participated in every time point of the longitudinal study are shown in Figure 2.9.

30 M)

 25 Uracil Hypoxanthine 20 Xanthine 15 Adenosine

10 Inosine Guanosine 5 Uridine

Median concentration ( concentration Median Uric acid 0

1-4 days 6 weeks 6 months 12 months

Figure 2‎ .8 Median concentrations of nucleotide metabolites in whole saliva of full-term infants 1-4 days (n=60), 6 weeks (n=20), 6 months (n=19), 12 months (n=14).

63

100 100 1 1 Hypoxanthine Xanthine

80 2 80 2

M) M)   3 3 60 4 60 4 5 5 40 40 6 6

7 7 Concentration ( Concentration Concentration ( Concentration 20 20 8 8 0 9 0 9 10 10

1-4 days 6 weeks 6 months 1-4 days 6 weeks 6 months 12 months 12 months

100 100 1 1 Uracil Adenosine

80 2 80 2

M) M)   3 3 60 4 60 4 5 5 40 40 6 6

7 7 Concentration ( Concentration Concentration ( Concentration 20 20 8 8 0 9 0 9 10 10

1-4 days 6 weeks 6 months 1-4 days 6 weeks 6 months 12 months 12 months

100 100 1 1 Inosine Guanosine

80 2 80 2

M) M)   3 3 60 4 60 4 5 5 40 40 6 6

7 7 Concentration ( Concentration Concentration ( Concentration 20 20 8 8 0 9 0 9 10 10

1-4 days 6 weeks 6 months 1-4 days 6 weeks 6 months 12 months 12 months

100 1 Uridine

80 2 M)

 3 60 4 5 40 6 7 Concentration ( Concentration 20 8 0 9 10

1-4 days 6 weeks 6 months 12 months

Figure 2.9 Median concentrations of the major nucleotide metabolites in whole saliva of 10 followed up full-term infants at 1-4 days, 6 weeks, 6 months, 12 months.

64

Table 2‎ .9 Salivary levels (median with range, µM) of nucleotide metabolites in a longitudinal study of infants’ saliva

1-4 days 6 weeks 6 months 12 months Nucleotide (n=60) (n=20) (n=19) (n=14) p values metabolites Median Range Median Range Median Range Median Range Pseudouridine 2.1 0.5-11 0.9 0.3-1.1 0.3 *<0.2-6.0 <0.2 <0.2 <0.0001 Uracil 5.3 <1.5-95 <1.5 <1.5 1.6 <1.5-2.7 <1.5 <1.5-5.0 <0.0001

Hypoxanthine 27 4.1-113 6.9 0.3-31 1.1 0.4-2 0.8 <0.2-12 <0.001

Xanthine 19 1.2-92 7.0 0.7-25 1.7 <0.7-4.8 1.7 <0.7-12 <0.001

Adenine 0.9 <0.1-14 0.8 0.1-7.3 0.2 <0.1-0.8 0.7 0.3-3.9 0.001

Adenosine 12 0.6-76 7.0 1.2-29 0.9 <0.03-2.2 0.8 0.03-1.2 <0.0001

Inosine 11 0.8-99 4.9 0.3-41 0.3 0.1-1.3 0.4 0.1-1.5 <0.0001

Guanosine 6.8 0.9-35 5.6 0.2-21 0.5 0.1-1.5 0.4 0.1-1.0 <0.0001

Dihydrothymine 1.6 <1.5-3.9 <1.5 <1.5 <1.5 <1.5 <1.5 <1.5 n.a

Dihydrouracil 5.2 <3.0-25 <3.0 <3.0-10 <3.0 <3.0-4.5 <3.0 <3.0-5.6 n.a

Uridine 12 2.9-54 0.8 0.1-3.3 0.3 0.1-0.9 0.9 0.5-7.5 <0.0001

Thymine <1.5 <1.5 <1.5 <1.5 <1.5 <1.5-11 <1.5 <1.5-2.0 <0.0001

Orotic acid <0.5 <0.5-26 <0.5 <0.5-48 <0.5 <0.5-14 <0.5 <0.5-12 n.a

Uric acid 133 26-509 81.55 18-232 199 81-352 379 185-978 <0.0001

*<: less than LOQ; n.s.: not significant, n.a: not applicable

65

2.5.6 ANIMAL SALIVARY NUCLEOTIDE METABOLITES Aim 6: To conduct a survey of saliva from other mammalian species to investigate and compare the salivary nucleotide metabolite patterns between mammals and humans

Nucleotide metabolite patterns in mammalian saliva samples are summarised in Figure 2.10 and the descriptive statistics of the data are shown in Table 2.10 (see Appendices 18-24 for full data set). Substantial variability was observed among species. Similar to humans, salivary deoxy-nucleosides were generally negligible in most species, but horse saliva contained high levels of deoxyuridine, deoxyinosine, (deoxy) thymidine and . The pyrimidine base uracil was elevated in the non-human mammalian saliva (similar to neonate humans) compared to adult humans, with horse saliva showing the highest level (median=81 µM) followed by goats (45 µM), cats (33 µM), and then the camel (24 µM). A similar species distribution was found for the pyrimidine nucleoside uridine, and for the purine bases xanthine and hypoxanthine.

Overall, the lowest levels of nucleosides and bases were detected in bovine and ovine saliva, with the ovine results proving to be markedly different from the goat results, even though these animals are considered to be evolutionarily close. Curiously, canine saliva differed from all other animals with respect to adenine metabolites; it had high concentrations of adenosine (median=19 µM) and adenine (5.7 µM), which were very low in the other species. As expected, uric acid median concentrations were generally low (30-60 µM) compared to humans, because non-human mammals have uricase which metabolises uric acid to allantoin, however, in the horses the median concentration was 193 µM.

66

Group1 Group2 Group3 Group4 Group5

Figure 2‎ .10 Mammalian salivary nucleotide metabolite concentrations. Metabolites were divided into functional groups; Group 1: Pyrimidines (Pseudouridine to Orotate); Group 2: Purine bases (Hypoxanthine/Xanthine); Group 3: Purine nucleosides (Inosine/Guanosine), Group 4: ATP metabolites (Adenine/Adenosine); Group 5: DNA metabolites (Deoxyadenosine to Deoxyuridine). The values are shown as median concentrations, µM.

67

Table 2‎ .10 Salivary levels (median with range, µM) of nucleotide metabolites in a selection of mammals

Cow Sheep Goat Horse Camel Dog Cat Nucleotide (n=8) (n=5) (n=4) (n=5) (n=1) (n=7) (n=5) metabolites Median Range Median Range Median Range Median Range Median Median Range Median Range Pseudouridine 0.2 *<0.2-0.5 0.2 0.2-0.4 1.6 1.5-1.7 2.3 1.2-5.0 1.3 0.4 <0.2-1.1 1.2 0.4-2.2 Uracil 3.8 1.6-5.8 8.3 3.4-9.7 45 30-61 81 28-130 24 12 <1.5-37 33 25-190 Hypoxanthine 0.2 <0.2-0.4 1.1 0.7-1.3 22 13-32 6.9 0.6-19 15 0.2 <0.2-0.5 12 0.5-68 Xanthine <0.7 <0.7 1.7 0.9-7.8 27 14-31 34 2.4-51 9.3 1.3 <0.7-4.2 11 2.1-115 Adenine 0.1 <0.1-0.6 1.2 0.1-1.6 3.3 1.1-4.2 0.2 0.1-2.5 2.5 5.7 0.1-19.3 1.8 0.2-38 Adenosine 0.2 <0.03-1.2 8 2.1-11 2.7 1.4-4.0 0.1 0.1-0.2 1.4 19 0.8-36 2 0.7-10 Deoxyadenosine 0.1 <0.02-0.3 0.2 0.1-2 0.2 0.1-0.6 0.04 0.02-0.1 0.5 3.6 0.04-18 2.5 0.4-2.7 Deoxyuridine <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 13 6.2-25 4.8 1.1 <0.5-3.7 4.5 1.6-25 Inosine 0.7 <0.2-3 1.1 1.0-2.0 5.7 4.8-27 9.4 5.4-27 10 3.4 0.4-11 10 1.8-31 Guanosine 0.3 0.03-3.9 2.8 1.9-6.5 5.2 2.7-30 10 6.2-40 10 3.3 0.5-6.0 12 2.6-33 Deoxyinosine <0.4 <0.4-1.9 <0.4 <0.4 0.7 0.6-0.9 10 5.7-19 2.9 1.0 <0.4-4.8 5.2 2.1-27 Deoxyguanosine 0.4 0.1-2.0 0.2 0.2-0.3 0.4 0.3-0.8 7.4 3.0-15 2.6 2.3 0.1-6.0 3.6 1.5-21 Thymidine 0.6 <0.3-1.7 0.4 <0.3-0.6 2.4 1.8-5.1 18 7.6-25 5.5 2.6 <0.3-14 13 4.6-48 Dihydrouracil <3.0 <3.0-6.5 6.3 <3.0-10 7.6 <3.0-16 <3.0 <3.0 <3.0 5.0 <3.0-6.3 <3.0 <3.0 Uridine 0.4 0.1-3.9 2.2 1.5-3.7 10 5.6-48 30 20-94 18 3.9 0.4-9.3 13 3.3-40 Orotate <0.5 <0.5 <0.5 <0.5 <0.5 <0.5-1.0 1.2 0.5-1.4 <0.5 <0.5 <0.5 1.3 <0.5-3 Urate 35 26 -40 42 40-44 59 48-103 193 71-394 47 45 34-62 47 31-68 Thymine <1.5 <1.5 <1.5 <1.5 4.1 2.2-10 26 12-48 9 2.1 <1.5-8.4 16 12-75 *<: less than LOQ

68

2.6 DISCUSSION

This project was inspired initially from the use of reverse phase HPLC with UV-photodiode array (PDA), which has traditionally been used to screen purine and pyrimidine metabolites [146], to examine a small number of saliva samples as a pilot study. Subsequently, a pilot study of several infants ranging in age from newborn to 3 years led to the serendipitous finding that neonatal saliva had higher concentrations of nucleosides and bases compared to adults, and that the transition between infant and adult saliva patterns of purines and pyrimidines appeared to occur during the first year of life. This led to the formation of the first two Specific Hypotheses (Section 2.2).

Amino acids, proteins, anions, vitamins, hormones and other small molecules have been intensively studied in saliva and found to include important biomarkers for several diseases [91, 93, 154-156]. But only one study of nucleotide metabolites in human saliva has been published [117]. For this part of the project, it was hypothesised that the presence of nucleotide metabolites in saliva may have an important role in oral biology, especially in neonates. Characterisation of these metabolites in saliva also may facilitate investigation of inborn errors of metabolism of purines and pyrimidines, as it was found that the salivary patterns were unique compared to plasma or urine.

2.6.1 COLLECTION OF SALIVA SAMPLES Adult saliva collection by passive drool was found to be simple, non-invasive and easier to standardise than stimulated saliva. It was not, however, completely without ethical restrictions; some adults felt 'embarrassed' by having to salivate into a container and declined to participate. In some cases it may be simpler to use cotton buds for collection of adult saliva, to avoid this restriction.

However, neonates and early infants often produce only small amounts of saliva compared to adults [157], and this presented a particular challenge to this project. In addition, there were ethical concerns about stimulating saliva in neonates and early infants. It was found that obtaining adequate volumes of saliva directly from neonates (e.g. using plastic pipettes) proved difficult when compared to non-invasive oral swabbing. Saliva collections from neonates and infants by cotton swabs previously have been used to investigate various biochemical markers and proteins such as cortisol [158], melatonin [159, 160] and immunoglobulins [161]. The oral swab collection method needed to overcome problems of; (1) contaminants in the cotton; (2) absorbance of the metabolites onto the cotton, which then acted like a reversed-phase medium and required a 2-step solvent elution to fully remove the various nucleosides and bases; (3) careful sample handling and weighing

69 of swabs before and after sampling, was essential to obtain accurate values for the volumes of saliva collected so that concentrations could be accurately calculated.

2.6.2 HPLC-MS/MS DETERMINATION OF NUCLEOTIDE METABOLITES HPLC-PDA analysis was used to develop methods for the extraction of the nucleosides and bases from cotton swabs used to collect saliva from the young infants. However, LC-MS/MS was subsequently adopted as the preferred method for analysis of large numbers of saliva samples from screening adults and infants, and a selection of domestic animals. There were at least 4 reasons why MS/MS analysis was selected for saliva analyses; (1) Greater sensitivity, as the adult saliva in the pilot studies had metabolite values as low as 1 µM, similar to that reported by Kochanska et. al. [117] and this sensitivity was not achievable by UV/PDA; (2) Greater specificity, MS/MS provided better confirmation of the identity of metabolites than UV spectra; (3) Ability to measure metabolites with poor UV spectral absorbance, and this was especially true for the dihydropyrimidines; (4) The increased specificity and sensitivity inherent in MS/MS, which allows overlapping or co-eluting metabolite peaks to be distinguished by their mass characteristics, meant that a more efficient reverse phase HPLC method could be developed. This method required only an 18 min run time compared to 35 min using UV/PDA (for which metabolite peaks must be fully resolved to be quantifiable).

Nucleotides have not been analysed at this stage because of limitations on time and access to equipment during the thesis research, and this should be a focus of future work (see Future Studies, Chapter 5).

2.6.3 ADULT SALIVA In human adult saliva, nucleotide metabolite concentrations were found to be low (median ≤2 µM) except for uric acid. This agreed with the results of a previous small survey of 23 adults [117] in which saliva was reported to contain hypoxanthine, xanthine and inosine at low levels, and with uric acid concentrations similar to those in plasma. The concentrations of the metabolites reported here were similar to those previously reported, but the metabolites examined here provided an in- depth profile of purines and pyrimidines, and show a large range of values. Interestingly, hypoxanthine, xanthine, inosine and uracil were elevated in several adult participants, which suggested that there was either; (a) genetic polymorphisms in the population that regulate salivary nucleotide metabolites or; (b) that these specific metabolites may be markers for oral health, dental hygiene, diet, or commensal bacteria. The latter hypothesis, while simpler, was excluded by the

70 results found for infant saliva discussed below. These studies suggested that there may be genetic differences within populations with regard to salivary metabolites; this has not previously been examined. For the purposes of this project, all non-Caucasian participants were deliberately excluded in order to reduce genetic variability. Thus, there is scope for further studies of other ethnic/racial groups.

In plasma, nucleotide metabolites are normally present at less than 1 µM concentrations, with the exception of uric acid and uridine [4, 142-145, 162]. Importantly, measurement of hypoxanthine in plasma is only reliable when blood cells are separated from plasma within a very short time, as this metabolite (and ammonia) increases rapidly after blood sample collection due to the breakdown of blood cell ATP [163]. Concentrations of some purine and pyrimidine bases (but not nucleosides) can be higher in normal urine [146]. Nucleotides usually exist only intracellularly, and are indicators of cellular damage when they appear in extracellular fluids.

2.6.4 INFANT SALIVA Most of the salivary metabolite concentrations in neonates were high and not simply in equilibrium with plasma. These data suggest that specific transport mechanisms may operate in neonates. Nucleosides and bases from neonates have been measured in umbilical cord blood [163, 164], plasma [163], and urine [165]. These nucleotide metabolites are also in low concentrations in neonatal cerebrospinal fluid [162]. Raised hypoxanthine concentrations in newborn umbilical cord blood and plasma have been reported as a biomarker for birth hypoxia [166-168].

Urate was found in high concentrations in neonatal saliva. This purine metabolite accumulates in saliva of humans and higher apes due to the absence of uricase. In other mammals the role of urate is as a minor intermediate metabolite, and therefore mammals have not evolved specific urate transporters, and this absence continues in humans/higher apes. Urate is thought to undergo non- specific transport in mammalian cells via anion transporters and the fructose transporter SLC2A9 [169, 170], which have comparable activities in neonates and adults. The raised urate levels observed in the saliva of newborns presumably originate from raised maternal urate concentrations, which peak during birth. Saliva is produced by the tubuloacinar cells of the salivary glands, while uric acid in saliva appears to be a dialysate of plasma. It was observed that there is significant variability in uric acid concentrations in infant saliva from birth up to 12 months of age. The lowest concentration was detected at 6 weeks of age where breast milk is likely to be the main source of uric acid. In contrast, the uric acid concentration at birth was higher than at 6 weeks of age

71 suggesting that this was of maternal origin. The highest level of uric acid, which approached adult levels was detected at 12 months of age when most infants depend on breast and/or formula feeding as well as from introduction of solids [171].

The highly selective and species-dependent presence of such metabolites in high levels in early life suggested an essential role at this stage where the immunity of the newborn is still developing. This observation of a gradual decrease in nucleotide metabolites in young children may be explained by the growth and development of infants during this period.

2.6.5 TRANSITIONING FROM INFANT TO ADULT: THE LONGITUDINAL STUDY The investigation of nucleotide metabolite levels in the saliva of young children at various ages is of value in understanding the pattern of changes during growth. Therefore, having observed the significant differences between neonatal and adult saliva with respect to the pattern of purines and pyrimidines, a longitudinal study was undertaken to cover the first year of life of individual infants. The results showed that the median salivary concentrations of some nucleotide metabolites decreased rapidly particularly for pyrimidines such as uracil and uridine, while others declined more gradually over the first year of life. There were significant differences between all metabolites, with the lowest levels detected at 12 months of age, except where levels of some metabolites remained low throughout the study (orotic acid and adenine). However, a few infants showed elevated concentrations of the pyrimidine base orotic acid; this may have been secondary to the feeding of infant milk formula of bovine origin which contains high levels of orotic acid [172], in contrast to human milk, which contains high levels of the pyrimidine nucleoside uridine but not orotic acid.

Individual variation in salivary patterns was found in infants, as in adults. This was particularly noticeable in the longitudinal study. These highly variable levels of metabolites found in neonates cannot be attributed to oral health or microbiota as these subjects were only hours old and would be unlikely to have pathologies such as gum disease. All infants sampled for saliva were in good health, as exclusion factors for saliva collection included no history of infection and they were not receiving antibiotics. In addition, while they were all delivered vaginally they were sampled at least 6 hours after birth when it was judged that trauma from birth should have subsided. In fact, these differences in patterns of metabolites between infants persisted well beyond the neonatal period and were present in some infants at 6 weeks.

72

2.6.6 SALIVARY PATTERNS IN OTHER MAMMALS The third Specific Hypothesis (Section 2.2) was that purine and pyrimidine metabolites would be found also in other mammals. This was not a particularly bold proposal, but what surprising was that in non-human mammalian saliva the concentration of nucleotide metabolites showed significant inter-species variability. Generally, the (adult) animals studied showed elevated levels of various nucleosides and bases compared to human adults, except for sheep and cows, which usually had the lowest amounts of salivary metabolites. Uracil was the most prevalent of the nucleotide salivary metabolites, with the highest levels occurring in horse saliva (median 80.5 µM). This is not salvaged by mammalian cells, but can be utilised by bacteria to form pyrimidine nucleotides [36].

There was considerable individual variability within animal species. Similar to humans, there may be genetic variability in salivary metabolites among other mammals, but for the examples of domesticated animals used here, more variation would be expected among outbred animals (dogs, cats) or between breeds (horses), but not for inbred strains (sheep, cattle, goats). For example, levels of nucleotide metabolites in goats were elevated compared to sheep and cows. Elevated levels of deoxynucleosides were detected in all horse saliva and in some individual cats, but these were negligible in human saliva and in the other animals. The sample sizes for the domestic/farm animals here were low, so variability was not fully tested. It is not possible to comment on the camel results, as only one animal was tested. However, these animal studies were intended as a pilot to encourage future research.

The highest levels of adenosine and adenine were present in canine saliva. Adenosine and adenine are products of the breakdown of adenine nucleotides. Free adenine is unusual; this is usually recycled to ATP within cells via adenine phosphoribosyltransferase. Adenosine is not usually found in free form in body fluids, as it is efficiently salvaged back to ATP via adenosine kinase, however adenosine is thought to function as an extracellular 'alarm signal' [173]. Interestingly, adenosine in the brain is also proposed to act as a homeostatic regulator of sleep and to be a mediator between the humoral and neural mechanisms of sleep-wake regulation [174, 175].

With the exception of sheep and cattle, the high levels of salivary nucleosides and bases in the sampled non-human mammals more closely resembled those of human babies than human adults, with consideration given to intra-species variability. Presumably this may provide further insights into preferred metabolic pathways in differing species. The difference between the major milk pyrimidine noted above (uridine in humans vs orotic acid in cattle) is suggestive of preferential metabolic pathways in animal species, and this aspect in saliva warrants further investigation.

73

2.6.7 CONCLUSIONS AND FURTHER HYPOTHESES The present study provided some important biological data for salivary nucleotide metabolite levels in healthy neonates and adults (see Chapters 3 and 4 for investigation of proposed roles of nucleotide metabolites in neonatal saliva). The presence of these purine and pyrimidine metabolites was both species dependent and developmentally regulated (in humans). The absence of de novo synthesis intermediates, such as SAICAR or AICAR (as ribotides), and of catabolic intermediates such as the dihydropyrimidines, is evidence for the functional importance of the nucleosides and bases that were found in saliva.

Saliva is a bodily fluid that can be obtained non-invasively, and it appears that some classes of metabolites might be actively secreted into saliva, e.g. hypoxanthine and xanthine appear in saliva at much higher concentrations than in plasma. On the other hand, some metabolites occur in low concentrations in saliva despite having high levels in plasma (e.g. creatinine), therefore there may be advantages in using saliva in metabolic studies.

The biochemical investigations of nucleosides and bases in human saliva, presented in this chapter, led to the formulation of 3 further hypotheses: 1. Based on the substantially raised xanthine and hypoxanthine concentrations in neonatal saliva, it was considered that these two salivary metabolites, which are both substrates for xanthine oxidase, may react with this enzyme in milk during breast-feeding, to produce hydrogen peroxide (see Chapter 3). This peroxide is considered to be an important molecule in triggering other reaction cascades of immune reactive oxygen species which could be crucial for infant innate immunity, such as the lactoperoxidase system [76, 176]. 2. The other elevated bases and nucleosides in neonatal saliva may have a role as nucleotide metabolites enhancing the selection for oral commensal microbiome during early stages of life and, accordingly, enhance innate immunity, (see Chapter 4 for further details). 3. Furthermore, these nucleotide metabolites could be also utilised by the gut apical cells for their growth. Uauy et al (1994) reported that rapidly growing tissues such as intestinal epithelium cells lack significant capacity for de novo synthesis of nucleotides and require exogenous sources of purine and pyrimidine bases. The study suggested these metabolites are utilised by intestinal cells for proliferation with excess purines being converted to uric acid [17]. Furthermore, nucleoside supplementation in weanling rats showed acceleration of intestinal maturation compared with a control group on a nucleoside-free diet [177]. This thought-provoking hypothesis was not pursued for this project, but could be an interesting future study.

74

Chapter 3

Human milk xanthine oxidase

and neonatal salivary nucleotide metabolites:

Their role in generating hydrogen peroxide in the neonatal mouth

75

3.1 INTRODUCTION

As discussed in Section 1.2.8, xanthine oxidoreductase (XOR) is a complex molybdeno- flavoprotein which catalyses the oxidation of the purine bases hypoxanthine to xanthine to uric acid. The enzyme exists in two protein forms of the same gene product, referred to as xanthine oxidase (XO) and xanthine dehydrogenase (XDH): XO reduces oxygen and water to hydrogen peroxide + (H2O2) and superoxide anion, while XDH reduces NAD to NADH [43, 50, 55].

XO has been the focus of considerable research into ischemia reperfusion injury [62, 63], however there is increasing evidence that XO has important physiological functions associated with its synthesis of ROS and RNS where the enzyme is viewed as an evolutionarily conserved component of the innate immune system [59]. XO activity has been found in the milk of all mammals studied so far [48], being particularly high in bovine milk in comparison to human milk [67]. Its role is considered primarily to produce peroxide, which then acts as an antibacterial agent in both the milk glands (to prevent mastitis) and in the milk itself to inhibit microbial growth [79]. Furthermore, peroxide free radicals can combine with thiocyanate (SCN-), which is abundant in adult and infant saliva catalysed by the enzyme lactoperoxidase (LPO), which is also present in milk and saliva, to produce a more potent anti-bacterial free radical hypothiocyanite (OSCN-): this is known as the 'lactoperoxidase system' [76-78] (Fig. 3.1).

Figure ‎3.1 The lactoperoxidase system. In the first step, milk xanthine oxidase (XO) generates hydrogen peroxide (H2O2) which is used by lactoperoxidase (LPO) to convert thiocyanate into antibacterial hypothiocyanite. Note that hypoxanthine and xanthine as well as thiocyanate are abundant in neonatal saliva.

The current studies have revealed appreciable concentrations of xanthine and hypoxanthine in human neonatal saliva (see Chapter 2). This observation raised the possibility that these metabolites could play an important role in generating H2O2 in the upper gastrointestinal tract of the neonate during suckling by re-activating breast milk XO production of H2O2, especially within the first few weeks postpartum when milk XO activity is high.

This chapter is focused on a study aimed at validating an assay of milk XO and investigating its kinetics in fresh human breast milk, to confirm its ability to produce H2O2 when mixed with

76 neonatal saliva thereby emulating the reaction in the infant’s mouth. In addition, H2O2 concentrations in fresh breast milk were measured to determine the physiological range of peroxide, as a prelude to studying its effects on bacterial species (Chapter 4).

3.2 SPECIFIC HYPOTHESIS AND RATIONALE

Hypothesis: Xanthine and hypoxanthine in neonatal saliva react with breast milk xanthine oxidase (XO) in an infant’s mouth to boost hydrogen peroxide to biologically active levels.

Rationale: The mechanism of H2O2 production by XO in milk has been described previously for bovine milk, as a means of reducing mammary infections. But this has never been linked to saliva interactions during suckling. This hypothesis would provide an explanation of secretion of hypoxanthine and xanthine in the neonatal saliva, and XO in breast milk. The role of XO in milk can thus been seen to extend beyond the breast.

Infancy is a stage of life characterised by immaturity of the ‘specific’ immune system (i.e. antibody- immune cell regulated). Reactive oxygen species such as peroxide are a well-known biological defence mechanism to inhibit and kill microbial pathogens. A similar mechanism may occur for both the breast and infant mouth and digestive tract as part of the neonatal innate immune system.

3.3 SPECIFIC AIMS

1. To establish an assay for XO and determine the activity in breast milk collected during the first week after delivery (i.e. postpartum).

2. To estimate the endogenous concentration of H2O2 in fresh breast milk

3. To determine the H2O2 generation during the interaction between neonatal saliva and breast milk.

77

3.4 MATERIALS AND METHODS

3.4.1 CLINICAL SAMPLES Prior written ethical approval for the collection of breast milk and saliva samples was obtained from the Human Research Ethics Committees of Mater Health Services, and The University of Queensland (see Appendices 1&2). The participating breastfeeding mothers' ages and gestational period, and the postpartum ages of the milk samples are shown in Table 3.1. Participants were selected only from those identifying themselves as Caucasian in order to minimise potential race- related genomic variability.

Breast milk samples were collected from 24 mothers, 1 to 5 days postpartum, by a research nurse using sterile gloves and hand expression into 50 mL sterile containers (Sarstedt Ply Ltd, Mawson Lakes, SA, Australia). Milk samples were placed on ice without delay and then transferred to the laboratory where they were subdivided into portions for the immediate assay of endogenous H2O2, or for storage at -80oC for further analyses such as XO assays. Pre-term and term Aptamil Gold Plus® (Nutricia Australia Pty Ltd, Macquarie Park, NSW, Australia) infant formula and pasteurised breast milk were obtained for XO activity assays from the Division of Neonatology, Mater Mothers’ Hospital. Commercial pasteurised bovine milk was purchased fresh from a local retail outlet. Neonatal saliva samples were collected from 60 infants who were aged 1-4 days (see Chapter 2 for details).

Table 3‎ .1 Information about breast milk samples and donor mothers

Gestational age (weeks) Postpartum age (hours) Mother's Age (years) Mean 39.5 52.8 28.9

SD 1.5 28.0 6.0

Range 37.3 - 42.3 15 -117 20 - 39

3.4.2 REAGENT PREPARATIONS

H2O2 (Merck Pty Ltd, Kilsyth, VIC, Australia), 30% v/v, 9 M, was used as a calibration standard by dilution to approximately 9 mM in 100 mM Tris-HCl pH 7.5 ('Tris buffer'). Concentrations of H2O2 were accurately determined spectrophotometrically in triplicate at 240 nm, assuming a molar absorptivity of 43.6 M-1 cm-1.

Peroxidase reagent containing 100 µM of Ampliflu Red (Sigma-Aldrich Pty, Castle Hill, NSW, Australia) and 0.8 U/mL horseradish peroxidase (Sigma-Aldrich Pty, Castle Hill, NSW, Australia) 78 was supplemented with hypoxanthine to give the following series of reaction concentrations in Tris buffer: 0, 8, 12.5, 25, 50,100, 200, 300, 400 μM. Buffers and aqueous reagents for all experiments were prepared from MilliQ water.

3.4.3 XANTHINE OXIDASE ASSAYS

XO activity was measured by monitoring the rate of H2O2 production during the oxidation of hypoxanthine to uric acid. Note that one mole of hypoxanthine produces 2 moles of H2O2, via the sequential reaction with xanthine. In the presence of horseradish peroxidase, H2O2 converts Ampliflu Red (10-acetyl-3,7-dihydroxyphenoxazine) into resorufin, which is a red fluorescent product having excitation and emission peak wavelengths of 544 nm and 590 nm respectively (Fig. 3.2). The assay used raw breast milk, stored at -80 oC, without pre-treatment. Breast milk samples were diluted 1:30 (v/v) in Tris buffer, then 50 μL of diluted milk was mixed with 50 μL of peroxidase reagent in a 96-well microtiter plate (Becton Dickinson, North Ryde, NSW, Australia) to give a total reaction volume of 100 µL. The microtiter plate was incubated at 37oC for 60 min in a temperature-controlled FLUOstar Omega fluorimeter (BMG Labtech, Cary, NC, USA). The fluorescence of the formed product, resorufin, was measured fluorimetrically every 30 s for 5 min, then every 5 min up to 60 min at 544 nm (excitation) and 590 nm (emission). Each experiment was performed in duplicate.

For each breast milk sample, a corresponding assay substrate blank (i.e. excluding hypoxanthine) was subtracted from each data point. The XO kinetic parameters were estimated for 6 breast milk samples using hypoxanthine concentrations of 0, 8, 12.5, 25, 50, 100, 200, 300, 400 µM as the final reaction concentration. XO activity of the remaining milk samples (n=18) were assayed directly using only 400 µM hypoxanthine working solution as the reaction concentration. XO activity was also assessed in the infant Aptamil Gold Plus®, pasteurised breast milk, and pasteurised bovine milk as described above.

The H2O2 produced from the XO reaction was calculated from a standard curve generated by adding

50 µL of H2O2 (0, 0.03, 0.06, 0.13, 0.25, 0.50, 1, 2, 3, 4, 5, 6, 7 µM) to microtiter wells each containing 50 µL of peroxidase reagent to give a reaction volume of 100 µL. The reaction mixture was incubated at 37oC for 60 min and the resorufin florescence was measured as described above.

The rate of H2O2 production by XO was then estimated using the slope of the initial linear region of the enzyme reaction. The final activity of XO (U/L breast milk) was calculated from the following equation:

79

In this study the breast milk dilution factor was 60, and a unit (U) of XO activity was defined as the production of l µmol of H2O2 per min, using hypoxanthine as the substrate. The parameters of the

Michaelis-Menten kinetic model, Vmax and Km, for breast milk XO were estimated using nonlinear regression analysis. The lower limit for detection of activity under these conditions of assay was 0.1 U/L.

To confirm that the H2O2 was produced by the activity of breast milk XO and not another enzyme reaction, breast milk pooled from 10 samples was supplemented with oxypurinol (Sigma-Aldrich Pty, Castle Hill, NSW, Australia), a XO-specific inhibitor, to give reaction concentrations of (0, 0.38, 0.75, 1.5, 3, 6, 12, 25 µM) then incubated as above with peroxidase reagent containing 400 μM of hypoxanthine.

Figure 3‎ .2 Principle of coupled enzymatic assays of XO using horseradish peroxidase-Ampliflu Red reagent. Oxidation of hypoxanthine or xanthine by XO results in generation of H2O2, which is coupled to conversion of the Ampliflu Red reagent to fluorescent resorufin by horseradish peroxidase (HRP).

3.4.4 MEASUREMENT OF ENDOGENOUS HYDROGEN PEROXIDE IN HUMAN MILK Freshly expressed breast milk samples were diluted 1:5 (v/v) in Tris buffer, then in duplicate, 100 μL of diluted milk was added to 1.5 mL polypropylene Eppendorf tubes containing 100 μL of peroxidase reagent (100 µM Ampliflu Red and 0.8 U/mL horse radish peroxidase in Tris buffer). 80

The reaction mixture was incubated in a water bath at 37oC for 60 min, then 150 μL of the mixture was microcentrifuged through an ultrafiltration membrane (Grace Davison Discovery Sciences, Rowville, VIC, Australia), 0.45 µm pore size, at 4500 g for 3 min, 4oC, to remove turbidity. The clear filtrate was transferred to a 96-well microtiter plate and the fluorescence was measured as described above for the XO assay. A negative control contained Tris buffer in place of the breast milk, and the fluorescence value from this control was subtracted from the assay reaction.

To examine whether peroxide in breast milk is generated as a result of xanthine and hypoxanthine co-secretion, 10 breast milk samples were treated with 10% (w/v) trichloroacetic acid (TCA) immediately after the milk was expressed, to effect denaturation of XO and other proteins, then the samples were ultrafiltered and assayed immediately for xanthine and hypoxanthine (see Chapter 2 for LC-MS/MS method).

3.4.5 GENERATION OF HYDROGEN PEROXIDE BY MIXING HUMAN MILK WITH NEONATAL SALIVA As a proof-of-principle experiment, the amount of peroxide produced by a mixture of breast milk and neonatal saliva was measured. A pooled breast milk sample was diluted 1:6 in Tris buffer, while the pooled neonatal saliva (which was found to contain 70 µM hypoxanthine and 30 µM xanthine by LC-MS analysis) was diluted 1:5 in Tris buffer to produce 14 µM hypoxanthine and 6 µM xanthine as the reaction concentrations. In duplicate microtiter plate wells, an aliquot of 33 µL of diluted breast milk was mixed with 33 µL of diluted saliva and 34 µL of peroxidase reagent to give a reaction volume of 100 µL, which was incubated at 37oC for 60 min, with the fluorescence measured every 30 s for 5 min, then every 5 min up to 60 min, protected from light since peroxide is light-sensitive. Calculation of the H2O2 formed included correction for the dilutions above.

To assess the effect of endogenous breast milk lactoperoxidase and catalase enzymes on H2O2 consumption, 600 µL of diluted pooled breast milk (1:15) was prepared in two 5 mL glass test tubes. Then 10 µL of sodium azide solution (Sigma-Aldrich Pty, Australia), an inhibitor of haemoprotein enzymes (i.e. catalase and lactoperoxidase), was added to one of the tubes to give a reaction concentration of 400 µM. An equal volume of H2O2 was added to both tubes to give 60 µM reaction concentrations. Another control was prepared by mixing equal volumes of Tris buffer with o 60 µM H2O2. The reaction was incubated in a water bath at 37 C for 30 min. In a 96-well microtiter plate, 50 µL aliquots of the above tubes were added to 50 µL of peroxidase reagent at timed intervals (2, 5, 10, 15, and 30 min) to quantitate the concentration of peroxide. The fluorescence was measured as described above, and the result from a negative control containing peroxidase reagent and diluted breast milk was subtracted from the assay results.

81

3.5 RESULTS

3.5.1 BREAST MILK XANTHINE OXIDASE KINETICS Aim 1: To establish an assay for XO and determine the activity in breast milk collected during the first week after delivery (i.e. postpartum). The fluorescence standard curve generated by the reaction of Ampliflu Red with horseradish 2 peroxidase and varying H2O2 concentrations displayed acceptable linearity (r = 0.999) up to 7 µM, as shown in Figure 3.3. This standard curve was subsequently used for both measuring endogenous peroxide levels in milk and for assaying XO activity.

XO activity was assayed as the rate of formation of H2O2, by comparison with the peroxide standard curve. The enzyme rate was determined from the initial linear portion of the slope. H2O2 production versus incubation time was plotted for each substrate concentration (Fig. 3.4). Breast milk XO activity as a function of hypoxanthine concentration (8–400 µM) was analysed in 6 breast milk samples. The mean ± SD Km and Vmax values, estimated from the raw data shown in Figure 3.5, were 12.8 ± 2.1 µM and 8.9 ± 6.2 µmol/min respectively (Table 3.2). A hypoxanthine concentration of 400 µM was used for further studies.

Having validated a reliable and accurate assay method, XO activity was then determined in 24 samples of breast milk taken over period of 1-5 days postnatally. The activity of XO in the 24 breast milk samples ranged from 1.8 to 18.0 U/L with a mean activity of 8.0 U/L (Fig. 3.6). XO activity was undetectable (LOD < 0.1 U/L) in formula milk, pasteurised bovine milk, and pasteurised human breast milk.

XO activity versus the gestational age (in hours) and postpartum time of each milk sample were plotted to assess the correlation using linear regression analysis (Fig. 3.7). There was no correlation between XO activity and gestational age (r2<0.001). In comparison, XO vs postpartum time revealed a weak correlation (r2=0.2).

The addition of oxypurinol to breast milk showed an XO inhibitory, dose-dependent response with significant inhibition of H2O2 generation (p< 0.001). Maximum inhibition was obtained at 12 µM oxypurinol, with an apparent Ki 0.6 µM oxypurinol (Fig. 3.8).

82

50000 r2=0.999 y=6226x - 0.791 40000

30000

20000 Fluorescence 10000

0 0 1 2 3 4 5 6 7 8

H2O2 (M)

Figure 3‎ .3 Standard curve for H2O2 concentration versus fluorescence. The fluorescent red dye produced by conversion of Ampliflu Red to resorufin, in the presence horseradish peroxidase in a total volume of 100 µL including the assay reagents. Means and error bars (SD) were plotted. SDs were small and are masked by the symbols.

7 400 6 300 5 200

M) 100

 4

( 50 2

O 3

2 25 H 2 12.5 1 8 0 0 0 5 10 15 20 25 Time (min)

Figure 3‎ .4 Kinetics of H2O2 production by XO during conversion of hypoxanthine to uric acid. Diluted breast milk was incubated with peroxidase reagent containing varying concentrations of hypoxanthine. Rates were linear for at least 20 min. Each assay was conducted in duplicate. This graph represents one of 6 experiments with different breast milk samples.

83

16 14 12 10

mole/min) 8

 (

2 6 O

2 4 H 2 0 0 100 200 300 400 500 Hypoxanthine (M)

Figure 3‎ .5 Hyperbolic nonlinear regression curve showing XO velocity as a function of increasing hypoxanthine concentration. Each assay was conducted in duplicate. This graph represents one of 6 experiments with different breast milk samples.

Table 3‎ .2 XO kinetic parameters of breast milk samples (n=6), using hypoxanthine as substrate.

Sample Vmax (µmole/min) Km (µM)

1 14.2 10.8 2 2.40 15.9

3 2.20 13.2

4 15.0 12.7

5 14.0 13.9

6 5.50 10.1

Mean±SD 8.9±6.2 12.8±2.1

84

20

15

10

5 Xanthine oxidase (U /L) (U oxidase Xanthine

0

Figure 3‎ .6 Distribution of XO activity in breast milk. Samples were collected 1-5 day postpartum (n=24) and the milk was considered as colostrum. Error bars show mean±SD. The mean activity ± SD of XO was 8.0 ± 5.3 U/L.

85

A 2 0

)

L /

U 1 5

(

e

s

a

d i

x 1 0

o

e

n

i

h t

n 5

a X

0 6 0 0 0 6 5 0 0 7 0 0 0 7 5 0 0 G e s ta tio n a l a g e (h )

2 0

B

)

L /

U 1 5

(

e

s

a

d i

x 1 0

o

e

n

i

h t

n 5

a X

0 0 2 0 4 0 6 0 8 0 1 0 0 1 2 0 P o s tp a rtu m a g e (h )

Figure 3‎ .7 Scatterplots of XO activity versus the gestational and postpartum age in hour. No significant correlations were observed, although there appeared to be a weak positive correlation for postpartum age (r2=0.2).

86

1 6

1 4 )

n 1 2

i m

/ 1 0

e

l o

m 8

 (

2 6

O 2

H 4

2

0

0 8 5 .5 3 6 2 5 .3 .7 1 1 2 0 0

O x y p u rin o l (M )

Figure 3‎ .8 Dose dependent inhibition of breast milk XO by oxypurinol. The pooled breast milk sample were spiked with 0-25 µM oxypurinol concentrations in triplicate, then incubated with the peroxidase reagent. The graph shows an inhibitory dose-dependent response, with maximum inhibition obtained with 12 µM oxypurinol, confirming that H2O2 was produced by XO activity. Error bars show mean±SD.

87

3.5.2 ENDOGENOUS HYDROGEN PEROXIDE IN BREAST MILK Aim 2: To estimate the endogenous concentration of hydrogen peroxide in fresh breast milk

The mean ± SD endogenous H2O2 concentration was 27.3±12.2 µM (range: 11 to 51 µM), as shown in the vertical scatterplot in Figure 3.9. Samples treated with 10% TCA contained undetectable hypoxanthine and xanthine concentrations (LOQ= <0.2 and <0.7 µM respectively) using LC-

MS/MS. The XO activities of 21 of the 24 milk samples were plotted against the endogenous H2O2 concentration for each sample to assess any correlation, using a linear regression analysis method (Fig. 3.10). The data showed no correlation between the two parameters (r2< 0.001).

3.5.3 HYDROGEN PEROXIDE GENERATION Aim 3: To determine hydrogen peroxide generation during the interaction between neonatal saliva and breast milk.

The addition of neonatal saliva containing the substrates hypoxanthine and xanthine to breast milk generated an additional 40 µM of H2O2 after about 60 min of incubation, demonstrating that the reaction between XO in breast milk and the salivary XO substrates occurred in vitro and resulted in further peroxide generation, thus boosting the endogenous levels (Fig. 3.11).

The addition of exogenous H2O2 to breast milk resulted in a gradual consumption of peroxide with time. This H2O2 consumption was presumed to be caused by endogenous breast milk lactoperoxidase and catalase enzymes. This decrease caused by the enzymes was inhibited by sodium azide. A negative control showed stable peroxide concentrations during the experiment (Fig. 3.12).

88

60

50

40

M)

 (

2 30

O 2

H 20

10

0

Figure 3‎ .9 Distribution of concentrations of H2O2 in fresh breast milk samples (n=22). Error bars show the mean±SD (27.3±12.2 µM).

60

50

40

M)

 (

2 30

O 2

H 20

10

0 0 5 10 15 20 Xanthine oxidase (U/L)

Figure 3‎ .10 Scatterplot of XO activity versus endogenous H2O2 concentration in 21 of 24 breast 2 milk samples. There was negligible correlation between breast milk H2O2 and XO (r <0.001).

89

40

30

M)

 (

2 20

O

2 H 10

0 0 10 20 30 40 50 60 70 Time (min)

Figure 3‎ .11 Kinetics of H2O2 generation after mixing 33 µL of diluted breast milk, 33 µL of 1/3 diluted neonatal saliva and 34 µL peroxidase assay working solution (green circles). The negative control was buffer and peroxidase reagents (blue squares). The original concentrations of hypoxanthine and xanthine in the neonatal saliva were 70 µM and 30 µM respectively. Calculation of the H2O2 formed included correction for the dilution of saliva and milk.

50

40

30 mole/L)

 H2O2+Breast milk+ NaN3

( 2 20 H O +Buffer

O 2 2 2

H H O +Breast milk 10 2 2

0 0 5 10 15 20 25 30 Time (min)

Figure 3‎ .12 Assessment of H2O2 consumption by breast milk. Peroxide showed a gradual decrease in concentration due to endogenous breast milk lactoperoxidase and catalase (blue open circles). Addition of sodium azide (400 µM) to the breast milk inhibited H2O2 consumption (purple open diamonds). The negative control (green triangles) was H2O2 added to Tris buffer, which showed a stable concentration throughout the experiment, (n=3).

90

3.6 DISCUSSION

XO is widely distributed in nature [43, 44]. In humans XO tissue distribution is less widespread, with the highest activity occurring in small intestine and liver [45], but only low levels of XO RNA transcripts occur in human heart, brain, lung and kidney [46]. However, the high activity of XO in the milk of mammals, where it is associated with milk fat globule membrane (MFGM) [24, 47, 178], suggests that it has a special biological significance in this fluid [67]. This suggestion has been supported by the discovery that lactating female mice with xanthine oxidoreductase gene knockout exhibited mammary epithelium collapse and MFGM synthesis abnormalities, supporting the important role of XO in MFGM envelopment and secretion [179].

3.6.1 XO ACTIVITY AND KINETIC PARAMETERS The combination of peroxidase with Ampliflu™ Red assay reagent offers a sensitive probe for measuring oxidative enzymes that generate H2O2, such as XO [180]. In this chapter, the fluorimetric assay of human milk XO activity was validated and its enzymatic kinetics were determined in fresh human breast milk using the peroxidase-based assay system using hypoxanthine as a substrate. Determination of XO activity in 24 fresh breast milk produced a range of 1.8 to 18 U/L using hypoxanthine as a substrate, which was slightly higher than published values of 0.16 to 9.2 U/L [181] using radio-isotopic xanthine, and 0.01 to 8.3 U/L [123] using radio-isotopic hypoxanthine.

This 10-fold range may be attributed to at least 3 factors; (1) the postpartum time or gestational age of the milk samples, i.e. were some 'colostrum' samples and other samples from a much later postpartum age? (2) loss of activity due to poor sample handling; (3) genetic differences in XO activity among the mothers.

With regard to point (1) above, the milk samples were collected from the mothers 1 to 5 days postpartum, with colostrum considered as the major first fraction of the breast milk. The analysis of human milk in a previous study reported that XO activity greatly increased in the first 10 to 15 days after birth, but then fell to basal levels soon after [123], so the samples examined here fell into the period of increasing activity (i.e. 1-5 days). There was only a mild correlation between postpartum age and XO activity (and no correlation for gestational age) observed for the milk presented here, and this was considered evidence for a lack of a strong contribution to XO activity arising from either postpartum or gestational age of the samples.

With regard to point (2) above, as explained in the Methods the milk samples were immediately put on ice after being expressed, then delivered directly to the laboratory, subdivided into portions and 91 frozen at -80oC until assay. It is unlikely therefore that there was a large range of activity loss due to sample handling. It should be remembered that milk XO is localised within the milk fat globule membrane (MFGM), so that the substrates (hypoxanthine and xanthine) must diffuse into the

MFGM (no transporters are described) then the product (H2O2) must diffuse out. However, most publications of milk XO kinetics have studied the enzyme outside its physiological environment. In the results presented here, XO was assayed in situ, i.e. within the MFGM.

With regard to point (3), although complete XO deficiency is considered rare [182], in vivo studies of metabolism (which utilises liver XO) have indicated a wide range of activity in humans [183], although there are confounding factors such as cigarette smoking, drug therapies or liver disease [184]. Thus, the range of milk XO activities presently observed may arise from genetic variation, with some minor contribution due to postpartum age.

Kinetic studies of the in situ XO (i.e. still embedded into the milk fat globule membrane “MFGM”) found that the Km was 12.8 ± 2.1 µM, which is slightly higher than a published value for the purified enzyme (Km=8 µM) [185] but lower than that for purified enzyme (Km=18 µM) [178] using xanthine as the substrate.

The dose-dependent inhibition of H2O2 formation in breast milk treated with oxypurinol, was consistent with the action of oxypurinol as a specific inhibitor of XO [186], and confirmed that XO was responsible for the H2O2 production observed for the peroxidase assay. Within cells most xanthine oxidoreductase is in the form of xanthine dehydrogenase (XDH), however extracellular xanthine oxidoreductase is thought to exist solely as XO, e.g. in plasma XDH is transformed to XO by serum proteases, in other tissues the transformation is thought to be the result of oxidation of critical amino acids in the enzyme. Milk contains high amounts of proteases [187, 188], and it has been suggested that these may be responsible for XO as the dominant form in milk [189].

3.6.2 XO ACTIVITY IN INFANT FORMULA AND PASTEURISED MILK XO activity was undetectable in preterm and term infant formulas, indicating that the preparation, processing and manufacturing of milk formula (from bovine milk) leads to complete loss of the enzyme activity. Modern 'powdered milk' used in formula milk is spray-dried, which involves spraying the milk into a high temperature air stream; this appears to inactivate the XO component, which means that formula-fed babies do not experience oral hydrogen peroxide generation nor the lactoperoxidase system. It is not clear whether this is beneficial to the baby or not. This was the subject for further hypothesis development and for study in this thesis.

92

Similarly, XO activity was not detected in pasteurised breast milk. Pasteurised human milk was developed specifically to provide mothers who were unable to breast feed with milk for their newborn babies, which was considered more 'natural' and possibly less allergenic than bovine milk. However, the status of XO activity in this product was previously unknown. The findings presented here are consistent with those reporting significant activity loss after breast milk thermal pasteurisation for essential enzymes/proteins such as lysozyme [190, 191], lactoperoxidase, secretory IgA and lactoferrin [190]. It has been suggested that pasteurisation may not be ideal for human milk, and this practice remains a controversial issue in neonatal intensive care units worldwide [192]. Several studies have shown that human milk thermal pasteurisation induced a significant loss of bactericidal capacity against E. coli [193], and both E. coli and S. aureus [192]. It was also confirmed here that fresh commercially pasteurised bovine milk lacked XO activity. The arguments for and against regulatory pasteurisation of bovine milk have been ongoing and vociferous, and are beyond the scope of this project.

3.6.3 ENDOGENOUS BREAST MILK HYDROGEN PEROXIDE

The endogenous H2O2 concentration in 22 of the 24 fresh breast milk samples was 27.3 ± 12.2 µM, which was highly consistent with published data in which the mean concentration of H2O2 in a first week sample was 25 µM [194]. The presence of peroxide in bovine milk has been proposed to be caused by the co-secretion of XO substrates xanthine and hypoxanthine into the mammary gland lumen during continuous milk production [79, 189]. In bovine milk, H2O2 has been calculated to be produced at a rate of 260-360 µmol/h from secreted hypoxanthine, or 130-180 µmol/h from secreted xanthine [79], but there are no reports of endogenous H2O2 concentrations in bovine milk (or milk other than human). Presumably the same mechanism operates in all mammals including humans. To test for the presence of xanthine and hypoxanthine in fresh breast milk, 10 freshly expressed breast milk samples were immediately acid-denatured and subjected to LC-MS/MS analysis. However, these contained no detectable xanthine or hypoxanthine. The absence of these substrates in fresh breast milk may be due to their short half-lives in the mammary glands because of the high activity of XO.

The continuous generation of H2O2 within the mammary gland lumen has been proposed to have a bactericidal activity which may protect the breast against mastitis which is frequently caused by S. aureus [140, 195] (see Chapter 4). The concentration of peroxide in breast milk has been reported to peak in the first few weeks after birth, then declines to about 9 µM by the fourth week of infant life. This decline in peroxide concentration with postnatal age [194] is consistent with the report

93 that XO decreases in a similar pattern [123]. However, the data presented here do not support any association between endogenous breast milk H2O2 and XO activity.

In addition to direct inhibition of bacterial growth, the H2O2 generated in breast milk is also thought to serve as an electron acceptor for the lactoperoxidase reaction; lactoperoxidase is abundantly expressed in human milk [196] and converts thiocyanate (SCN-), as a physiological substrate in the presence of peroxide to the powerful bactericidal radical hypothiocyanite (OSCN-). This complex system may provide mammary glands with continuous passive protection against invading bacteria. Thus, as well as its classical role in purine catabolism, XO can also be considered a component of innate immunity in mammals [59, 71].

Interestingly, it has been found that H2O2 and other ROS and RNS play important roles as small- molecule second messengers, similar to nitric oxide. Pan et al (2011) [197] suggested that H2O2 at low levels, in particular around 20 µM, stimulates cell viability and facilitates adhesion, migration, and wound healing in cornea cells or other tissues. Furthermore, in activated T cells the superoxide anion or low micromolar concentrations of H2O2 increased the production of the T-cell growth factor interleukin- 2, an immunologically important T-cell protein [198]. Other studies found that

H2O2 has a role in inducing the expression of the haem oxygenase (HO-1) gene [199], and activation of the transcription factor nuclear factor kB (NF-B) [200]. In coordination with the role of H2O2 in cell signalling, nitric oxide (NO) has been found to have a regulatory role in the control of smooth muscle relaxation [201] and in the inhibition of platelet adhesion [202]. This suggests that milk H2O2 detected in this study may not only contribute to the innate immunity of neonates, but also may provide the appropriate concentrations for rapid cell signalling and growth.

3.6.4 GENERATION OF HYDROGEN PEROXIDE IN NEONATAL MOUTH In this chapter, it was demonstrated that mixing neonatal saliva (which contains high levels of xanthine and hypoxanthine, as per Chapter 2), with breast milk led to stimulation of H2O2 production. To maintain the reaction within the physiological pH range, both XO assays and peroxide stimulation by saliva were experimentally conducted at pH 7.5, which was within the normal pH range of breast milk (pH 6.4-7.6) and paediatric saliva (pH 6.4-8.24) [149]. It can be speculated that a similar mechanism occurs during breast feeding, especially as the median levels of xanthine (19 µM) and hypoxanthine (27 µM) in neonatal saliva are higher than the hypoxanthine Km (~13 µM) of breast milk XO.

94

The saliva sample used for the mixing experiment contained 70 µM hypoxanthine and 30 µM xanthine, which were then diluted and added to an equal volume of milk. Given that 1 mole hypoxanthine produces 2 moles of peroxide and 1 mole of xanthine produces 1 mole of peroxide in two-fold diluted saliva, the theoretical formation of H2O2 should be 70 µM peroxide from hypoxanthine and 15 µM peroxide from xanthine, yielding a total peroxide concentration of 85 µM.

However, the estimated H2O2 production was only 40 µM. The reduced amount of free peroxide generated in vitro may be the result of competing consumption of the H2O2 by catalase and/or lactoperoxidase, both of which are expressed abundantly in breast milk [203, 204]. Evidence for this was provided in the present study by inhibiting milk peroxide breakdown with azide which targets haem enzymes such as catalase and peroxidase.

The major enzymatic H2O2 consumption in normal human tracheal secretions is due to lactoperoxidase [205]. However, as fresh breast milk was found to contain concentrations of approximately 30 µM of peroxide, the additional amount generated by mixing with saliva was well within the levels reported to inhibit S. aureus growth (see Chapter 4). Even though some studies have suggested that H2O2 is sourced to the lactoperoxidase system by oral bacteria and/or by salivary leucocytes, the mechanism of the XO-peroxide regeneration in the neonatal oral cavity that we have identified could also serve as a rapid source of peroxide for the lactoperoxidase system similar to the mechanism discussed for the mammary gland, as the lactoperoxidase enzyme is also expressed in infant saliva [204, 206, 207].

The persistence of peroxide in the mixed breast milk/ neonatal saliva mixtures for upwards of 1 hour suggested that the system might also play a role in bacterial inhibition during its passage through the neonatal gut. It has been shown in bovine calves that the lactoperoxidase system is effective in preventing ‘scour’ (diarrhoea) [208, 209]. Importantly, the lack of breast milk feeding combined with the absence of saliva production due to tube feeding of early preterm babies may contribute to the high rate of necrotising enterocolitis (NEC) experienced by babies in neonatal intensive care units. In addition, the inhibition of this system by pasteurisation of human milk and in pre-term milk formula milks may be detrimental to the health of newborns by changing the selective processes that act on oral (and hence gut) microbiota. This theme was examined in Chapter 4.

95

Chapter 4 ‎

An innate immunity mechanism:

Regulation of the neonatal microbiome by breast milk and saliva

96

4.1 INTRODUCTION

4.1.1 THE ROLE AND REGULATION OF COMMENSAL BACTERIA An important outcome of studying the human microbiome, which has been greatly facilitated in recent years with the advent of molecular genetic techniques that allow identification of non- cultivable species, is an increased understanding of the complexity of the microbiota: this has provided new insights into human biology, as well as demonstrating the role of commensal (versus pathogenic) bacteria in human health [210].

The largest microbial community of the human microbiome is located in the gastrointestinal tract (GIT), especially the large intestine [211, 212]. The commensal microbiota has a number of crucial roles: it forms part of the intestinal barrier, which protects the host from invasive pathogens [213]; it modulates host immune responses and promotes health [214, 215]; it metabolises nutrients within the host GIT, contributing to the digestive process and providing essential nutrients such as vitamins, chelates, and short-chain fatty acids [216, 217]. Microbial imbalance in the GIT has been linked to diseases in both human adults and neonates, including colon cancer [218, 219], inflammatory bowel disease [220, 221], and necrotising enterocolitis (NEC) [222, 223]. A healthy mucosal microbiota is thus crucial to the survival of the human host and it has been suggested as a target for therapeutic intervention to prevent diseases including allergic diseases and inflammatory bowel disease [224-226].

Compared to adults, the infant GIT microbiota is more variable in its composition and less stable over time. During the first two years of life, the infant GIT microbiota develops to become more densely colonised, with a population of GIT microorganisms that are generally similar to those found in the adults [227]. Although it is widely accepted that the GIT microbiota is established after birth (colonisation occurring from the mouth downwards into the GIT), there is evidence that bacterial DNA is present within the human placenta and meconium [228, 229], suggesting that some bacteria may colonise neonates before delivery. In the first few hours of life, the mother’s vaginal and faecal microbiota are usually the most important source of neonatal oral and GIT bacterial colonisers [230, 231], however, it has been shown that the mode of delivery can influence the numbers of bacteria and the types of bacteria present within the GIT microbiota [111].

During a Caesarean delivery the mouth of the neonate is not exposed to the mother’s vaginal and intestinal microbiota, and this results in less diversity within the neonates’ GIT microbiota when compared to neonates that are delivered per vagina. Importantly, a low intestinal microbiota 97 biodiversity in early infancy appears to exacerbate the development of allergic disease later in life [232, 233]. For Caesarean-delivered neonates, GIT colonization by the genera Bifidobacterium and Bacteroidetes is also delayed [234, 235]. Bifidobacterium has been reported as the most predominant genus in the infant GIT [236]. However, interestingly, the numbers of Bifidobacterium within the GIT of neonates are higher in breast-fed infants compared to formula-fed neonates [237], in which the human milk itself appears to be the source of this commensal flora [109].

4.1.2 EFFECTS OF FORMULA VERSUS BREAST FEEDING ON NEONATAL GUT MICROBIOTA While formula feeding has been shown to induce the most rapid weight gain in infants (which has been linked to subsequent obesity), it has also been demonstrated that breast milk feeding is clearly more healthy for both term and premature infants with regard to resistance to infections [238, 239]. Breast milk has a lower protein and calorie content than formula milk, but it is rich in immunological factors including maternal immunoglobulins, enzymes and growth factors. There is accumulating evidence that in neonates the immune defence mechanisms that normally maintain healthy and balanced GIT microbial communities may be inadequate, or at least not functioning at levels seen in adults [240]. During the early postnatal period the neonatal GIT is thus susceptible to abnormal bacterial colonisation [241]. It has also been shown that formula-fed babies show an increased incidence of NEC, the most common and lethal GIT emergency in preterm infants in neonatal intensive care units [242]. In contrast, preterm babies who are fed breast milk are significantly less likely to develop NEC than formula-fed preterm babies [243, 244]. This evidence supports the concept that early infants rely on 'innate' immune mechanisms to regulate their GIT microbiota until the development of the more mature cellular and antibody-based immune system present in adults [245, 246].

4.1.3 OXIDATIVE MECHANISMS FOR ORAL MICROBIAL REGULATION IN NEONATES Xanthine oxidase (XO) has been proposed to trigger several antibacterial metabolic systems in the presence of its substrates (hypoxanthine / xanthine), producing microbiocidal reactive oxygen species (ROS): superoxide and hydrogen peroxide (H2O2). The latter subsequently acts as a substrate for the lactoperoxidase system, which in the presence of thiocyanate (CNS-) can generate the antibacterial reactive nitrogen species (RNS), hypothiocyanite (OSCN-) [189]. In addition to lactoperoxidase system activation, XO has been shown to catalyse the anaerobic reduction of - inorganic nitrite (NO2 ) to nitric oxide [72, 73]. Superoxide generated by XO in the presence of molecular oxygen also reacts rapidly with nitric oxide to yield the RNS, peroxynitrite (ONOO-),

98 which is also a powerful antibacterial agent [24, 73, 247, 248]. As discussed in Chapter 3, XO is highly active in human breast milk where it may function to provide an XO-induced “cascade” of oxidative molecular species which may regulate neonatal oral microbiota during breast-feeding.

But does the XO system remain active within the neonatal stomach environment? Stevens et al. (2000) have shown that during breast-feeding, XO (and other milk proteins and fats) can survive passage through the infant stomach. Superoxide and H2O2 generation is markedly stimulated above pH 7 in the presence of oxygen, but the neonatal stomach is not highly acidic (pH 4-6), with an oxygen tension less than 5%, and these conditions are sufficient for XO to generate NO and consequently ONOO- [71]. As a result, XO is considered to be a component of the innate immune system in mammals [59].

4.1.4 OXIDATIVE MECHANISMS FOR MAMMARY GLANDS INNATE IMMUNITY

Fresh milk contains a surprisingly high amount of H2O2. In bovine milk XO activity is about 10- fold higher than in human milk and it has been predicted that the H2O2 is accordingly higher [79].

In Chapter 3 it was shown that the mean H2O2 in human milk was approximately 30 µM, although this may theoretically be boosted by another 80-100 µM by the XO metabolism of the hypoxanthine and xanthine present in neonatal saliva during breastfeeding. However, the effectiveness of the micromolar concentration range of H2O2 on bacterial growth has never been directly assessed.

It has been proposed that ROS, particularly H2O2, are active in protecting mammals against mastitis, which is an inflammation of breast tissue caused by several species of microorganisms [79,

189]. In particular, it has been suggested that H2O2 generated by XO in mammary ducts may provide a defence mechanism against Staphylococcus aureus [137, 139, 140, 195]. Mastitis can occur at any stage of lactation but the rate of occurrence is highest in the first few weeks postpartum, with an incidence of 2-33% [136, 139].

Allied to this concept, the XO-catalysed superoxide generating system has also been proposed to have an antibacterial role in mammary glands. These XO-induced cascades are enhanced when additional xanthine & hypoxanthine are added to breast milk. It was discovered in Chapter 2 that elevated concentrations of xanthine and hypoxanthine, as well as several important nucleotide metabolites, are present in neonatal saliva, compared to adult. Furthermore, in Chapter 3 it was demonstrated that by adding neonatal saliva to breast milk, XO was activated to generate ROS which may fuel a unique innate antibacterial mechanism in the neonatal mouth and then the GIT at

99 a time when other immune mechanisms are not fully developed. These observations led to the formulation of Specific Hypothesis 1, below.

4.1.5 GROWTH STIMULATION MECHANISMS FOR MICROBIAL REGULATION A further hypothesis developed for this project, was that the high levels of nucleotide metabolites present in neonatal saliva may act as positive growth factors to selectively stimulate some species of commensal bacteria (Specific Hypothesis 2, below).

Central to this concept of stimulating specific bacteria by providing growth factors, is the division of microorganisms into those that can synthesise their own biochemical compounds needed for growth ('prototrophs'), and those that are unable to synthesise a particular organic compound required for growth ('auxotrophs'), as defined by the International Union of Pure and Applied Chemistry (IUPAC, Zurich Switzerland).

Prototrophic bacteria can thus synthesise de novo their own nucleotides (which are essential for energy and DNA synthesis – see Chapter 1). In contrast, bacteria that are auxotrophic with respect to nucleotide metabolites are poor at de novo synthesis of nucleotides but can effectively utilise external sources of purines and pyrimidines to enhance their growth, by using nucleotide salvage pathways to convert bases or nucleosides derived from their host to their corresponding nucleotides [18]. However, by salvaging nucleotide metabolites provided by the host, auxotrophic bacteria are able to synthesise nucleotides by less energy-demanding and more efficient biochemical pathways, and this presumably may give them a growth advantage over other bacteria that do not use salvage pathways for nucleotide synthesis. This may provide a positive selection mechanism for oral and perhaps GIT microbiota.

100

4.2 SPECIFIC HYPOTHESES AND RATIONALE

Hypothesis 1: That the H2O2 generated in a neonatal mouth, when breast milk XO mixes with its salivary substrates xanthine and hypoxanthine, inhibits potentially pathogenic bacteria within the mouth and the GIT.

Hypothesis 2: It is well known that some bacteria rely on external sources of purines and pyrimidines for their metabolism (see Section 1.2.7). It is proposed that bases and nucleosides present in neonatal saliva stimulate and select for commensal microbiota.

Hypothesis 3: That the oral microbiota of 4-8 week old infants varies depending on the mode of feeding (i.e. breast-fed versus formula-fed), at least in part as a result of these selective metabolites in Hypotheses 1 and 2.

Rationale: The effect of breastfeeding versus formula feeding has been assessed for infant faecal bacteria but it has never been determined for oral bacteria. Also, the foundation of the above hypotheses rests on the supposition that there will be a difference between oral microbial populations due to breastfeeding (i.e. XO-positive milk) versus formula feeding (XO- negative milk). Thus, it is essential to determine whether there is a measurable in vivo effect.

4.3 SPECIFIC AIMS

In relation to the previous studies (see Chapters 2 and 3), the aims of this study were to:

1. Investigate in vitro the effect of micromolar H2O2 on bacterial growth

2. Investigate in vitro the effects of adding breast milk (containing XO) to simulated neonatal saliva (with xanthine and hypoxanthine, and other nucleosides and bases) on bacterial growth

3. Compare and contrast the in vivo differences in oral microbiota within neonatal mouth from babies who have been breast-fed (i.e. exposed to the milk XO-generated superoxide system) to the oral microbiota of babies who were exclusively formula-fed, using PCR amplification of the bacterial 16S rRNA and deep sequencing

101

4.4 MATERIALS AND METHODS

4.4.1 NUCLEOTIDE METABOLITE PREPARATIONS Stock solutions (10 mM) of the nucleotide metabolites, inosine, guanosine, adenosine, uridine and uracil, and of hypoxanthine and xanthine (XO substrates), for reference standards and cell culture studies (Sigma-Aldrich Pty. Ltd, Castle Hill, NSW, Australia), were prepared in sterile distilled water except xanthine which was first dissolved in 20 mM NaOH. From these stock solutions 1 mM standards were prepared in sterile distilled water. The UV absorbance maxima for each purine and pyrimidine was measured using a spectrophotometer (Cary 50 Bio-visible, Wilson Drive, USA) and accurate concentrations were calculated for each compound from the UV molar extinction coefficients. Solutions were then sterilised by filtration through sterile 0.2 µm filters (Acrodisc® Syringe Filters, PALL, Cheltenham, VIC, Australia) and stored in aliquots at -80oC.

4.4.2 HYDROGEN PEROXIDE STANDARDIZATION Hydrogen peroxide (30% v/v, ~9 M) (Merck Pty Ltd, Kilsyth, VIC, Australia) was standardised by diluting to 9 mM in 100 mM Tris-HCl buffer, pH 7.5 then the peroxide concentration was calculated by its absorbance at 240 nm, against a Tris-HCl buffer blank, assuming a molar extinction of 43.6 mol-1L-1cm-1.

4.4.3 PREPARATION OF BIOLOGICAL SAMPLES Breast milk: Ten breast milk samples (1 mL each) were pooled to produce a stock milk pool, which was then divided into 50 µL portions and stored at -80oC.

'Supplemented saliva': Due to the difficulty in obtaining sufficient volumes from infants (mean volume of the samples was approx. 100 µL, see Chapter 2), simulated neonatal saliva was prepared for this study by supplementing heat-inactivated and sterilised adult saliva.

'Supplemented saliva' was prepared by pooling 50 adult salivary samples to produce 70 mL of pooled saliva. This was mixed thoroughly then heat inactivated at 56oC in a water bath for 35 min, with gentle mixing every 5 min to ensure uniform heating. The aim of heat inactivation was to remove antibody activity that is present in adult saliva, but which is negligible in neonatal saliva [161]. The 'supplemented saliva' was then sterilised by vacuum filtration through a sterile 0.45 µm filter (Grace Discovery Sciences, Rowville, VIC, Australia) with a sterile filtration apparatus.

102

'Supplemented saliva' was divided into four experimental aliquots as following:

A. The first saliva aliquot ('Control') contained no additional nucleotide metabolites. However, there were naturally low concentrations of purine and pyrimidine metabolites present in adult saliva; an assay of these metabolites by LC-MS/MS (see Chapter 2 for details) provided values that were close to the median concentrations shown in Table 2.8. B. The second saliva aliquot ('Full supplement, except X and HX') was supplemented with the major purine and pyrimidine nucleotide metabolites to produce the median concentrations found in neonatal saliva; inosine, adenosine, guanosine, uracil, and uridine (11, 12, 7, 5.3, 12 µM respectively), but excluding xanthine and hypoxanthine. C. The third aliquot ('Full supplement') was supplemented to produce the median concentrations of the major purine and pyrimidine metabolites found in neonatal saliva as above, plus xanthine (20 µM) and hypoxanthine (27 µM). D. The fourth saliva aliquot ('Full supplement + oxypurinol') was supplemented with the same concentrations of the major purine and pyrimidine metabolites as in 'C', as well as 100 µM oxypurinol, an inhibitor of XO.

See Table 4.1 for a summary of the saliva supplements. Each of these 'simulated neonatal saliva' aliquots was divided into smaller volumes and stored at -80oC. These saliva samples were also analysed using HPLC-PDA (Agilent, USA) to confirm the presence of the purines and pyrimidines at the correct concentrations (see Chapter 2 for HPLC-PDA method).

Table 4‎ .1 Saliva experimental solutions. Four experimental saliva formulations were used to evaluate in vitro the bactericidal activity of breast milk and simulated neonatal saliva. All saliva samples were collected from adults, heat inactivated, sterilised by filtration, and supplemented with bases and nucleosides.

Saliva Supplements Control No supplements: control saliva only Full supplement except X, HX Major purines and pyrimidines except XO substrates (xanthine and hypoxanthine)

Full supplement Major purines and pyrimidines present in neonatal saliva including XO substrates

Full supplement + oxypurinol Major purines and pyrimidines present in neonatal saliva plus oxypurinol (100 µM)

103

4.4.4 BACTERIAL PREPARATIONS Bacterial strains were obtained from the culture collection at the Queensland University of Technology, Faculty of Health, Brisbane, QLD, Australia. These included Escherichia coli (ATCC 2592), Staphylococcus aureus (ATCC 29213), Salmonella spp. (ATCC 1311), Lactobacillus plantarum (UQM 297). Bacteria were stored in liquid nitrogen then cultured by inoculating onto either nutrient agar, horse blood agar (HBA), or de Man, Rogosa and Sharpe (MRS) agar (Oxoid Pty Ltd, Adelaide, SA, Australia) and incubated aerobically at 37o C either for 24 h or 48 h. Bacteria then were subcultured into 25 mL of either MRS broth for Lactobacillus species, or nutrient broth for the other species. The suspensions were vortexed for 15 s and incubated aerobically for 24 h at 37oC in a shaking incubator. After overnight incubation, the broth incubations containing bacteria were transferred into sterile centrifuge tubes, and the cells were harvested by centrifugation at 4000 x g for 15 min at 4oC. The supernatants were then discarded and the pellets were resuspended in 3 mL nutrient or MRS broth containing an equal volume of 20% glycerol (w/v). Each suspension was then vortexed for 15 seconds and divided into 100 µL aliquots in sterile Cryovial tubes (Sarstedt Australia Pty Ltd, Mawson Lakes, SA, Australia) and stored at -80oC for future experiments.

The number of colony forming units (CFU) per mL in a thawed aliquot of the stock for each bacterial species was determined. For each bacterial species the thawed stock was serially diluted ten-fold (10-1 to 10-5) in phosphate buffered saline (PBS) and then 3 x 10 µL drops from each dilution were inoculated onto nutrient agar or MRS agar plate as appropriate. The plates then were incubated at 37oC for either 24 h or 48 h. The number of CFUs in 3 x 10 µL drops was counted visually on one plate (the dilution with well-separated CFU in numbers that could easily be counted) and the number of CFUs per mL was calculated for each species.

4.4.5 DETERMINATION OF PURINE AND PYRIMIDINE METABOLITES IN BACTERIAL MEDIA USING LC-MS To determine the concentrations of purine and pyrimidine metabolites in bacterial culture media, aliquots of sterile Nutrient and MRS broths were ultra-filtered by microfuging using 0.5 mL, 50 kDa membrane cut-off filtration units (Amicon Ultra, Merck Pty. Ltd., Kilsyth, VIC, Australia) at 9500 x g, for 10 min at 20°C, then 10 µL of each medium was subjected to LC-MS/MS analysis (see Chapter 2 for details).

104

4.4.6 IN VITRO ORAL MICROBIOTA STUDIES

4.4.6.1 Hydrogen peroxide titration against bacteria Bacterial growth inhibition assays were conducted using flat bottom 96-microtiter plates (Thermo Scientific, Scoresby, VIC, Australia). Bacterial stock being tested (prepared as described above) was diluted in nutrient broth and then kept on ice. The number of cells was adjusted to a final concentration of 200 CFU/mL. Serial two-fold dilutions of H2O2 in 50 µL nutrient broth were prepared within the rows of the microtiter plate to give final H2O2 concentrations of 400, 200, 100, 50, 25, 12, 6, 3, 1.5, 0.75, 0 µM (11 wells). 50 µL of the diluted stock bacteria was added to each of the wells of the broth-peroxide dilution series to give a total volume of 100 µL. The final well (12th) in each row was a negative control and contained H2O2 and nutrient broth only. Each bacterial species was tested in triplicate in three H2O2 dilution series. The microtiter plate was sealed with a sterile adhesive microtiter sealer (MP Biomedicals, Seven Hills, NSW, Australia) and incubated at 37oC in a shaking incubator (John Morris Scientific Pty Ltd, Chatswood, NSW, Australia) for either 24 h or 48 h. Bacterial growth was determined turbidometrically using a microtiter reader (xMark, Bio-Rad Laboratories Pty. Ltd., Gladesville, NSW, Australia) by measuring absorbance at 600 nm.

4.4.6.2 Effects on bacterial growth of saliva and breast milk Stock bacterial species (S. aureus, E.coli, L. plantarum, Salmonella spp.) were serially diluted in PBS and adjusted to 200 CFU/mL for all experiments. Breast milk was diluted 1:6 in sterile 100 mmol/L Tris-HCl buffer (pH 7.5). In a 96-microtiter plate, 50 µL of each simulated neonatal saliva formulation (see Table 4.1) was pipetted into triplicate wells. To ensure that the metabolic reactions started simultaneously, breast milk (25 µL) was inoculated with bacteria (25 µL) in separate plate wells. Using a multichannel pipette, this mixture then was added to saliva to give a final concentration of 200 CFU/mL bacteria in a total volume of 100 µL. The plates were also sealed with a sterile microtiter adhesive sealer and incubated overnight at 37oC in a mechanical agitator (200 rpm). After the incubation, viable colony counts were determined for each mixture in triplicate as described above.

To evaluate the dose dependent effects of xanthine and hypoxanthine concentrations on bacterial growth, serial dilutions of xanthine and hypoxanthine were generated in simulated neonatal saliva supplements. Bacteria and breast milk were then mixed with the supplemented saliva in triplicate as detailed above. The final concentrations of xanthine and hypoxanthine were 0, 6, 25, 50 µM. Bacteria growth was assessed by viable cell counts as above. See Figure 4.1 for a schematic summary of the experiment.

105

4.4.7 IN VIVO ORAL MICROBIOTA STUDIES OF BREAST-FED VERSUS FORMULA-FED NEONATES

4.4.7.1 Subjects and specimen collection This study was approved by the Human Research Ethics Committees’ (see Appendices). Parents of the eligible infants were interviewed after delivery by a Neonatal Research Nurse at the Mater Mothers’ Hospital, South Brisbane, QLD, Australia and they gave consent for their infants’ to be recruited for this study. Breast-fed infants on antibiotics or infants’ whose mothers were on antibiotics before or at the time of collection were excluded from the study. Samples of buccal cells including bacteria were collected from healthy, full-term, vaginally delivered, Caucasian infants aged 4-8 weeks who had not been fed for at least 30 min. Infants’ buccal swab specimens were collected at the participants’ homes from breast-fed (n=26) and formula-fed (n=12) infants. The buccal cavity of each infant was swabbed gently for approximately 5 min using sterile swabs (Copan flocked swabs, Italy); no infant experienced any obvious discomfort and sterile gloves were worn during the collection. Three swabs were collected from each infant and placed in sterile 1.5 mL Eppendorf tubes and frozen on dry ice immediately. Samples were immediately delivered to the laboratory and stored at -80oC for further analysis. See Table 4.3 for details of the infants’ characteristics.

4.4.7.2 DNA extraction and purification For total genomic DNA extraction, buccal swabs were mixed with 100 µL of 4 g/L lysozyme (20 mM Tris-HCl, pH 8.0; 2 mM EDTA; 1.2% Triton) (Sigma-Aldrich, Castle Hill, NSW, Australia) and incubated in a heating block for 35 min at 37oC. Lysozyme hydrolyzes glycosidic bonds within peptidoglycan (the bacterial ), increasing the bacterial DNA yield. After incubation, 100 µL of 100 µg/mL Proteinase K (QIAGEN Pty Ltd, Chadstone Centre, VIC, Australia) solution was added and this was further incubated for 30 min at 37oC. Finally, 400 µL of lysis buffer was added to the mixture, which then was incubated at 56°C for a further 10 mins. The total genomic DNA was subsequently purified using the QIAamp DNA Mini Kit (QIAGEN Pty Ltd, Chadstone Centre, VIC, Australia) following the manufacturer’s instructions for Spin-Column Protocol for "DNA Purification from Buccal Swabs". The genomic DNA was eluted in 100 µL of PCR grade DNAse/RNAse-free distilled water (GIBCO UltraPureTM, Invitrogen, Mulgrave, VIC, Australia). Concentrations and DNA quality were assessed using a NanoDrop ND-1000 spectrophotometer (Thermo Scientific, Scoresby, VIC, Australia) and agarose gel electrophoresis. A 2% w/v agarose gel was prepared using the following ingredients: 2 g agarose powder (BIORAD) and 100 mL of 0.5x tris-borate EDTA (TBE) Buffer (Sigma-Aldrich, Castle Hill, NSW, Australia), the mixture was then heated using a microwave to dissolve the agarose. As the gel cooled, 2 μL of ethidium bromide (Sigma-Aldrich, Castle Hill, NSW, Australia) was added and gel was poured into a mould. After the 106 gel had set, DNA samples were mixed with gel loading buffer, and added to the gel. The gel was run for 60 min at 80 V, then photographed. The intensity of staining of the DNA amplicons then was compared to the DNA molecular weight marker VIII (Roche, Castle Hill, NSW, Australia). The genomic DNA concentration was standardised for all samples to approximately 10 ng/µL before PCR amplification. These DNA samples then were sent to the Australian Genome Research Facility (University of Queensland, Brisbane, QLD, Australia) for bacterial 16S ribosomal RNA (rRNA) gene 'deep sequencing'.

4.4.7.3 PCR amplification of the bacterial 16S rRNA genes PCR was performed using a Forward 27F primer (AGAGTTTGATCMTGGCTCAG) and a Reverse 519R primer (GWATTACCGCGGCKGCTG) [249], with fusion primer sequences for both the forward and reverse directions for each amplicon. Each primer sequence was begun with the appropriate adapter sequence for Lib-L or Lib-A libraries followed by the key sequence of TCAG. This was then followed by the unique “barcode” sequences for a GS FLX MID for the forward and the reverse directions; the barcode sequences were used subsequently to identify and distinguish each of the samples. Table 4.2 summaries the design structure of the fusion primers, and the 16S rRNA target sequence primers [Adapter]-[Key]-[MID barcode]-[Target primers].

Table 4‎ .2 Fusion primers design (5' to 3') for 16S rRNA amplification. The fusion primer sequences contained [Adapter]-[Key]-[MID barcode]-[Target primers].

Forward fusion primer sequence Reverse fusion primer sequence Adaptor CCATCTCATCCCTGCGTGTCTCCGAC CCTATCCCCTGTGTGCCTTGGCAGTCTCAG Key TCAG TCAG

MID Barcode sample specific ------

Target primer* GWATTACCGCGGCKGCTG AGAGTTTGATCMTGGCTCAG

*Base redundancies were included within the target primers (W= A+T, K= T+G, M= A+C).

The PCR amplifications were conducted in 384 well plates using 8 µL total volume per reaction. Each PCR reaction contained: 2 µL pooled DNA templates, 4 µL master mix containing Ampli Taq Gold (AmpliTAQ GOLD 360, Applied Biosystems, USA), and 2.25 µM of primers. The PCR cycling parameters were: 95°C for 5 min; 35 cycles of 95°C for 30 sec, 60°C for 30 sec, and 72°C for 2 min; a final extension step at 72°C for 7 min; and then the reactions were held at 10°C. All amplicons were purified using a robotic liquid handler (Beckman Coulter 384 Biomek NX, CA, USA) and Agencourt AMPure (Beckman Coulter, CA, USA). 107

4.4.7.4 Sample QC post PCR purification The PCR amplicons were quantified using PicoGreen® (Invitrogen, Mulgrave, VIC, Australia) to determine the DNA concentration with a slight modification to the manufacturer's instructions. Briefly, the samples were diluted 1:100 using PCR grade water, then reduced equal volumes (4 µL) of standards, samples and PicoGreen were used for the measurements. All samples were processed in 384 well plates. Samples then were normalised to the same molarity then a qPCR reaction was performed to check each sample. The samples were re-normalised based on the qPCR copy number results. This double normalisation technique is effective at ensuring an even number of de- multiplexed reads post sequencing on the GSFLX.

4.4.7.5 16S rRNA gene sequencing ('deep sequencing') The pool of samples then was set up according to Roche’s standard Lib-L emPCR method for XLR70 chemistry on the GSFLX+ instrument, emPCR Method Manual – Lib-L MV GX FLX Titanium Series October 2009 (Rev. Jan 2010), Germany. Emulsions were setup at 2 x MV (medium volume), both at 0.35 cpb (copies per bead) to achieve recoveries of 18% and 20% (the recommended range is 5 to 20%).

Two million of such beads were sequenced in a half picotitre plate in Roche-454 GS FLX amplicon sequencing platform instrument (Roche, Castle Hill, NSW, Australia) according to Roche’s standard XLR70 protocol for the GSFLX+ instrument. Subsequently, the quantified libraries were amplified in micro-reactors through emulsion PCR (emPCR) followed by Streptavidin bead enrichment and emulsion breaking. The beads attached to amplified DNA fragments were denatured with 1N NaOH solution and annealed to a specific sequencing primer. All these steps and subsequent sequencing steps on Roche-454 GS FLX amplicon sequencing platform instrument were performed according to Roche-454 GS FLX amplicon sequencing manual protocol updated in October 2009 and revised by January 2010. Standard bioinformatics quality control (QC) involved reviewing internal QC metrics and de-multiplexing the results as per the barcodes included in the pool.

4.4.7.6 16S rRNA gene sequence analysis The composition and species diversity in high-throughput amplicon sequencing data was analysed using the Quantitative Insights Into Microbial Ecology (QIIME) software package version 1.7.0 (http://qiime.org/) [250]. A total of 589,945 reads of 250-550 bp in length were obtained from the infants’ oral swab samples. The sequence dataset was depleted of short reads (< 250bp) and then the reads from the forward primer were selected and assigned to sample identities (for breast-fed and formula-fed infants) based on their unique barcode sequences. Sequence accuracy was maintained

108 by setting the default Phred quality score on Q25 or above. The resultant 522,275 high quality reads were used to computationally describe (or define) the bacterial community diversity and richness within each sample, and the relative abundances taxonomic levels from Phyla to each operational taxonomic unit (OTU) using QIIME version 1.7.0. The oral microbiota of breast-fed and formula- fed infants were also compared.

The sequences were clustered into OTUs (6,101 in total) based on their similarity of sequences (97% identity cutoff) using the UCLUST clustering algorithm. For each OTU a representative sequence was selected (the most consistent sequence match for each OTU) for further analysis of bacterial community diversity and abundance. These representative OTU sequences were aligned and chimera checking was conducted against the 'greengenes' 16S rRNA reference core set as the template sequences using QIIME based Python Nearest Alignment Space Termination Tool (PyNAST) alignment. All unaligned and potential chimeric sequences were detected and removed from the aligned sequence dataset. For taxonomy based analysis, the Ribosomal Database Project (RDP) classifier [251] was used, with a confidence threshold of 85%. QIIME-based composition analysis was used to quantify the total number of reads per each OTU and their relative abundance at different taxonomic levels.

4.4.7.7 Statistical analysis Mean values were compared between two different groups using an unpaired, two-tailed t-test, and between more than two groups using one-way analysis of variance (ANOVA). Comparison of oral abundance of bacterial phyla in breast-fed versus formula-fed infants was performed using the Mann-Whitney test. A p-value of 0.05 was set as the level of significance. All analyses of bactericidal experiments were analysed using GraphPad Prism 5.

The study procedure is presented schematically in Figure 4.1.

109

Figure 4‎ .1 Schematic diagram summarises the overall bacterial experimental plan. The study investigated: (1) the in vitro bactericidal effect of breast milk and saliva; and (2) the in vivo oral biodiversity in breast-fed versus formula-fed infants. Bactericidal effects were studied in two experiments: (1a) the evaluation of the bactericidal effect of increasing hydrogen H2O2 concentrations and (1b) evaluation of the bacterial growth in four different simulated neonatal saliva aliquots, which were added to human breast milk. Abbreviations: Cont, control; FS (X&HX-), full supplement except xanthine and hypoxanthine; FS, full supplement; FS+oxypurinol, full supplement plus oxypurinol (XO inhibitor).

110

4.5 RESULTS

4.5.1 SUPPLEMENTED SALIVA PREPARATION The purine and pyrimidine contents of the four ‘supplemented saliva’ formulations (Section 4.4.3) were assayed by HPLC-PDA, and the results are shown in Figure 4.2. The HPLC analysis of the 'Control' simulated neonatal saliva (i.e. pooled and filtered, heat-inactivated adult saliva with no additional metabolites) found the concentrations of nucleosides and bases were below the limit of detection (Fig. 4.2 A). Analysis of the 'Full supplement except X, HX' simulated neonatal saliva confirmed the absence of hypoxanthine and xanthine (Fig. 4.2 B) and presence of the added nucleosides and bases at correct concentrations. Analysis of the 'Full supplement' simulated neonatal saliva (with all bases and nucleosides) showed the presence of uracil, hypoxanthine, xanthine, uridine, inosine, guanosine, and adenosine (Fig. 4.2 C) at correct concentrations. HPLC analysis of the fourth saliva formulation confirmed the full supplement with an additional peak for oxypurinol (not shown). The chromatograms also showed uric acid to be present in the pooled adult saliva; this end-product of purine catabolism is present also in similar concentrations in neonatal saliva (Section 2.5.4).

111

Figure 4‎ .2 Chromatograms showing UV-absorbing peaks at 268 nm present in “supplemented simulated neonatal saliva” preparations. An aliquot of each formulation (20 µL) was ultra-filtered to remove protein and then injected into HPLC-PDA (see Chapter 2 for method details). 268 nm was used for monitoring as this is a mean value for purine and pyrimidine peak absorption. (A) ‘Control’: no additional bases or nucleosides, (B) 'Full supplement except X, HX’: supplemented with bases and nucleosides except xanthine and hypoxanthine, (C) 'Full supplement’: supplemented with major bases and nucleosides including xanthine and hypoxanthine. Chromatograms also show the endogenous uric acid saliva peak. 112

4.5.2 DETERMINATION OF PURINE AND PYRIMIDINE METABOLITES IN BACTERIAL GROWTH MEDIA The concentrations of nucleotide metabolites in Nutrient and MRS broth were determined using LC-MS/MS are presented in Table 4.3 (see Chapter 2 for details). The concentrations ranged from as low as 1 µM for adenine in Nutrient Broth to 369 µM for adenosine in MRS. The method also detected high concentrations of creatinine: 1646 µM in Nutrient broth and 164 µM in MRS. All nucleotide metabolites were also detected in both media, some at high micromolar concentrations. Analysis of these commonly-used bacterial growth media has not been previously reported.

Table 4.3 Concentration of nucleotide metabolites in Nutrient and MRS broth (µM) as determined by LC-MS Nucleotide metabolites Nutrient Broth MRS Thymine 3 6 Creatinine 1646 164

Uracil 14 48

Hypoxanthine 121 31

Xanthine 36 54

Adenine 1 5

Adenosine 19 369

Deoxyuridine 4 1

Inosine 148 12

Guanosine 65 70

Thymidine 9 3

Uridine 87 183

113

4.5.3 IN VITRO ORAL MICROBIOTA STUDIES

4.5.3.1 Hydrogen peroxide titration against bacteria

Aim 1: Investigate in vitro the effect of micromolar H2O2 on bacterial growth

Increasing concentrations of H2O2 had an inhibitory effect on bacterial viability in Nutrient Broth, with remarkable differences observed between bacterial species. Viability was assessed by measurement of the turbidity. For example, high concentrations of H2O2 (≥200 µM) markedly reduced the viability of all bacteria tested (S. aureus, L. plantarum, Salmonella spp, and E.coli) (Fig 4.3). Of these four bacterial species, S.aureus, a Gram positive bacterium, was the most sensitive to

H2O2, where >25 µM was able to inhibit the growth. By contrast, the Gram negative bacterial species (Salmonella spp., and E.coli) were only inhibited by higher concentrations of H2O2:

Salmonella spp.>50 µM H2O2, and E.coli >200 µM H2O2. The Gram positive rod, L. plantarum was inhibited by >100 µM H2O2.

114

0.5 0.5 A 0.4 B 0.4

0.3 0.3

0.2 0.2

OD (600nm) OD OD (600nm) OD 0.1 0.1

0.0 0.0 0 10 20 30 40 50 60 200 400 0 10 20 30 40 50 60 200 400 H2O2 (M) H2O2 (M)

0.8 0.5

C D 0.4 0.6 0.3 0.4

0.2

OD (600nm) OD (600nm) OD 0.2 0.1

0.0 0.0 0 10 20 30 40 50 60 200 400 0 10 20 30 40 50 60 200 400

H2O2 (M) H2O2 (M)

Figure 4‎ .3 Effect of increasing H2O2 (0-400 µM) concentration on bacterial growth. Hydrogen peroxide was added to Nutrient Broth inoculated with bacteria and incubated for 24 h at 37oC, except for L. plantarum which was incubated for 48 h. The concentration of cells was determined by measuring turbidity at OD600. A) S. aureus, B) Salmonella spp., C) L. plantarum, and D) E.coli. Values represent the mean±SD of triplicate experiments.

115

4.5.3.2 Bacterial growth with saliva and breast milk

Aim 2: Investigate in vitro the effects of adding breast milk (containing XO) to simulated neonatal saliva (with xanthine and hypoxanthine, and other nucleosides and bases) on bacterial growth

The bactericidal activity of breast milk XO was evaluated by adding breast milk to the four different preparations of simulated neonatal saliva (Table 4.1) and these were then inoculated with bacteria. The growth of S. aureus did not benefit from the supplementation with nucleosides and bases (Full supplement except X, HX) when compared to the control (p=0.13). But the growth of S. aureus was significantly inhibited in the presence of xanthine and hypoxanthine (Full supplement), which cause

XO in the breast milk to produce H2O2 (p=0.016). The inactivation of XO by oxypurinol (Full supplement+oxypurinol) resulted in a significant increase in S. aureus growth (p=0.001) (Figure 4.4 A). These results are consistent with those observed when S. aureus was grown in the presence of

H2O2 (Fig. 4.3A).

Similarly, the growth of Salmonella spp. was not affected by supplementation of simulated saliva with nucleosides and bases in the media (Full supplement except X, HX) (p=0.62) (Fig. 4.4B). However, the growth of Salmonella spp. was significantly inhibited by the activity of breast milk XO in the presence of xanthine and hypoxanthine (Full supplement) (p=0.007) (Fig 4.4B), and the growth was significantly restored by the inactivation of XO by oxypurinol (p=0.013) (Fig 4.4 B).

This apparent sensitivity to H2O2 was in contrast to the direct H2O2 titration experiment, which showed that >50 µM H2O2 was needed to inhibit the growth of Salmonella spp (Fig. 4.3 B).

L. plantarum growth was noticeably stimulated in the presence of bases and nucleosides (Full supplement except X,HX) compared to the control (p=0.09) (Fig 4.4C). The growth of L. plantarum was noticeably inhibited by XO in the presence of its substrates (Full supplement) (p=0.08) (Fig 4.4C). However, the exclusion of xanthine and hypoxanthine (p=0.1) and inhibition of XO by oxypurinol (p=0.1) restored the bacterial growth. However, L. plantarum response to the above systems was not statistically significant. In comparison no significant or noticeable effect, either stimulation or inhibition, was observed when E.coli was introduced to these four different metabolism experiments (Fig 4.4 D).

To further define the effect of the metabolism of substrates (in neonatal saliva) by XO (within breast milk) on bacterial growth, simulated neonatal saliva was supplemented with serial dilutions of hypoxanthine and xanthine prior to the incubation with breast milk and bacteria. Enhanced growth inhibition was observed with increased XO substrate concentrations as illustrated in Figure 4.5. Xanthine and hypoxanthine inhibited, in a dose-dependent manner, the growth of S. aureus

116

(p=0.001), Salmonella spp. (p=0.003) and L. plantarum (p=0.001), where the increase in the XO substrates concentration was assumed to result in an increase of H2O2 generation. However, compared to the above species, E.coli was not affected by increasing H2O2 and the activity of the XO metabolic reaction system (p=0.3) (Figure 4.5 D), which was also consistent with the results of the previous experiment (Figure 4.4 D).

117

ns ns A 8 * ** B 8 ** *

6 6

4 4

Log CFU/mL Log Log CFU/mL Log 2 2

0 0

ns ns ns ns ns ns C 8 D 8

6 6

4 4

Log CFU/mL Log CFU/mL Log 2 2

0 0

Control Full supplement except X&HX Full supplement Full supplement + Oxypurinol

Figure 4‎ .4 The effect of breast milk plus simulated neonatal saliva (with and without supplements) on bacterial viability. Bacteria were incubated with breast milk and 3 “supplemented saliva aliquots” and a control (Table 4.1). The mixtures were incubated for 24 h at 37oC. Bacterial survival was assayed by agar plating and viable cells were measured as Log CFU/mL. A) S. aureus, B) Salmonella spp., C) L. plantarum, and D) E. coli . Values represent the mean±SD, from triplicate experiments. Symbols: ns, not significant (p>0.05); * p<0.05; ** p<0.01.

118

8 8 A B

6 6

4 p= 0.001 4 p= 0.003 Log CFU/mL Log CFU/mL Log 2 2

0 6 25 50 0 6 25 50 Xanthine/hypoxanthine conc (M) Xanthine/hypoxanthine conc (M)

8 8 C D

6 6

4 p=0.001 4 p= 0.3 Log CFU/mL Log CFU/mL Log

2 2

0 6 25 50 0 6 25 50 Xanthine/hypoxanthine conc (M) Xanthine/hypoxanthine conc (M)

Figure 4‎ .5 Bactericidal effects of H2O2 generated by breast milk (XO) and increasing concentrations of salivary xanthine and hypoxanthine (0-50 µM). Bacteria were incubated with breast milk and simulated neonatal saliva for 24 h at 37oC. The viable cells were assayed by the CFU grown on agar plates. Figures display the concentration-dependent effect of xanthine and hypoxanthine versus log CFU/mL: A) S. aureus, B) Salmonella spp., C) L. plantarum, and D) E.coli. Values represent the mean±SD, of triplicate experiments.

119

4.5.4 IN VIVO ORAL MICROBIOTA STUDIES OF BREAST-FED VERSUS FORMULA-FED NEONATES Aim 3: Compare and contrast the in vivo differences in oral microbiota within neonatal mouth from babies who have been breast-fed (i.e. exposed to the milk XO-generated superoxide system) to the oral microbiota of babies who were exclusively formula-fed, using PCR amplification of the bacterial 16S rRNA and deep sequencing.

4.5.4.1 Participants for in vivo studies of oral bacteria In this study, 38 neonates were enrolled (26 breast-fed and 12 formula-fed) with parents’ informed consent. For each infant recruited to this study, the data that was collected and included were the gestational age at delivery, the birth weight, and the age at the time of collection of the neonate’s saliva clinical specimen. These results are summarised in Table 4.4. There were no differences in the gestational age (p=0.3), the birth weight (p=0.2), or age (p=0.2) of breast-fed babies compared to formula- fed babies.

Table ‎4.4 Infants’ mean gestation and weight at delivery and the infants’ mean age at the time of collection of the oral bacteria samples by buccal swabs

Breast-fed Formula fed p value Male Female Total Male Female Total

12 14 26 5 7 12 Gestational age (w)1 Mean±SD 40±0.5 40±1 40±1 40±2 39±2 40±2 0.3 Range 39-41 38-42 38-41 37-41

Birth weight (g) Mean±SD 3488±316 3584±570 3539±464 3916±599 3664±227 3779±432 0.2 Range 3054-4110 2700-4492 3182-4720 3490-4092

Age (w)1 Mean±SD 6±1 6±2 6±2 7±1 6±2 6±1 0.2 Range 4-8 4-8 6-8 4-8

1w= weeks

120

4.5.4.2 Breast-fed and formula-fed associated oral taxa detected using 16S rRNA 'next- generation deep sequencing' The microbiota within neonatal saliva of breast-fed (n=26) and formula-fed (n=12) infants (age: 4 to 8 weeks) was examined using the Roche-454 GS FLX amplicon-sequencing platform. Sequencing of 38 samples yielded 522,275 curated high quality reads following extraction of 11.5% low quality sequences, which were removed from the original dataset. The maximum and minimum number of sequences for each sample were 18,903 and 8,132 respectively. A refraction method was applied to normalise the sequence counts, to obtain the similar proportion of OTUs for each sample with respect to its read count. This allowed a comparison of sample diversity at an equal sequencing depth. Bacterial loads in the collected samples were not compared, as there was no effective control for differences in the sampled volumes; instead, each sample DNA concentration was initially standardised to 10±0.9 ng/μL. The raw data of 16S rRNA are shown in Appendix 26.The curated sequence dataset were published in Figshare repository (Appendix 27).

4.5.4.3 Phylum level comparison of breast-fed and formula-fed oral bacterial communities The OTU analyses of oral swabs from 26 breast-fed and 12 formula-fed infants clustered within 5 bacterial phyla: Actinobacteria, Bacteroidetes, Firmicutes, Fusobacteria, and Proteobacteria. Phylum-level distribution patterns demonstrated that the microbiota within neonatal saliva of breast-fed babies was distinct from formula-fed babies.

From the breast-fed neonatal buccal swab specimens, the percent abundances of the most prevalent phyla identified were: Firmicutes, Actinobacteria, and Proteobacteria. In contrast, for the formula- fed infants the most prevalent phyla detected within the buccal swab specimens were: Firmicutes, Bacteroidetes, and Actinobacteria (Figure 4.6A).

Firmicutes represented the predominant phylum in the two experimental cohorts, accounting for (mean ± SD) 96.3%±0.04 in breast-fed infants, compared to 95.3%±0.05 in formula-fed infants (p= 0.5). Actinobacteria represented 2.1%±0.03 in breast-fed infants, compared to 1.6%±0.03 in formula-fed (p=0.8). Proteobacteria were significantly higher in breast-fed infants (0.23%±0.01) compared to in formula-fed infants (0.01%±0.001) (p=0.04). Finally, Bacteroidetes were significantly lower in breast-fed infants (0.11%±0.01) compared to in formula-fed infants (1.8%±0.02) (p=0.01).

Since the phylum Firmicutes was the predominant phylum within buccal swabs of both the breast- fed and formula-fed infants, the relative distributions of the remaining bacterial phyla were re- graphed after excluding the Firmicutes (Figure 4.6B).

121

A 100 Actinobacteria Bacteroidetes 80 Firmicutes Fusobacteria Proteobacteria 60

40 % total community total %

20

0

Breast-fed Formula-fed

15 B Actinobacteria Bacteroidetes Fusobacteria Proteobacteria 10

% total community total % 5

0

Breast-fed Formula-fed

Figure 4 .6 Bar graph showing the relative distributions of the major oral phyla operational taxonomic units (OTUs). (A) Breast-fed versus formula-fed and (B) Breast-fed vs formula-fed excluding the Firmicutes. Taxa were detected by 16S rRNA analysis using Roche-454 GS FLX deep sequencing. Each column within the bar graph represents the precentage of the bacterial 16SrRNA gene detected within a clinical specimen.

122

4.5.4.4 Family level comparison of breast-fed and formula-fed oral bacterial communities The bacterial families within buccal swabs of breast-fed and formula-fed infants were analysed. Distributions of the family of the major three bacterial phyla (i.e. Firmicutes, Actinobacteria, Bacteroidetes) found in this study are shown as pie graphs in Figure 4.7. The Firmicutes phylum was the predominate OTU identified in buccal swabs of both the breast-fed and formula-fed infants; however, the most abundant families within this phylum varied depending on the cohort. In the breast-fed neonates, the Gemellaceae and the Lactobacillaceae accounted for 5% and 2% of the bacteria identified within neonatal buccal swabs compared to 3% and 0% respectively within buccal swabs of formula-fed infants. The Streptococcaceae was the most abundant family detected within the Firmicutes phylum, accounting for 83% of bacterial oral isolates in breast-fed infants and 93% of isolates in formula-fed infants.

The abundance of the Actinobacteria phylum OTUs showed marked variations between the two experimental cohorts. Within the Actinobacteria the bacterial family Micrococcaceae accounted for the majority of bacteria, 57% of isolates within buccal swabs of breast-fed infants compared to 93% of isolates for formula-fed infants. However, the greatest differences in the microbiota noted between these two groups of infants were high abundances the Actinomycetaceae family (19% of all isolates) and Coriobacteriaceae family (22%) in the buccal swabs of breast-fed infants, compared to only 2% and 3% of isolates respectively in formula-fed infants. In addition, 1% of the bacterial isolates in buccal swabs of breast-fed infants were of the Bifidobacteriaceae family but these bacteria were not detected within the saliva of formula-fed infants.

Whilst Prevotellaceae was the major family of the Bacteroidetes phylum detected in both cohorts, the abundance of this family also varied dramatically. Prevotellaceae accounted for 65% of the family isolates within buccal swabs of breast-fed infants, compared to 91% of isolates for formula fed-infants. Similarly, marked variations were noticed in the abundance of Porphyromonadaceae and Flavobacteriaceae families in the buccal swab specimens from breast-fed infants, 33% and 2% respectively, compared to only 9% and 0% respectively in specimens from formula-fed infants.

123

Breast-fed Formula-fed

Staphylococcaceae Gemellaceae

Carnobacteriaceae Lactobacillaceae

Streptococcaceae Firmicutes Lachnospiraceae Veillonellaceae

Actinomycetaceae

Micrococcaceae

Propionibacteriaceae

Segniliparaceae

Bifidobacteriaceae Actinobacteria

Coriobacteriaceae

Porphyromonadaceae

Prevotellaceae

Flavobacteriaceae Bacteroidetes

Figure 4.7 Relative abundances of the most common Firmicutes, Actinobacteria, and Bacteroidetes families in buccal swabs of breast-fed vs formula-fed infants. The pie charts represent the mean values.

124

4.6 DISCUSSION

Human neonates and older infants are more vulnerable to infectious microorganisms than older children and adults. Increasing evidence suggests that neonatal immune responses may not be fully developed, thereby predisposing a susceptibility to microbial imbalances [252]. Although XO is well known to have antibacterial properties as a result of ROS generation [47], this enzyme’s ability to kill bacteria nonetheless depends on the presence of appropriate substrates. Superoxide radicals such as peroxide are a well-known defence mechanism to inhibit and kill microbial pathogens. A similar mechanism may occur for both the breast and neonatal mouth and digestive tract as part of their innate immunity. Additionally, peroxide may encourage or at least provide a growth advantage to commensal microbiota. The present study was initiated after elevated concentrations of the XO substrates xanthine and hypoxanthine were detected in neonatal saliva (Chapter 2). It was hypothesised that the concentrations of these substrates when reacted with XO may produce H2O2 at physiologically active levels, i.e. which could affect bacterial growth (Chapter 3). Breast milk XO and its substrates in neonatal saliva thus constituted a H2O2-generating system, and this was investigated in this chapter for its ability to inhibit bacterial growth.

4.6.1 HYDROGEN PEROXIDE TITRATION AGAINST BACTERIA

The direct effect of H2O2 at micromolar concentrations on bacterial growth was tested by titrating bacterial growth against H2O2, but importantly this was performed at micromolar concentrations that would be normally considered pharmacologically ineffective. Initial investigation of the antibacterial activity of H2O2 was conducted using a simple and rapid method: serial dilutions of

H2O2 against four species of bacteria, with a spectrophotometric-based assay for turbidity. This showed significant variation in susceptibility of bacteria to micromolar H2O2, with S. aureus being the most sensitive to H2O2 at a ‘physiological’ concentration, i.e. a concentration that was predicted from mixing breast milk XO with neonatal saliva. In contrast, Salmonella spp., E.coli, and L. plantarum were inhibited only at higher concentrations of H2O2.

However, the initial experiments were conducted using “Oxoid Nutrient Broth” growth media, which did not mimic the physiological conditions of a breast-feeding infant's mouth. For example, the lactoperoxidase system was not present in this experiment. In addition, the Nutrient Broth media was shown to contain high concentrations of purine and pyrimidine metabolites which support bacterial growth, consequently this may have confounded the effects of the inhibition studies. However, these experiments demonstrated for the first time that micromolar concentrations of

H2O2, which had previously been considered too low to be bactericidal actually, had strong 125 inhibitory effects on bacterial growth and that the response to H2O2 was species-specific for the 4 different bacterial strains tested. These experiments demonstrated clearly that H2O2 concentrations as low as 30 µM could inhibit the growth of S. aureus, and support a role for peroxide as a cellular messenger similar to nitric oxide [86].

4.6.2 BACTERIAL GROWTH WITH SALIVA AND BREAST MILK It was then shown that the addition of supplemented simulated neonatal saliva to breast milk also exhibited bactericidal effects. In order to mimic the physiological conditions within neonatal saliva, ‘simulated’ neonatal saliva was used, i.e. adult saliva that was heat-inactivated and supplemented with various components. This made it possible to 'dissect' the physiological activities of neonatal saliva. This was added to breast milk and bacteria, so that the culture assays were performed without the addition of bacterial growth enhancing media such as Nutrient Broth. This physiological approach also contained other antibacterial systems such as the lactoperoxidase-system and was therefore a more complex system compared to the direct H2O2 titration experiments that were initially conducted to demonstrate antibacterial activity.

Using the physiological simulated saliva it was demonstrated that nucleotide precursor supplementation (nucleosides, bases, but not xanthine, hypoxanthine) of saliva had no effect on S. aureus, Salmonella spp. and E.coli compared to the negative control, suggesting these species did not benefit from nucleosides and bases as growth factors, i.e. these species were exhibiting prototrophic behaviour. By contrast, the combination of breast milk and fully supplemented saliva (containing xanthine/hypoxanthine) exhibited a bactericidal effect on S. aureus and Salmonella spp., which was prevented/reversed by oxypurinol, an XO inhibitor. This demonstrated that the XO- saliva system produced sufficient amounts of H2O2 to inhibit the growth of bacteria. In particular, S. aureus grown within the physiological saliva/breast milk system was shown to have inhibition kinetics similar to those observed with the H2O2 titration experiments.

In early studies of the XO antibacterial effect, Green et al [253] and Lipmann et al., [254] demonstrated the inhibition of growth of S. aureus by purified bovine milk XO. They attributed these findings to the cytotoxic effect of XO-generated hydrogen peroxide. In addition, Stevens et al. [71] demonstrated that E. coli and Salmonella spp. were inhibited by this system, after adding 100 µM hypoxanthine to breast milk.

126

Although these some studies have proposed that bacterial growth inhibition was caused by H2O2, others have suggested that XO-derived H2O2 exerts its antibacterial effects by activating the lactoperoxidase system [255].

Whilst S. aureus asymptomatically colonises the surface of the skin, it is also a major opportunistic human pathogen [256], and the most common causative pathogen of mastitis [195]. S. aureus causes serious infections in neonatal units, being responsible for cellulitis, abscesses, bacteraemia, and sepsis infections in neonates [257-259]. These infections may also be difficult to treat as many S.aureus strains have acquired multiple antibiotic resistant genes, so that greater than 60% of S. aureus isolates are now resistant to methicillin [87].

Lactobacillus species are major commensals within the GIT [260] and within the female lower genital tract (LGT). These bacteria are transferred to babies by vertical transmission at the time of birth [261]. In this current study there was a noticeable, but not significant, stimulation of Lactobacillus growth in the presence of nucleotide metabolites in simulated saliva in comparison to the other tested bacteria, indicating that this commensal bacterium was auxotrophic in its growth habit. This increase in L. plantarum growth was reversed when xanthine and hypoxanthine were added to the supplemented saliva aliquots, and growth was restored when oxypurinol, the potent inhibitor of XO, was added to this mixture, however this was not significant (p=0.09).

The increase in the growth of L. plantarum in the presence of nucleotide metabolites in saliva is suggestive of an important role in the selection and the maintenance of commensal bacterial microbiota. Lactobacillus species are important commensal flora; in particular L. gasseri OLL2716 strain as probiotic bacteria, showed a significant suppression of the growth of H. pylori [262].The continuous secretion of these metabolites in neonatal saliva especially in the first few weeks postpartum (see Chapter 2, Longitudinal study) may select for commensal bacterial microbiota by enhancing the growth and consequently lead to establishment of important bacteria within normal flora communities in the early stages of life. Development of this bacterial community is crucial for the establishment of a suitable innate immunity [263, 264]. L. plantarum showed intolerance to

H2O2 generated by the xanthine/hypoxanthine-XO system (p =0.001). This was in contrast with direct H2O2 titration of Lactobaccilus growth, raising the possibility that the milk lactoperoxidase system contributed to the inhibition of this species.

For these bacterial growth experiments with supplemented saliva and breast milk, it was reasonable to assume that salivary peroxidase was inactivated by the heating process used to inactivate antibodies. However, the breast milk lactoperoxidase was highly active: this was confirmed by assays (not shown here). But in the adult mouth H2O2 produced by bacterial metabolism is 127 presumed to serve as a substrate for the salivary peroxidase system. In neonates these peroxidative bacteria may not be established and therefore the presence of the milk XO-saliva hypoxanthine/xanthine system may be of critical importance within the early neonatal period. Peroxidase has been shown to be important in the defence of the lung against bacterial infection [265]. Further studies of lactoperoxidase in terms of its protective effects for the mouth and upper GIT are needed, especially at the physiological concentrations present in human milk and neonatal saliva.

The use of a peroxide defence against bacteria exists in other natural mechanisms. Recently, reactive oxygen species (ROS) generation as a defence system has been described in honey. Honey contains the enzyme , which at some dilutions, reacts with its excess substrate, glucose, to generate H2O2. This reaction has substantial inhibitory effects against several bacterial species [266-268]. However, the concentrations of H2O2 produced from this system in honey are much greater (as high as 3,000 µM) [269] than those generated by the milk XO system (i.e. approximately 30-100 µM in human milk, and estimated up to 300 µM in bovine milk [79]). These findings indicated that honey may be used externally to treat a wide range of wounds including abscesses, surgical wounds, traumatic wounds, burns, and ulcers of varied aetiologies using a range of dilutions; this provides further evidence for the natural occurrence of ROS generation contributing to non-specific host defence systems.

The XO-peroxide system may contribute to neonatal GIT innate immunity by playing an important role in prevention of opportunistic infections. In particular, XO can generate ROS as well as RNS under conditions present within the neonatal GIT (of low oxygen tension and neutral stomach pH), which is optimal for the XO metabolic reaction. Absence of this system may increase the risk of opportunistic GIT infections such as NEC. This disease has a high risk in preterm babies, who rely on formula feeding (and who also invariably have their saliva suctioned) [270, 271]. It has been shown in this project that the activity of XO is absent from formula feed preparations as well as pasteurized breast milk (see Chapter 3). Lactoperoxidase is also absent from formula and pasteurised milks (results not shown). Therefore, in the absence of milk XO and lactoperoxidase, this antimicrobial defence mechanism in the neonatal mouth, and then the GIT, may partially explain the greater susceptibility of bottle-fed infants to infection.

4.6.3 IN VIVO ORAL MICROBIOTA STUDIES OF BREAST-FED VERSUS FORMULA-FED NEONATES Upon delivery, the neonate is exposed to a wide variety of microbes, many of which are provided by the mother during and after birth. Importantly, there is increasing evidence that an imbalance in 128 the microbial communities has an impact on human health [272, 273]. Therefore, in the current study, 16S rRNA gene deep sequencing, a culture-independent method, has been utilised to screen the oral microbiota composition of healthy breast-fed and formula-fed infants. This study was conducted to address the questions of how oral bacterial communities are structured by the mode of feeding, and the differences existing between the oral bacteria in breast-fed compared to formula- fed infants.

The microbiota within the infant GIT has been studied previously and remarkable variations were demonstrated according to the feeding style [227, 274-276]. For example, it has been found that at approximately 6 days after birth, the number of CFU/mL of Lactobacilli in the faeces of breast-fed infants was 1000 times higher than that of Enterobacteriae, but 10 times less in the faeces of bottle- fed infants [277]. At the time of writing, the study presented here is the first to investigate the effects of breast-feeding versus formula-feeding on the selection of microbiota within the neonatal mouth.

The study samples were only collected from vaginally delivered babies to minimize the effect of delivery mode on oral bacterial communities. This is important because the delivery mode has been shown to shape the oral bacteria in human infants [111]. In previous studies [278], 64-82% of reported cases of methicillin-resistant S. aureus (MRSA) infections in newborns occurred in Caesarean-delivered infants, suggesting that the delivery mode has an influence on bacterial communities.

Interestingly, the breast-fed and formula-fed groups had distinct microbiota distribution patterns. In addition, there was marked inter-individual variation within each group, in the taxon-level community compositions along with some variation at the phylum level. Experimental variability could have been further assessed by repeat analyses of all samples, repeat analysis of samples from each individual, or from obtaining a larger sample size, to improve evaluation of the inter- and intra- individual differences. However, despite the limitations provided by ethical restrictions, limited access to the sequencing facility (i.e. AGRF), and the inevitable restricted funding for a PhD thesis project, the results were consistent with other studies in which large inter-individual variations were reported in the healthy oral cavity [107] and intestinal tract [211, 275, 279, 280]. This suggests that various bacterial species might respond to differences in, for example, host metabolism that result in differences in the oral environmental niche of individuals. This may serve as a hallmark of microbiota stability, as well as potentially being a biomarker for human health – fruitful areas for further research.

129

This study established that formula-fed infants had a higher prevalence of Bacteroidetes within their saliva. This is consistent with a study of infant gut microbiota, which reported increased Bacteroidetes associated with formula feeding [227]. Apart from Bacteroidetes, the Bifidobacteriaceae family was detected in the mouths of breast-fed infants but was absent within the mouths of formula-fed infants. Bifidobacterium spp. are among the first anaerobic bacteria to reach high concentrations within the GIT of most neonates within the first few weeks of life. The prevalence of this bacterial species has been shown to be enhanced by breast feeding [237].

Since saliva and breast milk contain growth-enhancing and antibacterial factors such as amino acids, organic acids, proteins, enzymes, and immunoglobulins, the variations reported here in infant oral microbiota cannot be attributed solely to the XO-system. However, from the in vitro experimental data shown above, it is apparent that XO-system has an effect on bacterial selection and growth. It is also important to mention here that nucleotide metabolites are present in bacterial growth media as demonstrated in this study (Table 4.3), some of these metabolites were detected at physiological concentrations, similar to the concentrations detected within neonatal saliva; this confirms their importance in enhancing bacterial growth, and thereby modulating oral microbiota.

130

Chapter 5

Conclusions and future directions

131

5.1 CONCLUSIONS

Purines and pyrimidines in adult and neonatal saliva: It is clear from the studies presented here that nucleotide metabolites are excreted actively by the salivary glands and that saliva is not just a plasma filtrate. The significance of this has not appeared in the literature, namely that saliva may provide a unique fluid to aid in the diagnosis of metabolic disorders. Furthermore, it has been shown for the first time that higher concentrations of these metabolites exist in neonatal saliva compared to adult saliva, and that the transition occurs during the first 6 months of life. It is reasonable to conclude that the raised concentrations of nucleotide metabolites in neonatal saliva (compared to adult saliva) must play a physiological role rather than being simply the result of 'birth stress'. This was demonstrated by the highly selective nature of the nucleosides and bases found in saliva (i.e. non-functional metabolites were not found in saliva), and the fact the effect persisted beyond the neonatal period. This led to the conclusion that raised nucleotide metabolites in neonatal saliva have a role in regulating microbiota and consequently the enhancement of neonatal innate immunity.

Breast milk XO and generation of peroxide: In particular, high levels of hypoxanthine and xanthine were found in the neonatal mouth. From these elevated levels of XO substrates, it was hypothesised that there must be some link to XO in milk. From the evidence presented, it can be seen that xanthine and hypoxanthine in the neonatal mouth react with milk XO to generate reactive oxygen species (ROS). These highly reactive radicals are well known to have antibacterial properties. It was thus considered that this mechanism was likely have a role in neonatal innate immunity.

Bacterial studies: The bacterial studies were divided into 2 aspects: in vitro inhibition and stimulation of 4 test species of bacteria by peroxide and nucleotide metabolites, and an in vivo study of the effect of breast milk on selection of oral bacteria. The conclusions reached from the in vitro experiments were that an ‘ROS cascade’ produced by XO and its substrates can inhibit pathogenic bacteria at physiological concentrations, suggesting that the XO system may have an important role in neonatal innate immunity as well as preventing mastitis in the mother. This was the first time that the milk XO-peroxide system was considered as working beyond the breast. In addition, it was shown that other nucleosides and bases in neonatal saliva could serve as growth factors for some commensal bacteria and consequently may help to regulate the commensal bacterial community in the early stages of life. These nucleosides and bases may also have a role in feeding developing gut cells in the newborn, although this was not examined here. The in vivo study was important, as it

132 was the first to show that there was an effect of breastfeeding on oral bacteria. Interestingly, if no effect had been demonstrated then the other hypotheses would have been more difficult to justify.

Overall conclusions: The results presented here demonstrated the following: nucleosides and bases are higher in neonatal saliva than adult saliva; xanthine and hypoxanthine in saliva react with breast milk XO to generate H2O2, which can inhibit the growth of specific bacteria. The same mechanism may also occur in the mammary gland, supported by the presence of H2O2 in fresh breast milk. The nucleoside and bases in neonatal saliva may provide a good source for commensal flora to enhance the microbiota communities in the neonates’ mouth and gastrointestinal tract. Overall, the experiments provided evidence for the important role of regulation of the neonatal microbiome by breast milk and saliva.

5.2 FUTURE STUDIES

1. Measurement of nucleotides (i.e. in addition to nucleosides and bases presented here) might be important future approaches. Due the difficulty in measuring nucleotides at low concentrations, as previously reported for adults, a HPLC method compatible with mass spectrometry was not available, and there was insufficient time to establish such a method for nucleotides – this work is continuing. 2. It was noted that some adults tend to have higher concentrations of nucleotide metabolites such as xanthine, hypoxanthine and inosine, which might be explained by the genetic variation (e.g. SNPs), which includes racial/ethnic groups. Therefore, investigation of the transport mechanisms of these metabolites into saliva for both neonates and adults and monitoring the plasma level in neonates after birth would be an important focus of the future studies. 3. It has been also demonstrated in this thesis that mixing breast milk XO with neonatal saliva elicits ROS generation, and this system could be of importance at early stages of life. Besides the

importance of ROS as bactericidal factors, the XO system serves as donor of H2O2 for the lactoperoxidase system to generate other antibacterial metabolites (e.g. hypothiocyanite). There

is accumulating evidence that in adults the salivary lactoperoxidase system may obtain H2O2 from the metabolism of oral microbiota, however when the microbiota are not established in the

neonatal mouth another source of H2O2, namely the milk XO system, might be critical. Thus, further investigation of these antibacterial systems using physiological conditions, rather than growth media, may improve our understanding of the innate immunity of neonates and early infants.

133

4. Future studies might also include the ability of XO to directly generate RNS. XO can convert .- - NO2 to NO, the latter can react with O2 to produce NOO a powerful antibacterial agent. The .- importance of the XO system as demonstrated here is to donate O2 to the above mechanism to generate bactericidal RNS, particularly in the neonatal stomach where conditions are appropriate for this mechanism.

5. There is some evidence that ROS such as H2O2 play an important role in cell signalling and regulation of cell proliferation and apoptosis. Importantly, it has been demonstrated in this thesis

that H2O2 is present in breast milk as well as regenerated in vitro by adding neonatal saliva to breast milk at physiological concentration. Thus, this might be essential for gut cell signalling and growth, during the rapid cell growth in neonates. This topic is important to investigate in the future. Neonatal gut cell growth may also benefit from salivary nucleosides/bases. 6. Some the in vitro experiments of bacterial growth reported in this thesis will benefit from being repeated and extended, e.g. stimulation studies with supplemented saliva: a) One control experiment to perform would be to incubate bacteria with supplemented saliva alone (i.e. no milk) – this is what happens a lot of the time in the baby's mouth. This would show whether the purine and pyrimidine supplementation actually assisted growth of some bacteria. This has never been done correctly because previous researchers have used growth media that are not physiological. Indeed, 'supplemented saliva' may be an excellent physiological medium for growing some commensal bacterial species. b) In retrospect, the generation times (24 h) used were probably too short for Lactobacillus, so growth stimulation was not fully measured. 7. Only 4 species of bacteria were examined: other species are awaiting study. 8. The presence of nucleotide metabolites could be helpful markers for the diagnosis of some metabolic disorders.

134

5.3 SUMMARY OF RESEARCH RATIONALE, OUTCOMES AND BENEFITS

1. By studying one of the earliest postnatal mechanisms of regulating oral and gut microbiota, this research may improve the oral and gut health of newborns through to adults, by providing an improved understanding of natural anti-bacterial and pro-biotic mechanisms. 2. Specifically, this research could assist in understanding the pathogenesis of necrotising enterocolitis (NEC) that can affect premature and some full-term babies. It may also assist in providing new methods of prevention of NEC, e.g., through new milk formulations and feeding mechanisms that more closely imitate the natural regulation of microbiota. 3. It may contribute to the prevention of middle ear infections, which are more common in non- breast-fed infants and after weaning. 4. The study may also initiate new lines of inquiry into how best to establish or conserve normal oral and gut flora in young children with severe illness or requiring major surgery (e.g. for abdominal, chest or cardiac malformations). 5. It is likely that this project will provide yet more evidence to support breast feeding compared to milk formula feeding. 6. Lastly, these considerations may also apply to some nursing mothers, e.g. with metabolic disorders. For example, although small in number, there has never been a survey of oral/gastric infections among babies of mothers with (inherited xanthine oxidase deficiency), or for nursing mothers on . For example, there are presently 2 identified xanthinuric women of child-bearing age in the Brisbane area, and many more world-wide.

135

Bibliography

136

1. Geldart, S.E. and P.R. Brown, Separation of purine and pyrimidine bases by capillary zone electrophoresis with carbonate buffers. J Chromatogr A, 1999. 831: p. 123-129. 2. Rudolph, F.B., The biochemistry and physiology of nucleotides. J Nutr, 1994. 124: p. 124S- 127S. 3. Garret, R. and C. Grisham, Biochemistry. 4th ed. 2010, Boston: Cole. Boston. 4. Traut, T.W., Physiological concentrations of purines and pyrimidines. Mol Cell Biochem, 1994. 140: p. 1-22. 5. Zimmermann, H., Extracellular metabolism of ATP and other nucleotides. Naunyn Schmiedebergs Arch Pharmacol, 2000. 362: p. 299-309. 6. Venkatakrishnan, A.J., X. Deupi, G. Lebon, C.G. Tate, G.F. Schertler, and M.M. Babu, Molecular signatures of G-protein-coupled receptors. Nature, 2013. 494: p. 185-194. 7. Newton, R.P. and C.J. Smith, Cyclic nucleotides. Phytochemistry, 2004. 65: p. 2423-2437. 8. Connolly, G.P. and J.A. Duley, Uridine and its nucleotides: biological actions, therapeutic potentials. Trends Pharmacol Sci, 1999. 20: p. 218-225. 9. Evans, D.R. and H.I. Guy, Mammalian pyrimidine biosynthesis: Fresh insights into an ancient pathway. J Biol Chem, 2004. 279: p. 33035-33038. 10. Fischer, M. and A. Bacher, Biosynthesis of flavocoenzymes. Nat Prod Rep, 2005. 22: p. 324-350. 11. Koch-Nolte, F., F. Haag, A.H. Guse, F. Lund, and M. Ziegler, Emerging roles of NAD+ and its metabolites in cell signaling. Sci Signal, 2009. 2: p. 1-5. 12. Loffler, M., L.D. Fairbanks, E. Zameitat, A.M. Marinaki, and H.A. Simmonds, Pyrimidine pathways in health and disease. Trends Mol Med, 2005. 11: p. 430-437. 13. Nyhan, W.L., Disorders of purine and pyrimidine metabolism. Mol Genet Metab, 2005. 86: p. 25-33. 14. Dudzinska, W., A.J. Hlynczak, E. Skotnicka, and M. Suska, The of human erythrocytes. Biochemistry Biokhimiia, 2006. 71: p. 467-475. 15. Kahner, B.N., H. Shankar, S. Murugappan, G.L. Prasad, and S.P. Kunapuli, Nucleotide signaling in platelets. J Thromb Haemost, 2006. 4: p. 2317-2326. 16. Pontarin, G., A. Fijolek, P. Pizzo, P. Ferraro, C. Rampazzo, T. Pozzan, L. Thelander, P.A. Reichard, and V. Bianchi, Ribonucleotide reduction is a cytosolic process in mammalian cells independently of DNA damage. Proc Natl Acad Sci U S A, 2008. 105: p. 17801-17806. 17. Uauy, R., R. Quan, and A. Gil, Role of Nucleotides in Intestinal Development and Repair: Implications for Infant Nutrition. J Nutr, 1994. 124: p. 1436S-1441S. 18. Kilstrup, M., K. Hammer, J.P. Ruhdal, and J. Martinussen, Nucleotide metabolism and its control in lactic acid bacteria. FEMS Microbiol Rev, 2005. 29: p. 555-590. 137

19. Camici, M., V. Micheli, P.L. Ipata, and M.G. Tozzi, Pediatric neurological syndromes and inborn errors of purine metabolism. Neurochem Int, 2010. 56: p. 367-378. 20. Hoeger, U., N. Rebscher, and G. Geier, Metabolite supply in oocytes of Nereis virens: role of nucleosides. Hydrobiologia, 1999. 402: p. 163-174. 21. de, B.A.P., B.H. van, S.B. Nabuurs, W.F. Arts, J. Christodoulou, and J. Duley, PRPS1 mutations: four distinct syndromes and potential treatment. Am J Hum Genet, 2010. 86: p. 506-518. 22. Cameron, J.S., F. Moro, and H.A. Simmonds, , uric acid and purine metabolism in paediatric nephrology. Pediatr Nephrol, 1993. 7: p. 105-118. 23. Bzowska, A., E. Kulikowska, and D. Shugar, Purine nucleoside phosphorylases: properties, functions, and clinical aspects. Pharmacol Ther, 2000. 88: p. 349-425. 24. Harrison, R., Physiological roles of xanthine oxidoreductase. Drug Metab Rev, 2004. 36: p. 363-375. 25. van, K.A.B., L.H. van, M. Loffler, and G.A.H. van, Analysis of pyrimidine synthesis "de novo" intermediates in urine and dried urine filter- paper strips with HPLC-electrospray tandem mass spectrometry. Clin Chem, 2004. 50: p. 2117-2124. 26. Davis, R.H., Carbamyl phosphate synthesis in Neurospora crassa. II. Genetics, metabolic position, and regulation of arginine-specific carbamyl phosphokinase. Biochim Biophys Acta, 1965. 107: p. 54-68. 27. Williams, L.G. and R.H. Davis, Pyrimidine-Specific Carbamyl Phosphate Synthetase in Neurospora crassa. J Bacteriol, 1970. 103: p. 335-341. 28. Jurecka, A., Inborn errors of purine and pyrimidine metabolism. J Inherit Metab Dis, 2009. 32: p. 247-263. 29. van, K.A.B., L.H. van, and G.A.H. van, Activity of pyrimidine degradation enzymes in normal tissues. Nucleosides Nucleotides Nucleic Acids, 2006. 25: p. 1211-1214. 30. Levine, R.L., N.J. Hoogenraad, and N. Kretchmer, A review: biological and clinical aspects of pyrimidine metabolism. Pediatr Res, 1974. 8: p. 724-734. 31. Marsh, S., Thymidylate synthase pharmacogenetics. Invest New Drugs, 2005. 23: p. 533- 537. 32. Zaharevitz, D.W., L.W. Anderson, N.M. Malinowski, R. Hyman, J.M. Strong, and R.L. Cysyk, Contribution of de-novo and salvage synthesis to the uracil nucleotide pool in mouse tissues and tumors in vivo. FEBS J, 1992. 210: p. 293-296. 33. Gupta, R., K. Ramachandran, and J. Chowhan, Diagnostic and prognostic value of the urinary excretion of beta-aminoisobutyric acid in patients with urinary bladder cancer. Indian J of Clin Biochem, 1996. 11: p. 59-61. 138

34. Van Gennip, A.H., E.J. Van Bree-blom, N.G.G.M. Abeling, A.J. Van Erven, and P.A. Voûte, [beta]-Aminoisobutyric acid as a marker of thymine catabolism in malignancy. Clin Chim Acta, 1987. 165: p. 365-377. 35. Zrenner, R., M. Stitt, U. Sonnewald, and R. Boldt, Pyrimidine and purine biosynthesis and degradation in plants. Annu Rev Plant Biol, 2006. 57: p. 805-836. 36. Arsene-Ploetze, F., H. Nicoloff, B. Kammerer, J. Martinussen, and F. Bringel, Uracil salvage pathway in Lactobacillus plantarum: Transcription and genetic studies. J Bacteriol, 2006. 188: p. 4777-4786. 37. Beck, C.F., J.L. Ingraham, J. Neuhard, and E. Thomassen, Metabolism of pyrimidines and pyrimidine nucleosides by Salmonella typhimurium. J Bacteriol, 1972. 110: p. 219-228. 38. O'Donovan, G.A. and J. Neuhard, Pyrimidine metabolism in microorganisms. Bacteriol Rev, 1970. 34: p. 278-343. 39. Martinussen, J., P.S. Andersen, and K. Hammer, Nucleotide metabolism in Lactococcus lactis: salvage pathways of exogenous pyrimidines. J Bacteriol, 1994. 176: p. 1514-1516. 40. Jyssum, S. and K. Jyssum, Metabolism of pyrimidine bases and nucleosides in Neisseria meningitidis. J Bacteriol, 1979. 138: p. 320-323. 41. Miyamoto, Y., T. Masaki, and S. Chohnan, Characterization of N-deoxyribosyltransferase from Lactococcus lactis subsp. lactis. Biochim Biophys Acta, 2007. 1774: p. 1323-1330. 42. Suzuki, I., S. Kato, T. Kitada, N. Yano, and T. Morichi, Growth of Lactobacillus bulgaricus in Milk. 2. Characteristics of Purine Nucleotides, Pyrimidine Nucleotides, and Synthesis. J Dairy Sci, 1986. 69: p. 971-978. 43. Harrison, R., Structure and function of xanthine oxidoreductase: where are we now? Free Radic Biol Med, 2002. 33: p. 774-797. 44. Parks, D.A. and D.N. Granger, Xanthine oxidase: biochemistry, distribution and physiology. Acta Physiol Scand Suppl, 1986. 548: p. 87-99. 45. Pritsos, C.A., Cellular distribution, metabolism and regulation of the xanthine oxidoreductase enzyme system. Chem Biol Interact, 2000. 129: p. 195-208. 46. Saksela, M., R. Lapatto, and K.O. Raivio, Xanthine oxidoreductase and enzyme activity in developing human tissues. Biol Neonate, 1998. 74: p. 274-280. 47. Martin, H.M., J.T. Hancock, V. Salisbury, and R. Harrison, Role of Xanthine Oxidoreductase as an Antimicrobial Agent. Infect Immun, 2004. 72: p. 4933-4939. 48. Fox, P.F. and A.L. Kelly, Indigenous enzymes in milk: Overview and historical aspects— Part 1. Int Dairy J, 2006. 16: p. 500-516. 49. Raivio, K., M. Saksela, and R. Lapatto, Xanthine oxidoreductase-role in human pathophysiology and in hereditary xanthinuria. OMMBID, 2001. 2: p. 2639-2652. 139

50. Enroth, C., B.T. Eger, K. Okamoto, T. Nishino, T. Nishino, and E.F. Pai, Crystal structures of bovine milk xanthine dehydrogenase and xanthine oxidase: structure-based mechanism of conversion. Proc Natl Acad Sci U S A, 2000. 97: p. 10723-10728. 51. Xu, P., X. Zhu, T. Huecksteadt, A. Brothman, and J. Hoidal, Assignment of human xanthine dehydrogenase gene to p 22. Genomics(San Diego, Calif), 1994. 23: p. 289- 291. 52. Brondino, C.D., M.J. Romao, I. Moura, and J.J. Moura, Molybdenum and tungsten enzymes: the xanthine oxidase family. Curr Opin Chem Biol, 2006. 10: p. 109-114. 53. Okamoto, K., K. Matsumoto, R. Hille, B.T. Eger, E.F. Pai, and T. Nishino, The crystal structure of xanthine oxidoreductase during catalysis: implications for reaction mechanism and enzyme inhibition. Proc Natl Acad Sci U S A, 2004. 101: p. 7931-7936. 54. Kisker, C., H. Schindelin, and D.C. Rees, Molybdenum-cofactor-containing enzymes: structure and mechanism. Annu Rev Biochem, 1997. 66: p. 233-267. 55. Rasmussen, J.T., M.S. Rasmussen, and T.E. Petersen, Cysteines Involved in the Interconversion Between Dehydrogenase and Oxidase Forms of Bovine Xanthine Oxidoreductase. J Dairy Sci, 2000. 83: p. 499-506. 56. Nishino, T., K. Okamoto, Y. Kawaguchi, H. Hori, T. Matsumura, B.T. Eger, and E.F. Pai, Mechanism of the conversion of xanthine dehydrogenase to xanthine oxidase: identification of the two cysteine disulfide bonds and crystal structure of a non-convertible rat liver xanthine dehydrogenase mutant. J Biol Chem, 2005. 280: p. 24888-24894. 57. Della Corte, E. and F. Stirpe, The regulation of rat liver xanthine oxidase. Involvement of thiol groups in the conversion of the enzyme activity from dehydrogenase (type D) into oxidase (type O) and purification of the enzyme. Biochem J, 1972. 126: p. 739-745. 58. Godber, B., S. Sanders, R. Harrison, R. Eisenthal, and R.C. Bray, > or = 95% of xanthine oxidase in human milk is present as the demolybdo form, lacking . Biochem Soc Trans, 1997. 25: p. 519-525. 59. Vorbach, C., R. Harrison, and M.R. Capecchi, Xanthine oxidoreductase is central to the evolution and function of the innate immune system. Trends Immunol, 2003. 24: p. 512-517. 60. Zweier, J.L. and M.A. Talukder, The role of oxidants and free radicals in reperfusion injury. Cardiovasc Res, 2006. 70: p. 181-190. 61. Mittal, A., A.R. Phillips, B. Loveday, and J.A. Windsor, The potential role for xanthine oxidase inhibition in major intra-abdominal surgery. World J Surg, 2008. 32: p. 288-295. 62. Peglow, S., A.H. Toledo, R. Anaya-Prado, F. Lopez-Neblina, and L.H. Toledo-Pereyra, Allopurinol and xanthine oxidase inhibition in liver ischemia reperfusion. J Hepatobiliary Pancreat Sci, 2011. 18: p. 137-146. 140

63. Agarwal, A., A. Banerjee, and U.C. Banerjee, Xanthine oxidoreductase: a journey from purine metabolism to cardiovascular excitation-contraction coupling. Crit Rev Biotechnol, 2011. 31: p. 264-280. 64. Kooij, A., K.S. Bosch, W.M. Frederiks, and N.C.J. Van, High levels of xanthine oxidoreductase in rat endothelial, epithelial and connective tissue cells. A relation between localization and function? Virchows Arch B Cell Pathol Incl Mol Pathol, 1992. 62: p. 143- 150. 65. Linder, N., J. Rapola, and K. Raivio, Cellular expression of xanthine oxidoreductase protein in normal human tissues. Lab Invest, 1999. 79: p. 967-974. 66. Van, D.M.R.J., H. Vreeling-Sindelarova, J.P. Schellens, N.C.J. Van, and W.M. Frederiks, Ultrastructural localization of xanthine oxidase activity in the digestive tract of the rat. Histochem J, 1995. 27: p. 897-905. 67. Harrison, R., Milk xanthine oxidase: Properties and physiological roles. Int Dairy J, 2006. 16: p. 546-554. 68. Mather, I.H., A Review and Proposed Nomenclature for Major Proteins of the Milk-Fat Globule Membrane. J Dairy Sci, 2000. 83: p. 203-247. 69. Keenan, T.W., Historical Perspective: Milk Lipid Globules and Their Surrounding Membrane: A Brief History and Perspectives for Future Research. J Mammary Gland Biol Neoplasia, 2001. 6: p. 365-371. 70. McManaman, J.L., C.A. Palmer, R.M. Wright, and M.C. Neville, Functional regulation of xanthine oxidoreductase expression and localization in the mouse mammary gland: evidence of a role in lipid secretion. J Physiol, 2002. 545: p. 567-579. 71. Stevens, C.R., T.M. Millar, J.G. Clinch, J.M. Kanczler, T. Bodamyali, and D.R. Blake, Antibacterial properties of xanthine oxidase in human milk. Lancet, 2000. 356: p. 829-830. 72. Godber, B., J. Doel, G. Sapkota, D. Blake, C. Stevens, R. Eisenthal, and R. Harrison, Reduction of nitrite to nitric oxide catalyzed by xanthine oxidoreductase. J Biol Chem, 2000. 275: p. 7757-7763. 73. Millar, T.M., C.R. Stevens, N. Benjamin, R. Eisenthal, R. Harrison, and D.R. Blake, Xanthine oxidoreductase catalyses the reduction of nitrates and nitrite to nitric oxide under hypoxic conditions. FEBS Lett, 1998. 427: p. 225-228. 74. Lundberg, J.O., E. Weitzberg, and M.T. Gladwin, The nitrate-nitrite-nitric oxide pathway in physiology and therapeutics. Nat Rev Drug Discov, 2008. 7: p. 156-167. 75. Zhou, H., Y. Xu, T. Chen, I. Suzuki, and G. Li, Electrochemistry of xanthine oxidase and its interaction with nitric oxide. Anal Sci J, 2006. 22: p. 337-340.

141

76. Yener, F.Y., F. Korel, and A. Yemenicioglu, Antimicrobial activity of lactoperoxidase system incorporated into cross-linked alginate films. J Food Sci, 2009. 74: p. M73-79. 77. Welk, A., C. Meller, R. Schubert, C. Schwahn, A. Kramer, and H. Below, Effect of lactoperoxidase on the antimicrobial effectiveness of the thiocyanate hydrogen peroxide combination in a quantitative suspension test. BMC Microbiol, 2009. 9: p. 134. 78. Touch, V., S. Hayakawa, S. Yamada, and S. Kaneko, Effects of a lactoperoxidase- thiocyanate-hydrogen peroxide system on Salmonella enteritidis in animal or vegetable . Int J Food Microbiol, 2004. 93: p. 175-183. 79. Silanikove, N., F. Shapiro, and G. Leitner, Posttranslational ruling of xanthine oxidase activity in bovine milk by its substrates. Biochem Biophys Res Commun, 2007. 363: p. 561- 565. 80. Brieger, K., S. Schiavone, F.J. Miller, Jr., and K.H. Krause, Reactive oxygen species: from health to disease. Swiss Med Wkly, 2012. 142: p. w13659. 81. Guzik, T.J., N.E. West, E. Black, D. McDonald, C. Ratnatunga, R. Pillai, and K.M. Channon, Vascular superoxide production by NAD(P)H oxidase: association with endothelial dysfunction and clinical risk factors. Circ Res, 2000. 86: p. E85-90. 82. Droge, W., Free radicals in the physiological control of cell function. Physiol Rev, 2002. 82: p. 47-95. 83. Harman, D., Aging: a theory based on free radical and radiation chemistry. J Gerontol, 1956. 11: p. 298-300. 84. Harman, D., The aging process. Proc Natl Acad Sci U S A, 1981. 78: p. 7124-7128. 85. Alfadda, A.A. and R.M. Sallam, Reactive oxygen species in health and disease. J Biomed Biotechnol, 2012. 2012: p. 936486. 86. Linnane, A.W., M. Kios, and L. Vitetta, Healthy aging: regulation of the metabolome by cellular redox modulation and prooxidant signaling systems: the essential roles of superoxide anion and hydrogen peroxide. Biogerontology, 2007. 8: p. 445-467. 87. Feng, H., H. Xiang, J. Zhang, G. Liu, N. Guo, X. Wang, X. Wu, X. Deng, and L. Yu, Genome-wide transcriptional profiling of the response of Staphylococcus aureus to cryptotanshinone. J Biomed Biotechnol, 2009. 2009: p. 617509. 88. Perera, R.M. and N. Bardeesy, Cancer: when antioxidants are bad. Nature, 2011. 475: p. 43- 44. 89. Roussa, E., Channels and transporters in salivary glands. Cell Tissue Res, 2011. 343: p. 263- 287. 90. Dodds, M.W., D.A. Johnson, and C.K. Yeh, Health benefits of saliva: a review. J Dent, 2005. 33: p. 223-233. 142

91. de Almeida, P., A. Grégio, M. Machado, A. De Lima, and L. Azevedo, Saliva composition and functions: a comprehensive review. J Contemp Dent Pract, 2008. 9: p. 72-80. 92. Aps, J.K.M. and L.C. Martens, Review: The physiology of saliva and transfer of drugs into saliva. Forensic Sci Int, 2005. 150: p. 119-131. 93. Oppenheim, F.G., T. Xu, F.M. Mcmillian, S.M. Levitz, R.D. Diamond, G.D. Offner, and R.F. Troxler, Histatins, a Novel Family of Histidine-Rich Proteins in Human-Parotid Secretion - Isolation, Characterization, Primary Structure, and Fungistatic Effects on Candida-Albicans. J Biol Chem, 1988. 263: p. 7472-7477. 94. O'Connell, B.C., T. Xu, T.J. Walsh, T. Sein, A. Mastrangeli, R.G. Crystal, F.G. Oppenheim, and B.J. Baum, Transfer of a gene encoding the anticandidal protein histatin 3 to salivary glands. Hum Gene Ther, 1996. 7: p. 2255-2261. 95. Humphrey, S.P. and R.T. Williamson, A review of saliva: normal composition, flow, and function. J Prosthet Dent, 2001. 85: p. 162-169. 96. Nieuw Amerongen, A.V., A.J. Ligtenberg, and E.C. Veerman, Implications for diagnostics in the biochemistry and physiology of saliva. Ann N Y Acad Sci, 2007. 1098: p. 1-6. 97. Aps, J.K. and L.C. Martens, Review: The physiology of saliva and transfer of drugs into saliva. Forensic Sci Int, 2005. 150: p. 119-131. 98. Structure, function and diversity of the healthy human microbiome. Nature, 2012. 486: p. 207-214. 99. Segata, N., S.K. Haake, P. Mannon, K.P. Lemon, L. Waldron, D. Gevers, C. Huttenhower, and J. Izard, Composition of the adult digestive tract bacterial microbiome based on seven mouth surfaces, tonsils, throat and stool samples. Genome Biol, 2012. 13: p. R42. 100. Carpenter, G.H., The secretion, components, and properties of saliva. Annu Rev Food Sci Technol, 2013. 4: p. 267-276. 101. Marcotte, H. and M.C. Lavoie, Oral Microbial Ecology and the Role of Salivary Immunoglobulin A. Microbiol Mol Biol Rev, 1998. 62: p. 71-109. 102. Scannapieco, F.A., Saliva-Bacterium Interactions in Oral Microbial Ecology. Crit Rev Oral Biol Med, 1994. 5: p. 203-248. 103. Wade, W.G., The oral microbiome in health and disease. Pharmacol Res, 2013. 69: p. 137- 143. 104. Clarridge, J.E., Impact of 16S rRNA gene sequence analysis for identification of bacteria on clinical microbiology and infectious diseases. Clin Microbiol Rev, 2004. 17: p. 840-862. 105. Woo, P.C., S.K. Lau, J.L. Teng, H. Tse, and K.Y. Yuen, Then and now: use of 16S rDNA gene sequencing for bacterial identification and discovery of novel bacteria in clinical microbiology laboratories. Clin Microbiol Infect, 2008. 14: p. 908-934. 143

106. Dewhirst, F.E., T. Chen, J. Izard, B.J. Paster, A.C. Tanner, W.H. Yu, A. Lakshmanan, and W.G. Wade, The human oral microbiome. J Bacteriol, 2010. 192: p. 5002-5017. 107. Bik, E.M., C.D. Long, G.C. Armitage, P. Loomer, J. Emerson, E.F. Mongodin, K.E. Nelson, S.R. Gill, C.M. Fraser-Liggett, and D.A. Relman, Bacterial diversity in the oral cavity of 10 healthy individuals. ISME J, 2010. 4: p. 962-974. 108. Jenkinson, H.F. and R.J. Lamont, Oral microbial communities in sickness and in health. Trends Microbiol, 2005. 13: p. 589-595. 109. Ward, T.L., S. Hosid, I. Ioshikhes, and I. Altosaar, Human milk metagenome: a functional capacity analysis. BMC Microbiol, 2013. 13: p. 116. 110. Lif Holgerson, P., L. Harnevik, O. Hernell, A.C. Tanner, and I. Johansson, Mode of birth delivery affects oral microbiota in infants. J Dent Res, 2011. 90: p. 1183-1188. 111. Dominguez-Bello, M.G., E.K. Costello, M. Contreras, M. Magris, G. Hidalgo, N. Fierer, and R. Knight, Delivery mode shapes the acquisition and structure of the initial microbiota across multiple body habitats in newborns. Proc Natl Acad Sci U S A, 2010. 107: p. 11971- 11975. 112. Goi, N., Y. Hirai, H. Harada, A. Ikari, T. Ono, N. Kinae, M. Hiramatsu, K. Nakamura, and K. Takagi, Comparison of peroxidase response to mental arithmetic stress in saliva of smokers and non-smokers. J Toxicol Sci, 2007. 32: p. 121-127. 113. Enberg, N., H. Alho, V. Loimaranta, and M. Lenander-Lumikari, Saliva flow rate, amylase activity, and protein and electrolyte concentrations in saliva after acute alcohol consumption. Oral Surg Oral Med Oral Pathol Oral Radiol Endod, 2001. 92: p. 292-298. 114. Dawes, C., Physiological factors affecting salivary flow rate, oral sugar clearance, and the sensation of dry mouth in man. J Dent Res, 1987. 66: p. 648-653. 115. Kaufman, E. and I.B. Lamster, The diagnostic applications of saliva--a review. Crit Rev Oral Biol Med, 2002. 13: p. 197-212. 116. Pfaffe, T., J. Cooper-White, P. Beyerlein, K. Kostner, and C. Punyadeera, Diagnostic potential of saliva: current state and future applications. Clin Chem, 2011. 57: p. 675-687. 117. Kochanska, B., R.T. Smolenski, and N. Knap, Determination of adenine nucleotides and their metabolites in human saliva. Acta Biochim Pol, 2000. 47: p. 877-879. 118. Teeninga, N., Z. Guan, J. Freijer, A.F. Ruiter, M.T. Ackermans, J.E. Kist-van Holthe, T. van Gelder, and J. Nauta, Monitoring Prednisolone and Prednisone in Saliva: A Population Pharmacokinetic Approach in Healthy Volunteers. Ther Drug Monit, 2013: p. aheadofprint. 119. Tan, X. and Z. Song, Continuous, quantitative monitoring of roxithromycin in human saliva by flow injection chemiluminescence analysis. Appl Spectrosc, 2013. 67: p. 54-58.

144

120. Zuidema, J., K.M. Hold, D. de Boer, and R.A. Maes, Saliva as a specimen for therapeutic drug monitoring in pharmacies. Pharm World Sci, 1996. 18: p. 193-194. 121. Leung, A.K. and R.S. Sauve, Breast is best for babies. J Natl Med Assoc, 2005. 97: p. 1010- 1019. 122. Picciano, M.F., Nutrient composition of human milk. Pediatr Clin North Am, 2001. 48: p. 53-67. 123. Brown, A.M., M. Benboubetra, M. Ellison, D. Powell, J.D. Reckless, and R. Harrison, Molecular activation-deactivation of xanthine oxidase in human milk. Biochim Biophys Acta, 1995. 1245: p. 248-254. 124. do Nascimento, M.B. and H. Issler, Breastfeeding: making the difference in the development, health and nutrition of term and preterm newborns. Rev Hosp Clin Fac Med Sao Paulo, 2003. 58: p. 49-60. 125. Hamosh, M., Bioactive factors in human milk. Pediatr Clin North Am, 2001. 48: p. 69-86. 126. Lonnerdal, B., Biochemistry and physiological function of human milk proteins. Am J of Clin Nutr, 1985. 42: p. 1299-1317. 127. Lonnerdal, B., Nutritional and physiologic significance of human milk proteins. Am J Clin Nutr, 2003. 77: p. 1537-1543. 128. Saarinen, U.M., M.A. Siimes, and P.R. Dallman, Iron absorption in infants: High bioavailability of breast milk iron as indicated by the extrinsic tag method of iron absorption and by the concentration of serum ferritin. J Pediatr, 1977. 91: p. 36-39. 129. Picciano, M.F., Human milk: nutritional aspects of a dynamic food. Biol Neonate, 1998. 74: p. 84-93. 130. Janas, L.M. and M.F. Picciano, The nucleotide profile of human milk. Pediatr Res, 1982. 16: p. 659-662. 131. Sugawara, M., N. Sato, T. Nakano, T. Idota, and I. Nakajima, Profile of nucleotides and nucleosides of human milk. J Nutr Sci Vitaminol (Tokyo), 1995. 41: p. 409-418. 132. Cosgrove, M., Perinatal and infant nutrition. Nucleotides. Nutr, 1998. 14: p. 748-751. 133. Lerner, A. and R. Shamir, Nucleotides in infant nutrition: a must or an option. Isr Med Assoc J, 2000. 2: p. 772-774. 134. Yau, K.I., C.B. Huang, W. Chen, S.J. Chen, Y.H. Chou, F.Y. Huang, K.E. Kua, N. Chen, M. McCue, P.A. Alarcon, R.L. Tressler, G.M. Comer, G. Baggs, R.J. Merritt, and M.L. Masor, Effect of nucleotides on diarrhea and immune responses in healthy term infants in Taiwan. J Pediatr Gastroenterol Nutr, 2003. 36: p. 37-43. 135. Robinson, J.L., Bovine-Milk Orotic-Acid - Variability and Significance for Human- Nutrition. J Dairy Sci, 1980. 63: p. 865-871. 145

136. Barbosa-Cesnik, C., K. Schwartz, and B. Foxman, Lactation mastitis. JAMA, 2003. 289: p. 1609-1612. 137. Fetherston, C., Mastitis in lactating women: physiology or pathology? Breastfeed Rev, 2001. 9: p. 5-12. 138. Michie, C., F. Lockie, and W. Lynn, The challenge of mastitis. Arch Dis Child, 2003. 88: p. 818-821. 139. Foxman, B., H. D'Arcy, B. Gillespie, J.K. Bobo, and K. Schwartz, Lactation mastitis: Occurrence and medical management among 946 breastfeeding women in the United States. Am J Epidemiol, 2002. 155: p. 103-114. 140. Osterman, K.L. and V.A. Rahm, Lactation mastitis: bacterial cultivation of breast milk, symptoms, treatment, and outcome. J Hum Lact, 2000. 16: p. 297-302. 141. Van Gennip, A., Defects in metabolism of purines and pyrimidines. Ned Tijdschr Klin Chem, 1999. 24: p. 171-175. 142. Jabs, C.M., P. Neglen, B. Eklof, and E.J. Thomas, Adenosine, inosine, and hypoxanthine/xanthine measured in tissue and plasma by a luminescence method. Clin Chem, 1990. 36: p. 81-87. 143. Yamamoto, T., H. Koyama, M. Kurajoh, T. Shoji, Z. Tsutsumi, and Y. Moriwaki, Biochemistry of uridine in plasma. Clin Chim Acta, 2011. 412: p. 1712-1724. 144. Wung, W.E. and S.B. Howell, Simultaneous liquid chromatography of 5-fluorouracil, uridine, hypoxanthine, xanthine, uric acid, allopurinol, and in plasma. Clin Chem, 1980. 26: p. 1704-1708. 145. Tavazzi, B., G. Lazzarino, P. Leone, A.M. Amorini, F. Bellia, C.G. Janson, P.V. Di, L. Ceccarelli, S. Donzelli, J.S. Francis, and B. Giardina, Simultaneous high performance liquid chromatographic separation of purines, pyrimidines, N-acetylated amino acids, and dicarboxylic acids for the chemical diagnosis of inborn errors of metabolism. Clin Biochem, 2005. 38: p. 997-1008. 146. Simmonds, H., J. Duley, and P. Davies, Analysis of purines and pyrimidines in blood, urine, and other physiological fluids. Techniques in diagnostic human biochemical genetics: a laboratory manual, 1991: p. 397-424. 147. Schipper, R.G., E. Silletti, and M.H. Vingerhoeds, Saliva as research material: biochemical, physicochemical and practical aspects. Arch Oral Biol, 2007. 52: p. 1114-1135. 148. Punyadeera, C., G. Dimeski, K. Kostner, P. Beyerlein, and J. Cooper-White, One-step homogeneous C-reactive protein assay for saliva. J Immunol Methods, 2011. 373: p. 19-25. 149. Diem K and L. C., eds. Geigy Scientific Tables. 7 ed. 1970, Ciba-Geigy Limited: Basle. 643.

146

150. Hartmann, S., J.G. Okun, C. Schmidt, C.D. Langhans, S.F. Garbade, P. Burgard, D. Haas, J.O. Sass, W.L. Nyhan, and G.F. Hoffmann, Comprehensive detection of disorders of purine and pyrimidine metabolism by HPLC with electrospray ionization tandem mass spectrometry. Clin Chem, 2006. 52: p. 1127-1137. 151. de Weerth, C., J. Jansen, M.H. Vos, I. Maitimu, and E.G. Lentjes, A new device for collecting saliva for cortisol determination. Psychoneuroendocrinology, 2007. 32: p. 1144- 1148. 152. Feldman, H. and J.P. Guignard, Plasma creatinine in the first month of life. Arch Dis Child, 1982. 57: p. 123-126. 153. Guignard, J.P. and A. Drukker, Why do newborn infants have a high plasma creatinine? Pediatrics, 1999. 103: p. e49. 154. Francis, S.J., R.F. Walker, D. Riadfahmy, D. Hughes, J.F. Murphy, and O.P. Gray, Assessment of Adrenocortical Activity in Term Newborn-Infants Using Salivary Cortisol Determinations. J Pediatr, 1987. 111: p. 129-133. 155. Henkin, R.I., I. Velicu, and A. Papathanassiu, cAMP and cGMP in human parotid saliva: relationships to taste and smell dysfunction, gender, and age. Am J Med Sci, 2007. 334: p. 431-440. 156. Tanaka, S., M. Machino, S. Akita, Y. Yokote, and H. Sakagami, Changes in salivary amino acid composition during aging. In Vivo, 2010. 24: p. 853-856. 157. Ben-Aryeh, H., S. Lapid, R. Szargel, A. Benderly, and D. Gutman, Composition of whole unstimulated saliva of human infants. Arch Oral Biol, 1984. 29: p. 357-362. 158. Nelson, N., K. Arbring, and E. Theodorsson, Neonatal salivary cortisol in response to heelstick: method modifications enable analysis of low concentrations and small sample volumes. Scand J Clin Lab Invest, 2001. 61: p. 287-291. 159. Bagci, S., A. Mueller, J. Reinsberg, A. Heep, P. Bartmann, and A.R. Franz, Utility of salivary melatonin measurements in the assessment of the pineal physiology in newborn infants. Clin Biochem, 2010. 43: p. 868-872. 160. Bagci, S., A. Mueller, J. Reinsberg, A. Heep, P. Bartmann, and A.R. Franz, Saliva as a valid alternative in monitoring melatonin concentrations in newborn infants. Early Hum Dev, 2009. 85: p. 595-598. 161. Wan, A.K., W.K. Seow, D.M. Purdie, P.S. Bird, L.J. Walsh, and D.I. Tudehope, Immunoglobulins in saliva of preterm and full-term infants. Oral Microbiol Immunol, 2003. 18: p. 72-78. 162. Eells, J.T. and R. Spector, Purine and pyrimidine base and nucleoside concentrations in human cerebrospinal fluid and plasma. Neurochem Res, 1983. 8: p. 1451-1457. 147

163. O'Connor, M.C., R.A. Harkness, R.J. Simmonds, and F.E. Hytten, The measurement of hypoxanthine, xanthine, inosine and uridine in umbilical cord blood and fetal scalp blood samples as a measure of fetal hypoxia. Br J Obstet Gynaecol, 1981. 88: p. 381-390. 164. Maguire, M.H., I. Szabo, I.E. Valko, B.E. Finley, and T.L. Bennett, Simultaneous measurement of adenosine and hypoxanthine in human umbilical cord plasma using reversed-phase high-performance liquid chromatography with photodiode-array detection and on-line validation of peak purity. J Chromatogr B Biomed Sci Appl, 1998. 707: p. 33- 41. 165. Vidotto, C., D. Fousert, M. Akkermann, A. Griesmacher, and M.M. Muller, Purine and pyrimidine metabolites in children's urine. Clin Chim Acta, 2003. 335: p. 27-32. 166. Bratteby, L.E. and S. Swanstrom, Hypoxanthine concentration in plasma during the first two hours after birth in normal and asphyxiated infants. Pediatr Res, 1982. 16: p. 152-155. 167. Buonocore, G., S. Perrone, M. Longini, P. Vezzosi, B. Marzocchi, P. Paffetti, and R. Bracci, Oxidative stress in preterm neonates at birth and on the seventh day of life. Pediatr Res, 2002. 52: p. 46-49. 168. Harkness, R.A., R.T. Geirsson, and I.R. McFadyen, Concentrations of hypoxanthine, xanthine, uridine and urate in amniotic fluid at caesarean section and the association of raised levels with prenatal risk factors and fetal distress. Br J Obstet Gynaecol, 1983. 90: p. 815-820. 169. Anzai, N. and H. Endou, Urate transporters: an evolving field. Semin Nephrol, 2011. 31: p. 400-409. 170. Vitart, V., I. Rudan, C. Hayward, N.K. Gray, J. Floyd, C.N. Palmer, S.A. Knott, I. Kolcic, O. Polasek, J. Graessler, J.F. Wilson, A. Marinaki, P.L. Riches, X. Shu, B. Janicijevic, N. Smolej-Narancic, B. Gorgoni, J. Morgan, S. Campbell, Z. Biloglav, L. Barac-Lauc, M. Pericic, I.M. Klaric, L. Zgaga, T. Skaric-Juric, S.H. Wild, W.A. Richardson, P. Hohenstein, C.H. Kimber, A. Tenesa, L.A. Donnelly, L.D. Fairbanks, M. Aringer, P.M. McKeigue, S.H. Ralston, A.D. Morris, P. Rudan, N.D. Hastie, H. Campbell, and A.F. Wright, SLC2A9 is a newly identified urate transporter influencing serum urate concentration, urate excretion and gout. Nat Genet, 2008. 40: p. 437-442. 171. Hamilton, K., L. Daniels, N. Murray, K.M. White, and A. Walsh, Mothers' perceptions of introducing solids to their infant at six months of age: identifying critical belief-based targets to promote adherence to current infant feeding guidelines. J Health Psychol, 2012. 17: p. 121-131. 172. Motyl, T., J. Krzeminski, M. Podgurniak, C. Witeszczak, and P. Zochowski, Variability of orotic acid concentration in cow's milk. Endocr Regul, 1991. 25: p. 79-82. 148

173. Sitkovsky, M.V. and A. Ohta, The 'danger' sensors that STOP the immune response: the A2 adenosine receptors? Trends Immunol, 2005. 26: p. 299-304. 174. Huang, Z.L., Y. Urade, and O. Hayaishi, The role of adenosine in the regulation of sleep. Curr Top Med Chem, 2011. 11: p. 1047-1057. 175. Porkka-Heiskanen, T. and A.V. Kalinchuk, Adenosine, energy metabolism and sleep homeostasis. Sleep Med Rev, 2011. 15: p. 123-135. 176. Welk, A., P. Rudolph, J. Kreth, C. Schwahn, A. Kramer, and H. Below, Microbicidal efficacy of thiocyanate hydrogen peroxide after adding lactoperoxidase under saliva loading in the quantitative suspension test. Arch Oral Biol, 2011. 56: p. 1576-1582. 177. Uauy, R., G. Stringel, R. Thomas, and R. Quan, Effect of dietary nucleosides on growth and maturation of the developing gut in the rat. J Pediatr Gastroenterol Nutr, 1990. 10: p. 497- 503. 178. Abadeh, S., J. Killacky, M. Benboubetra, and R. Harrison, Purification and partial characterization of xanthine oxidase from human milk. Biochim Biophys Acta, 1992. 1117: p. 25-32. 179. Vorbach, C., A. Scriven, and M.R. Capecchi, The housekeeping gene xanthine oxidoreductase is necessary for milk fat droplet enveloping and secretion: gene sharing in the lactating mammary gland. Genes Dev, 2002. 16: p. 3223-3235. 180. Zhou, M., Z. Diwu, N. Panchuk-Voloshina, and R.P. Haugland, A stable nonfluorescent derivative of resorufin for the fluorometric determination of trace hydrogen peroxide: applications in detecting the activity of phagocyte NADPH oxidase and other oxidases. Anal Biochem, 1997. 253: p. 162-168. 181. Zikakis, J.P., T.M. Dougherty, and N.O. Biasotto, The presence and some properties of xanthine oxidase in human milk and colostrum. J Food Sci, 1976. 41: p. 1408-1412. 182. Harkness, R.A., G.M. McCreanor, D. Simpson, and I.R. MacFadyen, Pregnancy in and incidence of xanthine oxidase deficiency. J Inherit Metab Dis, 1986. 9: p. 407-408. 183. Kashuba, A.D., J.S. Bertino, Jr., G.L. Kearns, J.S. Leeder, A.W. James, R. Gotschall, and A.N. Nafziger, Quantitation of three-month intraindividual variability and influence of sex and menstrual cycle phase on CYP1A2, N-acetyltransferase-2, and xanthine oxidase activity determined with caffeine phenotyping. Clin Pharmacol Ther, 1998. 63: p. 540-551. 184. Kalow, W., Variability of caffeine metabolism in humans. Arzneimittelforschung, 1985. 35: p. 319-324. 185. Benboubetra, M., A. Baghiani, D. Atmani, and R. Harrison, Physicochemical and kinetic properties of purified sheep's milk xanthine oxidoreductase. J Dairy Sci, 2004. 87: p. 1580- 1584. 149

186. Graham, S., R.O. Day, H. Wong, A.J. McLachlan, L. Bergendal, J.O. Miners, and D.J. Birkett, Pharmacodynamics of oxypurinol after administration of allopurinol to healthy subjects. Br J Clin Pharmacol, 1996. 41: p. 299-304. 187. Heegaard, C.W., L.B. Larsen, L.K. Rasmussen, K.E. Hojberg, T.E. Petersen, and P.A. Andreasen, Plasminogen activation system in human milk. J Pediatr Gastroenterol Nutr, 1997. 25: p. 159-166. 188. Lindberg, T., K. Ohlsson, and B. Westrom, Protease inhibitors and their relation to protease activity in human milk. Pediatr Res, 1982. 16: p. 479-483. 189. Silanikove, N., F. Shapiro, A. Shamay, and G. Leitner, Role of xanthine oxidase, lactoperoxidase, and NO in the innate immune system of mammary secretion during active involution in dairy cows: manipulation with casein hydrolyzates. Free Radic Biol Med, 2005. 38: p. 1139-1151. 190. Akinbi, H., J. Meinzen-Derr, C. Auer, Y. Ma, D. Pullum, R. Kusano, K.J. Reszka, and K. Zimmerly, Alterations in the host defense properties of human milk following prolonged storage or pasteurization. J Pediatr Gastroenterol Nutr, 2010. 51: p. 347-352. 191. Tully, D.B., F. Jones, and M.R. Tully, Donor milk: what's in it and what's not. J Hum Lact, 2001. 17: p. 152-155. 192. Van Gysel, M., V. Cossey, S. Fieuws, and A. Schuermans, Impact of pasteurization on the antibacterial properties of human milk. Eur J Pediatr, 2012. 171: p. 1231-1237. 193. Silvestre, D., P. Ruiz, C. Martinez-Costa, A. Plaza, and M.C. Lopez, Effect of pasteurization on the bactericidal capacity of human milk. J Hum Lact, 2008. 24: p. 371-376. 194. Al-Kerwi, E.A., A.H. Al-Hashimi, and A.M. Salman, Mother's milk and hydrogen peroxide. Asia Pac J Clin Nutr, 2005. 14: p. 428-431. 195. Delgado, S., P. Garcia, L. Fernandez, E. Jimenez, M. Rodriguez-Banos, R. del Campo, and J.M. Rodriguez, Characterization of Staphylococcus aureus strains involved in human and bovine mastitis. FEMS Immunol Med Microbiol, 2011. 62: p. 225-235. 196. Shin, K., H. Hayasawa, and B. Lonnerdal, Purification and quantification of lactoperoxidase in human milk with use of immunoadsorbents with antibodies against recombinant human lactoperoxidase. Am J Clin Nutr, 2001. 73: p. 984-989. 197. Pan, Q., W.Y. Qiu, Y.N. Huo, Y.F. Yao, and M.F. Lou, Low levels of hydrogen peroxide stimulate corneal epithelial cell adhesion, migration, and wound healing. Invest Ophthalmol Vis Sci, 2011. 52: p. 1723-1734. 198. Roth, S. and W. Droge, Regulation of T-cell activation and T-cell growth factor (TCGF) production by hydrogen peroxide. Cell Immunol, 1987. 108: p. 417-424.

150

199. Storz, G., L.A. Tartaglia, and B.N. Ames, Transcriptional regulator of oxidative stress- inducible genes: direct activation by oxidation. Science, 1990. 248: p. 189-194. 200. Schreck, R., P. Rieber, and P.A. Baeuerle, Reactive oxygen intermediates as apparently widely used messengers in the activation of the NF-kappa B transcription factor and HIV-1. EMBO J, 1991. 10: p. 2247-2258. 201. Ignarro, L.J. and P.J. Kadowitz, The pharmacological and physiological role of cyclic GMP in vascular smooth muscle relaxation. Annu Rev Pharmacol Toxicol, 1985. 25: p. 171-191. 202. Radomski, M.W., R.M. Palmer, and S. Moncada, The anti-aggregating properties of vascular endothelium: interactions between prostacyclin and nitric oxide. Br J Pharmacol, 1987. 92: p. 639-646. 203. Aycicek, A., O. Erel, A. Kocyigit, S. Selek, and M.R. Demirkol, Breast milk provides better power than does formula. Nutr, 2006. 22: p. 616-619. 204. Gothefors, L. and S. Marklund, Lactoperoxidase activity in human milk and in saliva of newborn infants. Infect Immun, 1975. 11: p. 1210-1215. 205. El-Chemaly, S., M. Salathe, S. Baier, G.E. Conner, and R. Forteza, Hydrogen peroxide- scavenging properties of normal human airway secretions. Am J Respir Crit Care Med, 2003. 167: p. 425-430. 206. Hyyppa, T., L. Karhuvaara, J. Tenovuo, M. Lumikari, and P. Vilja, Antimicrobial factors in whole saliva of human infants: a longitudinal study. Pediatr Dent, 1989. 11: p. 30-36. 207. Tenovuo, J., O.P. Lehtonen, A.S. Aaltonen, P. Vilja, and P. Tuohimaa, Antimicrobial factors in whole saliva of human infants. Infect Immun, 1986. 51: p. 49-53. 208. Reiter, B., R.J. Fulford, V.M. Marshall, N. Yarrow, M.J. Ducker, and M. Knutsson, An evaluation of the growth promoting effect of the lactoperoxidase system in newborn calves. Anim Sci, 1981. 32: p. 297-306. 209. Millar, T.M., J.M. Kanczler, T. Bodamyali, D.R. Blake, and C.R. Stevens, Xanthine oxidase is a peroxynitrite synthase: newly identified roles for a very old enzyme. Redox Rep, 2002. 7: p. 65-70. 210. Matamoros, S., C. Gras-Leguen, F. Le Vacon, G. Potel, and M.F. de La Cochetiere, Development of intestinal microbiota in infants and its impact on health. Trends Microbiol, 2013. 21: p. 167-173. 211. 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. Le 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. 151

Dore, F. Guarner, K. Kristiansen, O. Pedersen, J. Parkhill, J. Weissenbach, P. Bork, S.D. Ehrlich, and J. Wang, A human gut microbial gene catalogue established by metagenomic sequencing. Nature, 2010. 464: p. 59-65. 212. Arumugam, M., J. Raes, E. Pelletier, D. Le 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. de Vos, S. Brunak, J. Dore, M. Antolin, 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. van de Guchte, E. Guedon, F. Haimet, W. Huber, J. van Hylckama-Vlieg, A. Jamet, C. Juste, G. Kaci, J. Knol, O. Lakhdari, S. Layec, K. Le Roux, E. Maguin, A. Merieux, R. Melo 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, Enterotypes of the human gut microbiome. Nature, 2011. 473: p. 174-180. 213. Kamada, N., G.Y. Chen, N. Inohara, and G. Nunez, Control of pathogens and pathobionts by the gut microbiota. Nat Immunol, 2013. 14: p. 685-690. 214. Maslowski, K.M. and C.R. Mackay, Diet, gut microbiota and immune responses. Nat Immunol, 2011. 12: p. 5-9. 215. Rudin, A. and A.C. Lundell, Infant B cell memory and gut bacterial colonization. Gut Microbes, 2012. 3: p. 474-475. 216. Sommer, F. and F. Backhed, The gut microbiota--masters of host development and physiology. Nat Rev Microbiol, 2013. 11: p. 227-238. 217. Ramakrishna, B.S., Role of the gut microbiota in and metabolism. J Gastroenterol Hepatol, 2013. 28 Suppl 4: p. 9-17. 218. Compare, D. and G. Nardone, Contribution of gut microbiota to colonic and extracolonic cancer development. Dig Dis, 2011. 29: p. 554-561. 219. Sobhani, I., A. Amiot, Y. Le Baleur, M. Levy, M.L. Auriault, J.T. Van Nhieu, and J.C. Delchier, Microbial dysbiosis and colon carcinogenesis: could colon cancer be considered a bacteria-related disease? Therap Adv Gastroenterol, 2013. 6: p. 215-229. 220. Duboc, H., S. Rajca, D. Rainteau, D. Benarous, M.A. Maubert, E. Quervain, G. Thomas, V. Barbu, L. Humbert, G. Despras, C. Bridonneau, F. Dumetz, J.P. Grill, J. Masliah, L. Beaugerie, J. Cosnes, O. Chazouilleres, R. Poupon, C. Wolf, J.M. Mallet, P. Langella, G.

152

Trugnan, H. Sokol, and P. Seksik, Connecting dysbiosis, bile-acid dysmetabolism and gut inflammation in inflammatory bowel diseases. Gut, 2013. 62: p. 531-539. 221. Bien, J., V. Palagani, and P. Bozko, The intestinal microbiota dysbiosis and Clostridium difficile infection: is there a relationship with inflammatory bowel disease? Therap Adv Gastroenterol, 2013. 6: p. 53-68. 222. Azcarate-Peril, M.A., D.M. Foster, M.B. Cadenas, M.R. Stone, S.K. Jacobi, S.H. Stauffer, A. Pease, and J.L. Gookin, Acute necrotizing enterocolitis of preterm piglets is characterized by dysbiosis of ileal mucosa-associated bacteria. Gut Microbes, 2011. 2: p. 234-243. 223. Torrazza, R.M. and J. Neu, The altered gut microbiome and necrotizing enterocolitis. Clin Perinatol, 2013. 40: p. 93-108. 224. Isolauri, E., S. Rautava, and S. Salminen, Probiotics in the development and treatment of allergic disease. Gastroenterol Clin North Am, 2012. 41: p. 747-762. 225. Jonkers, D., J. Penders, A. Masclee, and M. Pierik, Probiotics in the management of inflammatory bowel disease: a systematic review of intervention studies in adult patients. Drugs, 2012. 72: p. 803-823. 226. Jia, W., H. Li, L. Zhao, and J.K. Nicholson, Gut microbiota: a potential new territory for drug targeting. Nat Rev Drug Discov, 2008. 7: p. 123-129. 227. Koenig, J.E., A. Spor, N. Scalfone, A.D. Fricker, J. Stombaugh, R. Knight, L.T. Angenent, and R.E. Ley, Succession of microbial consortia in the developing infant gut microbiome. Proc Natl Acad Sci U S A, 2011. 108 Suppl 1: p. 4578-4585. 228. Satokari, R., T. Gronroos, K. Laitinen, S. Salminen, and E. Isolauri, Bifidobacterium and Lactobacillus DNA in the human placenta. Lett Appl Microbiol, 2009. 48: p. 8-12. 229. Jimenez, E., M.L. Marin, R. Martin, J.M. Odriozola, M. Olivares, J. Xaus, L. Fernandez, and J.M. Rodriguez, Is meconium from healthy newborns actually sterile? Res Microbiol, 2008. 159: p. 187-193. 230. Gueimonde, M., S. Sakata, M. Kalliomaki, E. Isolauri, Y. Benno, and S. Salminen, Effect of maternal consumption of lactobacillus GG on transfer and establishment of fecal bifidobacterial microbiota in neonates. J Pediatr Gastroenterol Nutr, 2006. 42: p. 166-170. 231. Vaishampayan, P.A., J.V. Kuehl, J.L. Froula, J.L. Morgan, H. Ochman, and M.P. Francino, Comparative metagenomics and population dynamics of the gut microbiota in mother and infant. Genome Biol Evol, 2010. 2: p. 53-66. 232. Forno, E., A.B. Onderdonk, J. McCracken, A.A. Litonjua, D. Laskey, M.L. Delaney, A.M. Dubois, D.R. Gold, L.M. Ryan, S.T. Weiss, and J.C. Celedon, Diversity of the gut microbiota and eczema in early life. Clin Mol Allergy, 2008. 6: p. 11.

153

233. Abrahamsson, T.R., H.E. Jakobsson, A.F. Andersson, B. Bjorksten, L. Engstrand, and M.C. Jenmalm, Low diversity of the gut microbiota in infants with atopic eczema. J Allergy Clin Immunol, 2012. 129: p. 434-440. 234. Jakobsson, H.E., T.R. Abrahamsson, M.C. Jenmalm, K. Harris, C. Quince, C. Jernberg, B. Bjorksten, L. Engstrand, and A.F. Andersson, Decreased gut microbiota diversity, delayed Bacteroidetes colonisation and reduced Th1 responses in infants delivered by Caesarean section. Gut, 2013: p. aheadofprint. 235. Penders, J., C. Thijs, C. Vink, F.F. Stelma, B. Snijders, I. Kummeling, P.A. van den Brandt, and E.E. Stobberingh, Factors influencing the composition of the intestinal microbiota in early infancy. Pediatrics, 2006. 118: p. 511-521. 236. Turroni, F., C. Peano, D.A. Pass, E. Foroni, M. Severgnini, M.J. Claesson, C. Kerr, J. Hourihane, D. Murray, F. Fuligni, M. Gueimonde, A. Margolles, G. De Bellis, P.W. O'Toole, D. van Sinderen, J.R. Marchesi, and M. Ventura, Diversity of bifidobacteria within the infant gut microbiota. PLoS ONE, 2012. 7: p. e36957. 237. Roger, L.C., A. Costabile, D.T. Holland, L. Hoyles, and A.L. McCartney, Examination of faecal Bifidobacterium populations in breast- and formula-fed infants during the first 18 months of life. Microbiology, 2010. 156: p. 3329-3341. 238. Bertino, E., F. Giuliani, L. Occhi, A. Coscia, P. Tonetto, F. Marchino, and C. Fabris, Benefits of donor human milk for preterm infants: current evidence. Early Hum Dev, 2009. 85: p. S9-S10. 239. Kelley, L., Increasing the consumption of breast milk in low-birth-weight infants: can it have an impact on necrotizing enterocolitis? Adv Neonatal Care, 2012. 12: p. 267-272. 240. Neu, J., M. Chen, and E. Beierle, Intestinal innate immunity: how does it relate to the pathogenesis of necrotizing enterocolitis. Semin Pediatr Surg, 2005. 14: p. 137-144. 241. Hunter, C.J., J.S. Upperman, H.R. Ford, and V. Camerini, Understanding the susceptibility of the premature infant to necrotizing enterocolitis (NEC). Pediatr Res, 2008. 63: p. 117- 123. 242. Gephart, S.M., J.M. McGrath, J.A. Effken, and M.D. Halpern, Necrotizing enterocolitis risk: state of the science. Adv Neonatal Care, 2012. 12: p. 77-87. 243. Schurr, P. and E.M. Perkins, The relationship between feeding and necrotizing enterocolitis in very low birth weight infants. Neonatal Netw, 2008. 27: p. 397-407. 244. Meinzen-Derr, J., B. Poindexter, L. Wrage, A.L. Morrow, B. Stoll, and E.F. Donovan, Role of human milk in extremely low birth weight infants' risk of necrotizing enterocolitis or death. J Perinatol, 2009. 29: p. 57-62.

154

245. Firth, M.A., P.E. Shewen, and D.C. Hodgins, Passive and active components of neonatal innate immune defenses. Anim Health Res Rev, 2005. 6: p. 143-158. 246. Levy, O., Innate immunity of the newborn: basic mechanisms and clinical correlates. Nat Rev Immunol, 2007. 7: p. 379-390. 247. Maia, L.B. and J.J. Moura, Nitrite reduction by xanthine oxidase family enzymes: a new class of nitrite reductases. J Biol Inorg Chem, 2011. 16: p. 443-460. 248. Doel, J.J., B.L. Godber, R. Eisenthal, and R. Harrison, Reduction of organic nitrates catalysed by xanthine oxidoreductase under anaerobic conditions. Biochim Biophys Acta, 2001. 1527: p. 81-87. 249. Turner, S., K.M. Pryer, V.P. Miao, and J.D. Palmer, Investigating deep phylogenetic relationships among cyanobacteria and plastids by small subunit rRNA sequence analysis. J Eukaryot Microbiol, 1999. 46: p. 327-338. 250. 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, QIIME allows analysis of high-throughput community sequencing data. Nat Methods, 2010. 7: p. 335-336. 251. Wang, Q., G.M. Garrity, J.M. Tiedje, and J.R. Cole, Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol, 2007. 73: p. 5261-5267. 252. Marodi, L., Neonatal innate immunity to infectious agents. Infect Immun, 2006. 74: p. 1999- 2006. 253. Green, D.E. and R. Pauli, The Anti-Bacterial Action of the Xanthine Oxidase System. Proc Soc Exp Biol Med, 1943. 54: p. 148-150. 254. Lipmann, F. and C.R. Owen, The Antibacterial Effect of Enzymatic Xanthine Oxidation. Science, 1943. 98: p. 246-248. 255. Björck, L. and O. Claesson, Xanthine Oxidase as a Source of Hydrogen Peroxide for the Lactoperoxidase System in Milk. J Dairy Sci, 1979. 62: p. 1211-1215. 256. Chua, K.Y., T.P. Stinear, and B.P. Howden, Functional genomics of Staphylococcus aureus. Brief Funct Genomics, 2013: p. 305-315. 257. Denniston, S. and F.A. Riordan, Staphylococcus aureus bacteraemia in children and neonates: a 10 year retrospective review. J Infect, 2006. 53: p. 387-393. 258. Carey, A.J. and S.S. Long, Staphylococcus aureus: a continuously evolving and formidable pathogen in the neonatal intensive care unit. Clin Perinatol, 2010. 37: p. 535-546. 155

259. Ganatra, H.A. and A.K. Zaidi, Neonatal infections in the developing world. Semin Perinatol, 2010. 34: p. 416-425. 260. Boesten, R.J. and W.M. de Vos, Interactomics in the human intestine: Lactobacilli and Bifidobacteria make a difference. J Clin Gastroenterol, 2008. 42 Suppl 3 Pt 2: p. S163-167. 261. Matsumiya, Y., N. Kato, K. Watanabe, and H. Kato, Molecular epidemiological study of vertical transmission of vaginal Lactobacillus species from mothers to newborn infants in Japanese, by arbitrarily primed polymerase chain reaction. J Infect Chemother, 2002. 8: p. 43-49. 262. Fujimura, S., A. Watanabe, K. Kimura, and M. Kaji, Probiotic mechanism of Lactobacillus gasseri OLL2716 strain against Helicobacter pylori. J Clin Microbiol, 2012. 50: p. 1134- 1136. 263. Stockinger, S., M.W. Hornef, and C. Chassin, Establishment of intestinal homeostasis during the neonatal period. Cell Mol Life Sci, 2011. 68: p. 3699-3712. 264. Johnson, C.L. and J. Versalovic, The human microbiome and its potential importance to pediatrics. Pediatrics, 2012. 129: p. 950-960. 265. Wijkstrom-Frei, C., S. El-Chemaly, R. Ali-Rachedi, C. Gerson, M.A. Cobas, R. Forteza, M. Salathe, and G.E. Conner, Lactoperoxidase and human airway host defense. Am J Respir Cell Mol Biol, 2003. 29: p. 206-212. 266. Anthimidou, E. and D. Mossialos, Antibacterial Activity of Greek and Cypriot Honeys Against Staphylococcus aureus and Pseudomonas aeruginosa in Comparison to Manuka Honey. J Med Food, 2013. 16: p. 42-47. 267. Oelschlaegel, S., L. Pieper, R. Staufenbiel, M. Gruner, L. Zeippert, B. Pieper, I. Koelling- Speer, and K. Speer, Floral markers of cornflower (Centaurea cyanus) honey and its peroxide antibacterial activity for an alternative treatment of digital dermatitis. J Agric Food Chem, 2012. 60: p. 11811-11820. 268. Brudzynski, K., K. Abubaker, and T. Wang, Powerful bacterial killing by buckwheat honeys is concentration-dependent, involves complete DNA degradation and requires hydrogen peroxide. Front Microbiol, 2012. 3: p. 242. 269. Bang, L.M., C. Buntting, and P. Molan, The effect of dilution on the rate of hydrogen peroxide production in honey and its implications for wound healing. J Altern Complement Med, 2003. 9: p. 267-273. 270. Lin, P.W. and B.J. Stoll, Necrotising enterocolitis. Lancet, 2006. 368: p. 1271-1283. 271. Sisk, P.M., C.A. Lovelady, R.G. Dillard, K.J. Gruber, and T.M. O'Shea, Early human milk feeding is associated with a lower risk of necrotizing enterocolitis in very low birth weight infants. J Perinatol, 2007. 27: p. 428-433. 156

272. Blaut, M. and T. Clavel, Metabolic diversity of the intestinal microbiota: implications for health and disease. J Nutr, 2007. 137: p. 751S-755S. 273. Turnbaugh, P.J., R.E. Ley, M.A. Mahowald, V. Magrini, E.R. Mardis, and J.I. Gordon, An obesity-associated gut microbiome with increased capacity for energy harvest. Nature, 2006. 444: p. 1027-1031. 274. Martin, V., A. Maldonado-Barragan, L. Moles, M. Rodriguez-Banos, R.D. Campo, L. Fernandez, J.M. Rodriguez, and E. Jimenez, Sharing of bacterial strains between breast milk and infant feces. J Hum Lact, 2012. 28: p. 36-44. 275. Palmer, C., E.M. Bik, D.B. DiGiulio, D.A. Relman, and P.O. Brown, Development of the human infant intestinal microbiota. PLoS Biol, 2007. 5: p. e177. 276. Jost, T., C. Lacroix, C.P. Braegger, and C. Chassard, New insights in gut microbiota establishment in healthy breast fed neonates. PLoS ONE, 2012. 7: p. e44595. 277. Rinne, M., M. Kalliomaki, H. Arvilommi, S. Salminen, and E. Isolauri, Effect of probiotics and breastfeeding on the bifidobacterium and lactobacillus/enterococcus microbiota and humoral immune responses. J Pediatr, 2005. 147: p. 186-191. 278. Watson, J., R. Jones, C. Cortes, S. Gerber, R. Golash, J. Price, E. Bancroft, L. Mascola, R. Gorwitz, and D. Jernigan, Community-associated methicillin-resistant Staphylococcus aureus infection among healthy newborns–Chicago and Los Angeles County, 2004. JAMA, 2006. 296: p. 36-38. 279. Wang, Y., J.D. Hoenig, K.J. Malin, S. Qamar, E.O. Petrof, J. Sun, D.A. Antonopoulos, E.B. Chang, and E.C. Claud, 16S rRNA gene-based analysis of fecal microbiota from preterm infants with and without necrotizing enterocolitis. ISME J, 2009. 3: p. 944-954. 280. 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, Diversity of the human intestinal microbial flora. Science, 2005. 308: p. 1635-1638.

157

Appendices

158

Appendix 1 Mater Health Services Human Research Ethics Committee Approval.

159

160

Appendix 2 University of Queensland Human Research Ethics Committee Approval.

161

Appendix 3 Mater Health Services Human Research Ethics Committee Approval.

162

163

Appendix 4 University of Queensland Human Research Ethics Committee Approval.

164

Appendix 5 Mater Health Services Human Research Ethics Committee Approval.

165

166

Appendix 6 University of Queensland Human Research Ethics Committee Approval.

167

Appendix 7 Golden Casket research fund approval.

168

169

Appendix 8 NHMRC research fund approval.

170

171

Appendix 9 Paper published in J Chromatogr B, 2013. 931C: p. 140-147.

172

173

174

175

176

177

178

179

Appendix 10 Abstract Published in Journal of Inherited Metabolic Disease.

Al-Shehri S, Shaw PN, Cowley D, Liley H, Henman M, Tomarchio A, Charles B, Duley JA (2011) Nucleotide precursors in neonatal and adult saliva. J. Inherit. Metab. Dis. 34:S73. Poster was also presented at SSIEM, 2011, Geneva.

NUCLEOTIDE PRECURSORS IN NEONATAL AND ADULT SALIVA

Al-Shehri S1*, Shaw PN1, Cowley D2, Liley H2, Henman M2, Tomarchio A2, Charles B1, Duley JA1,2

1 School of Pharmacy, The University of Queensland, Brisbane 4072, Australia; 2Mater Medical Research Institute, Brisbane 4101, Australia.

Background: Human saliva has been extensively described for its composition of major proteins, electrolytes, cortisol, melatonin and some metabolites such as amino acids. Little is known, however, about nucleotide precursors in human saliva. Nucleotides are essential for synthesis of DNA/RNA, supplying energy, regulating G-protein signalling, and biosynthesis. Methods: Saliva samples were collected from full-term neonates, aged 1-3 days, using cotton swabs. Unstimulated fasting (morning) saliva samples were collected directly from adults. The samples were extracted and ultrafiltered, then nucleotide precursors were analysed by reversed-phase HPLC with UV-detection and mass spectrometry (with stable- isotope internal standards). Results: Concentrations of salivary nucleobases and nucleosides as µM±SEM for Neonates\Adults respectively were: Pseudouridine 2±0.4\0.27±0.07, Uracil 15±4\4.4±1.5, Thymine 0.2±0.1\0.35±0.16, Dihydrouracil 8±1\1.1±0.21, Dihydrothymine 2±02\0.43±0.1, Uridine 19±4\1.7±0.7, Urate 157±23\203±24, Hypoxanthine 41±9\3.8±1.5, Xanthine 29±7\4.9±2.3, Adenosine 13±3\0.04±0.02, Inosine 22±7\0.43±0.13, Guanosine 11±3\0.22±0.05, Succinyladenosine 6±2\0.63±0.25. Deoxynucleosides and dihydroxy-adenine were negligible. Discussion: Salivary concentrations of purine nucleotide precursors such as hypoxanthine, xanthine, adenosine, inosine and guanosine are surprisingly high in neonates and much higher than in plasma. Transition to lower adult levels appears to occur during the first year. These precursors in saliva may be useful biomarkers for diagnosis of inborn errors of nucleotide metabolism and the investigation of some pharmacogenetic disorders.

180

Appendix 11 A poster presented at SSIEM, 2011, Geneva.

181

Appendix 12 An abstract and a poster presented at ASBMB ComBio2012

Longitudinal‎study‎of‎nucleotide‎metabolites‎in‎infants’‎saliva‎compared‎to‎adults‎and‎mammals‎

Al-Shehri S1*, Henman M2, Cowley D2, Charles B1, Shaw PN1, Liley H2, Duley JA1,2

1 School of Pharmacy, The University of Queensland, Brisbane 4072, Australia; 2Mater Medical Research Institute & Health Services, Brisbane 4101, Australia

Background: Human saliva has been extensively described for its composition of major proteins, electrolytes, cortisol, melatonin and some metabolites such as amino acids. Little is known, however, about nucleotide precursors in human saliva. Nucleotides are essential for synthesis of DNA/RNA, supplying energy, regulating G-protein signalling, and biosyntheses. Methods: Saliva swabs were collected from full-term neonates, aged 1-4 days, 6-weeks, 6-months and 12- months. Unstimulated fasting (morning) saliva samples were collected directly from adults. Samples were extracted and ultrafiltered, then nucleotide metabolites were analysed by RP HPLC with UV-photodiode array and ESI-QTrap mass spectrometry. Results: Median concentrations (µM) of salivary nucleobases and nucleosides in infants for neonates/6-weeks/6- months/12-months respectively were: uracil 5.3/0.80/1.4/0.70, hypoxanthine 27/7.0/1.1/0.80, xanthine 19/7.0/2.0/2.0, adenosine 12/7.0/0.90/0.80, inosine 11/5.0/0.30/0.40, guanosine 7.0/6.0/0.50/0.40, uridine 12/0.80/0.30/0.90. Deoxynucleosides, dihydropyrimidines and dihydroxyadenine concentrations were essentially negligible. The concentrations in human adults/cows/horses/dogs respectively were: uracil 0.80/3.8/81/12, hypoxanthine 2.0/0.20/7.0/0.20, xanthine 2.0/0.40/34/1.2, adenosine 0.10/0.20/0.10/19, inosine 0.20/0.50/9.4/3.4, guanosine 0.10/0.30/10/30, uridine 0.40/0.40/27/4.0. Salivary deoxyuridine (13µM), deoxyinosine (10µM), deoxyguanosine (7µM), and thymidine (18µM) were unusually high in horses. Saliva from cats, goats, sheep and camel were also assayed. Discussion: Salivary concentrations of purine nucleotide precursors such as hypoxanthine, xanthine, adenosine, inosine and guanosine were surprisingly high in neonates and higher than in plasma. Transition from human neonate to adult levels occurred during the first year. These precursors in saliva may be useful biomarkers for diagnosis of inborn errors of nucleotide metabolism and the investigation of some pharmacogenetic disorders.

182

Appendix 13 A poster presented at ASBMB ComBio, 2012, Adelaide.

183

Appendix 14 Abstract Published Journal of Paediatrics and Child Health

Al-Shehri S, Liley HG, Knox CL, Henman M, Cowley DM, Charles BG, Duley JA. Human milk xanthine oxidase and neonatal salivary nucleotide precursors generate hydrogen peroxide; a novel pathway with potential role in regulating oral micro flora. J Paediatrics Child Health 2013;49 (S2), (Suppl 2), p.121

Human milk xanthine oxidase and neonatal salivary nucleotide precursors generate hydrogen peroxide; a novel pathway with potential role in regulating oral microflora.

Al-Shehri SS1, Liley HG2, Knox CL3, Henman M2, Cowley DM2, Charles BG1, Duley JA1,2

1School of Pharmacy, University of Queensland, Brisbane, Australia,2Mater Medical Research Institute and Mater Health Services, Brisbane, Australia,3 Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, Australia

Email: [email protected]

Background: Xanthine oxidase (XO) is a molybdeno-flavoprotein occurring with high activity in the milk fat globule membrane and is involved in the final stage of degradation of purine nucleotides. It catalyzes the sequential oxidation of hypoxanthine to xanthine and uric acid, accompanied by production of hydrogen peroxide and superoxide anion. Human saliva has been extensively described for its composition of proteins, electrolytes, cortisol, melatonin and some metabolites such as amino acids, but little is known about nucleotide metabolites. Method: Saliva was collected with swabs from 60 babies at full-term 1-4 days, 6-weeks, 6-months and 12-months. Unstimulated fasting (morning) saliva samples were collected directly from 77 adults. Breast milk was collected from 24 new mothers. Saliva was extracted from swabs and ultra-filtered. Nucleotide metabolites were analyzed by RP- HPLC with UV-photodiode array and ESI-MS/MS mass spectrometry. XO activity was measured as peroxide production from hypoxanthine. Bacterial inhibition over time was assessed using CFU/mL or OD. Results: Median concentrations (μmol/L) of salivary nucleobases and nucleosides for neonates/6-weeks/6-months/12- months/adult respectively were: uracil 5.3/0.8/1.4/0.7/0.8, hypoxanthine 27/7.0/1.1/0.8/2.0, xanthine 19/7.0/2.0/2.0/2.0, adenosine 12/7.0/0.9/0.8/0.1, inosine 11/5.0/0.3/0.4/0.2, guanosine 7.0/6.0/0.5/0.4/0.1, uridine 12/0.8/0.3/0.9/0.4. Deoxynucleosides and dihydropyrimidines concentrations were essentially negligible. XO activity (Vmax:mean±SD) in breast milk was 8.9±6.2 μmol/min/L and endogenous peroxide was 27±12 μmol/L; mixing breast milk with neonate saliva generated ~40 μmol/L peroxide, which inhibited Staphylococcus aureus. Conclusions: Salivary metabolites, particularly xanthine/hypoxanthine, are high in neonates, transitioning to low adult levels between 6-weeks to 6-months (P<0.001, Mann-Whitney). Peroxide occurs in breast milk and is boosted during suckling as an antibacterial system.

184

Appendix 15 A poster presented at PSANZ, 2013, Adelaide.

185

Appendix 16 Raw data set of adult salivary nucleotide precursors concentrations (µM). Subject Creatinine Pseudouracil Uracil Hypoxanthine Xanthine Adenine Adenosine Inosine Guanosine Dihydrothymine Dihydrouracil Uridine Thymine Orotate Urate 1 <5 <0.2 <1.5 0.9 0.8 0.1 <0.03 0.3 0.3 <1.5 <3 <0.02 <1.5 <0.5 224.0 2 <5 <0.2 1.6 2.5 1.1 0.8 <0.03 0.3 0.1 <1.5 2.5 0.4 <1.5 <0.5 153.0 3 <5 0.2 <1.5 3.1 2.6 0.4 0.0 0.4 0.3 <1.5 <3 0.8 <1.5 0.5 271.0 4 <5 <0.2 4.5 0.7 <0.7 0.1 <0.03 0.1 0.1 <1.5 <3 0.5 <1.5 <0.5 146.0 5 <5 0.6 20.0 10.0 12.3 1.3 0.2 1.0 0.4 <1.5 <3 2.9 7.2 1.0 211.0 6 <5 <0.2 <1.5 0.3 <0.7 0.1 <0.03 <0.06 0.1 <1.5 <3 <0.02 <1.5 <0.5 131.0 7 5.6 1.0 10.9 22.0 32.7 1.3 <0.03 1.9 0.7 <1.5 <3 2.4 8.4 <0.5 396.0 8 <5 <0.2 <1.5 1.1 0.8 0.2 <0.03 0.2 0.2 <1.5 <3 0.4 <1.5 <0.5 135.0 9 6.9 <0.2 8.5 3.4 4.5 1.0 <0.03 <0.06 0.1 <1.5 <3 0.0 <1.5 2.2 162.0 10 8.6 0.4 7.4 2.3 5.6 0.9 0.2 0.6 0.1 <1.5 <3 1.5 2.3 <0.5 294.0 11 <5 0.4 5.3 3.5 6.3 1.1 0.1 0.6 0.1 <1.5 <3 <0.02 2.0 0.5 91.0 12 <5 <0.2 <1.5 1.3 <0.7 0.1 <0.03 0.2 0.2 <1.5 <3 <0.02 <1.5 <0.5 201.0 13 <5 0.2 <1.5 1.3 1.1 0.2 <0.03 0.3 0.3 <1.5 <3 <0.02 <1.5 <0.5 105.0 14 8.4 0.2 <1.5 0.6 0.7 0.2 0.1 0.1 0.2 <1.5 <3 0.1 <1.5 0.5 318.0 15 <5 0.3 <1.5 5.2 2.4 1.0 0.1 0.2 0.2 <1.5 <3 0.1 <1.5 0.8 184.0 16 11.5 0.6 8.0 10.3 15.8 1.5 <0.03 0.4 0.2 <1.5 <3 <0.02 1.9 0.9 248.0 17 8.8 0.2 1.5 3.4 3.7 0.5 <0.03 0.3 0.2 <1.5 <3 <0.02 <1.5 <0.5 214.0 18 6.9 <0.2 <1.5 1.2 <0.7 0.1 <0.03 0.1 0.1 <1.5 <3 <0.02 <1.5 <0.5 212.0 19 <5 <0.2 <1.5 0.6 <0.7 0.2 <0.03 <0.06 <0.03 <1.5 <3 <0.02 <1.5 <0.5 162.0 20 6.5 0.4 2.1 3.0 9.6 0.9 <0.03 0.3 0.2 <1.5 <3 8.4 1.7 0.8 287.0 21 5.9 0.2 <1.5 1.8 2.0 0.6 0.1 0.2 0.2 <1.5 <3 <0.02 <1.5 <0.5 109.0 22 6.1 0.2 <1.5 1.0 1.7 0.9 0.7 0.3 0.2 <1.5 <3 <0.02 <1.5 0.8 142.0 23 5.8 0.2 <1.5 0.7 1.1 0.2 <0.03 0.1 0.1 <1.5 <3 0.1 <1.5 <0.5 178.0 24 7.7 <0.2 <1.5 10.4 13.0 3.1 0.5 0.9 0.5 <1.5 <3 0.2 <1.5 0.5 177.0 25 <5 0.2 <1.5 1.2 1.7 0.8 <0.03 <0.06 0.1 <1.5 <3 0.4 <1.5 <0.5 98.7 26 <5 <0.2 <1.5 0.4 <0.7 0.2 <0.03 <0.06 <0.03 <1.5 <3 0.4 <1.5 <0.5 78.8 27 9.8 0.2 <1.5 1.9 1.6 0.7 <0.03 <0.06 <0.03 <1.5 <3 0.4 <1.5 <0.5 162.0 28 5.2 <0.2 <1.5 <0.2 <0.7 <0.1 <0.03 <0.06 <0.03 <1.5 <3 <0.02 <1.5 <0.5 92.7 29 <5 <0.2 <1.5 2.5 2.0 1.3 0.1 0.2 0.1 <1.5 <3 0.0 <1.5 0.5 74.7 30 <5 0.2 <1.5 5.4 1.4 0.8 0.4 1.0 0.3 <1.5 <3 0.0 <1.5 <0.5 95.8

186

Appendix 16 (continued) Subject Creatinine Pseudouracil Uracil Hypoxanthine Xanthine Adenine Adenosine Inosine Guanosine Dihydrothymine Dihydrouracil Uridine Thymine Orotate Urate 31 8.5 0.2 <1.5 2.2 2.2 0.4 <0.03 0.1 0.1 <1.5 <3 <0.02 <1.5 <0.5 170.0 32 7.6 <0.2 <1.5 0.9 <0.7 0.6 <0.03 <0.06 0.0 <1.5 <3 0.1 <1.5 0.6 39.0 33 5.1 <0.2 <1.5 0.6 0.7 0.8 0.2 <0.06 0.1 <1.5 <3 0.1 <1.5 0.5 195.0 34 7.1 <0.2 <1.5 0.8 2.8 0.5 0.1 0.1 0.2 <1.5 <3 0.5 <1.5 0.5 105.0 35 6.6 0.3 3.3 6.9 7.8 2.3 <0.03 0.2 0.1 <1.5 <3 0.6 <1.5 1.6 197.0 36 <5 <0.2 <1.5 0.6 <0.7 0.6 0.1 0.1 0.0 <1.5 <3 0.1 <1.5 <0.5 89.5 37 <5 0.4 <1.5 3.9 3.6 0.6 0.2 0.3 0.2 <1.5 <3 0.9 <1.5 0.7 246.0 38 <5 0.2 <1.5 0.7 <0.7 0.5 <0.03 0.2 <0.03 <1.5 <3 0.2 <1.5 0.9 112.0 39 <5 0.3 <1.5 2.8 1.0 0.7 0.3 51.0 0.3 <1.5 <3 0.1 <1.5 0.9 89.5 40 <5 0.2 <1.5 3.6 0.9 0.8 0.1 0.1 0.1 <1.5 <3 0.7 <1.5 0.9 242.0 41 <5 <0.2 7.7 3.2 2.1 2.2 0.4 0.4 0.2 <1.5 <3 0.4 <1.5 1.2 107.0 42 <5 <0.2 9.2 5.4 6.2 3.7 0.3 0.2 0.2 <1.5 <3 0.3 <1.5 2.7 176.0 43 <5 0.2 <1.5 2.0 1.5 1.6 0.4 0.3 0.2 <1.5 <3 0.8 <1.5 1.0 240.0 44 < <0.2 <1.5 0.3 1.2 1.2 0.4 0.2 0.3 <1.5 3.4 0.2 <1.5 0.8 84.1 45 <5 0.2 1.6 2.3 1.1 2.1 0.3 0.3 0.2 <1.5 <3 0.9 <1.5 1.0 59.7 46 <5 0.2 <1.5 2.1 0.9 1.2 0.1 0.2 0.1 <1.5 <3 0.2 <1.5 0.7 222.0 47 <5 0.3 <1.5 1.7 1.3 0.8 0.3 0.4 0.4 <1.5 <3 0.4 <1.5 1.1 140.0 48 5.8 0.4 18.3 1.9 16.5 0.8 <0.03 <0.06 <0.03 <1.5 <3 <0.02 <1.5 <0.5 102.0 49 <5 0.3 <1.5 0.7 1.1 1.0 0.1 0.1 0.1 <1.5 <3 0.2 <1.5 1.8 131.0 50 <5 0.3 1.6 5.2 4.8 1.4 0.4 0.5 0.3 <1.5 <3 0.6 <1.5 0.7 176.0 51 6.6 <0.2 1.7 2.4 2.0 0.4 0.5 0.7 0.3 <1.5 <3 0.0 1.8 1.4 101.0 52 8.2 <0.2 3.0 2.1 0.9 1.0 0.2 0.1 0.2 <1.5 <3 0.4 <1.5 1.1 212.0 53 <5 <0.2 <1.5 1.0 <0.7 0.5 0.1 0.1 0.1 <1.5 <3 0.2 <1.5 1.0 168.0 54 <5 0.5 <1.5 6.8 3.5 1.2 0.3 0.8 0.2 <1.5 <3 0.4 <1.5 1.4 284.0 55 <5 <0.2 <1.5 1.5 <0.7 0.5 0.3 0.4 0.2 <1.5 <3 0.2 <1.5 0.9 178.0 56 <5 <0.2 <1.5 1.9 0.7 0.5 0.3 0.3 0.2 <1.5 <3 0.6 <1.5 1.1 108.0 57 <5 0.3 <1.5 2.2 1.4 1.0 0.3 0.5 0.3 <1.5 <3 0.9 <1.5 1.2 161.0 58 <5 0.2 <1.5 1.8 <0.7 0.7 0.7 0.7 0.2 <1.5 <3 0.4 <1.5 0.9 210.0 59 <5 0.3 <1.5 0.6 0.7 0.2 0.3 0.3 0.2 <1.5 <3 0.4 <1.5 0.7 205.0 60 <5 0.4 <1.5 1.2 <0.7 0.7 0.2 0.3 0.2 1.8 3.4 0.6 <1.5 1.5 521.0

187

Appendix 16 (continued) Subject Creatinine Pseudouracil Uracil Hypoxanthine Xanthine Adenine Adenosine Inosine Guanosine Dihydrothymine Dihydrouracil Uridine Thymine Orotate Urate 61 <5 <0.2 3.9 3.5 3.7 3.5 1.2 0.6 0.4 <1.5 <3 0.2 <1.5 2.1 146.0 62 <5 0.4 <1.5 <0.2 <0.7 0.3 <0.03 <0.06 <0.03 <1.5 <3 0.3 <1.5 <0.5 235.0 63 <5 0.4 13.9 2.2 9.0 0.9 <0.03 <0.06 <0.03 <1.5 <3 <0.02 <1.5 0.8 72.8 64 <5 <0.2 5.0 0.6 2.1 0.3 <0.03 <0.06 <0.03 <1.5 <3 0.2 <1.5 <0.5 205.0 65 5.7 <0.2 9.6 2.4 5.2 0.6 0.1 0.1 0.1 <1.5 <3 0.1 <1.5 0.5 130.0 66 <5 <0.2 9.0 1.5 2.8 0.6 <0.03 <0.06 <0.03 <1.5 <3 0.3 <1.5 0.5 131.0 67 <5 <0.2 9.4 2.1 7.1 1.0 <0.03 <0.06 <0.03 <1.5 <3 0.0 <1.5 1.6 360.0 68 5.8 0.2 5.7 1.1 5.3 0.1 <0.03 <0.06 <0.03 <1.5 <3 0.3 <1.5 <0.5 268.0 69 <5 0.2 10.9 4.2 6.3 1.0 <0.03 <0.06 <0.03 <1.5 <3 0.1 <1.5 2.9 217.0 70 <5 0.2 5.3 5.2 5.3 0.4 0.1 0.3 0.1 <1.5 <3 0.1 <1.5 0.5 123.0 71 <5 <0.2 9.7 2.6 5.3 0.7 <0.03 <0.06 <0.03 <1.5 <3 <0.02 <1.5 3.1 128.0 72 5.3 <0.2 20.9 3.3 9.9 1.0 <0.03 <0.06 <0.03 <1.5 <3 <0.02 <1.5 4.4 179.0 73 <5 <0.2 <1.5 1.4 0.7 0.2 0.1 0.1 0.0 <1.5 <3 0.1 <1.5 1.9 189.0 74 6.1 <0.2 <1.5 3.3 8.2 0.6 <0.03 <0.06 <0.03 <1.5 <3 <0.02 <1.5 1.1 252.0 75 5.3 0.4 8.2 4.8 14.2 1.1 0.2 0.5 0.4 <1.5 <3 <0.02 2.9 0.5 611.0 76 <5 0.3 4.6 3.9 <0.7 0.4 0.0 0.1 0.0 <1.5 <3 <0.02 <1.5 <0.5 235.0 77 7.3 0.2 21.1 0.8 6.5 0.7 <0.03 <0.06 <0.03 <1.5 <3 <0.02 <1.5 <0.5 206.0

188

Appendix 17 Raw data set of neonatal salivary nucleotide precursors concentrations (µM). Subject Creatinine Pseudouracil Uracil Hypoxanthine Xanthine Adenine Adenosine Inosine Guanosine Dihydrothymine Dihydrouracil Uridine Thymine Orotate Urate 1 6.8 0.5 3.0 14.0 4.4 <0.1 5.4 4.8 2.1 1.8 <3 6.5 <1.5 <0.5 78.0 2 21.0 1.6 <1.5 16.0 7.1 3.1 12.3 14.0 8.7 2.2 3.9 11.2 <1.5 <0.5 182.7 3 29.0 3.0 31.0 61.0 41.6 <0.1 6.5 15.1 8.4 3.9 18.5 25.1 <1.5 <0.5 359.8 4 25.0 1.4 5.7 27.0 14.4 0.8 7.9 8.0 4.1 2.5 7.1 11.0 <1.5 <0.5 124.0 5 22.0 4.9 37.0 95.0 85.4 9.6 26.0 41.0 24.6 3.0 9.8 45.3 <1.5 <0.5 61.6 6 24.0 2.0 <1.5 4.1 4.5 1.4 11.5 21.2 14.6 1.9 4.7 13.1 <1.5 1.9 156.0 7 29.0 2.6 37.0 59.0 37.8 <0.1 7.0 8.5 4.9 3.7 13.5 16.2 <1.5 <0.5 291.0 8 32.0 2.2 16.0 31.0 8.4 0.1 7.8 7.3 6.2 2.9 10.6 16.3 <1.5 <0.5 163.2 9 26.0 4.6 53.0 88.0 68.4 0.4 14.0 20.2 14.2 <1.5 7.6 8.9 <1.5 <0.5 439.0 10 6.2 0.6 2.2 8.6 5.9 1.1 4.5 1.4 1.7 1.6 <3 3.7 <1.5 <0.5 159.0 11 36.0 4.1 17.0 86.0 47.6 2.8 27.2 66.5 22.3 2.1 12.4 32.3 <1.5 0.7 97.3 12 37.0 6.0 31.0 59.0 92.2 0.8 12.5 19.6 9.7 1.8 25.3 16.4 <1.5 <0.5 267.8 13 19.0 2.5 3.3 15.0 32.1 0.9 10.3 16.3 8.6 1.9 6.9 19.6 <1.5 <0.5 101.4 14 15.0 1.2 4.0 10.0 14.7 <0.1 5.5 1.8 1.8 1.5 <3 6.0 <1.5 <0.5 129.0 15 17.0 2.5 5.4 34.0 32.9 1.3 8.3 18.2 8.6 2.1 7.5 9.4 <1.5 <0.5 105.8 16 20.0 1.6 3.6 11.0 10.8 3.6 10.3 6.2 4.0 <1.5 3.5 8.9 <1.5 <0.5 123.0 17 29.0 5.4 7.6 80.0 60.8 5.2 43.9 85.3 34.6 3.1 7.3 43.2 <1.5 0.9 156.0 18 12.0 2.3 40.0 65.0 46.1 0.1 7.0 11.8 5.3 1.6 15.4 15.6 <1.5 <0.5 80.8 19 29.0 2.5 31.0 56.0 34.8 5.6 12.2 10.3 9.0 <1.5 5.1 11.8 <1.5 26.4 58.5 20 27.0 4.0 28.0 47.0 46.1 0.3 29.3 10.7 7.8 <1.5 9.0 23.2 <1.5 <0.5 273.0 21 6.8 1.0 3.4 10.0 6.0 0.1 6.0 1.8 2.0 <1.5 <3 5.1 <1.5 0.9 45.6 22 8.8 1.5 5.7 23.0 17.3 14.4 6.9 4.1 3.4 1.5 <3 4.1 <1.5 <0.5 260.0 23 32.0 1.9 <1.5 10.0 17.0 1.8 11.5 7.6 6.6 2.3 5.3 8.3 <1.5 <0.5 134.0 24 24.0 6.6 95.0 113.0 85.3 1.5 11.3 43.1 28.8 2.0 18.1 42.6 <1.5 0.6 180.8 25 23.0 1.7 18.0 35.0 22.1 0.3 7.0 9.0 5.2 1.9 7.1 12.3 <1.5 <0.5 240.0 26 6.9 0.5 6.5 4.7 2.1 0.1 3.0 1.5 1.8 1.5 <3 3.4 <1.5 1.0 34.8 27 <5 0.5 7.0 5.7 2.5 0.7 9.0 5.4 3.6 <1.5 <3 4.3 <1.5 3.5 25.6 28 18.0 1.5 19.0 28.0 11.5 0.3 11.5 5.9 4.2 2.1 6.8 9.6 <1.5 <0.5 163.8 29 16.0 2.3 3.0 28.0 12.4 3.2 13.1 11.0 8.9 1.8 7.7 7.3 <1.5 0.9 66.3

189

Appendix 17 (continued) Subject Creatinine Pseudouracil Uracil Hypoxanthine Xanthine Adenine Adenosine Inosine Guanosine Dihydrothymine Dihydrouracil Uridine Thymine Orotate Urate 30 20.0 1.0 4.3 10.0 7.5 0.1 4.0 1.2 2.0 2.4 <3 4.7 <1.5 <0.5 124.0 31 42.0 9.0 5.5 84.0 69.0 9.7 36.8 42.8 18.2 2.6 <3 54.0 1.5 2.2 73.1 32 26.0 3.5 <1.5 48.0 33.0 2.2 14.4 29.3 14.0 1.7 8.0 14.4 <1.5 0.6 95.6 33 50.0 3.6 <1.5 49.0 23.0 2.7 23.9 39.5 16.9 2.6 7.4 17.9 <1.5 0.8 92.2 34 43.0 4.0 4.6 56.0 42.0 4.4 76.4 45.0 25.4 <1.5 9.8 34.7 <1.5 0.9 147.0 35 25.0 4.2 85.0 105.0 81.0 0.4 31.5 29.0 20.5 2.6 18.6 39.8 <1.5 0.6 119.0 36 16.0 2.5 17.0 20.0 36.0 3.2 10.1 11.1 8.6 1.5 5.6 16.5 <1.5 0.6 132.0 37 17.0 1.6 <1.5 17.0 8.7 1.8 11.9 14.1 5.0 <1.5 2.8 10.1 <1.5 0.8 109.0 38 58.0 7.4 <1.5 83.0 49.0 9.3 25.5 22.1 9.6 <1.5 3.0 23.9 <1.5 1.0 400.0 39 13.0 1.1 <1.5 12.0 7.5 1.4 16.6 3.2 3.7 <1.5 <3 6.2 <1.5 <0.5 178.0 40 49.0 3.9 56.0 55.0 56.0 0.3 12.4 11.3 7.4 3.4 12.8 13.8 <1.5 0.6 253.4 41 17.0 1.5 <1.5 23.0 14.0 1.7 31.5 16.6 7.0 <1.5 2.6 11.3 <1.5 <0.5 84.0 42 18.0 0.9 5.0 5.9 9.8 0.1 16.8 5.5 4.1 1.7 2.7 11.6 <1.5 <0.5 173.0 43 15.0 1.0 1.7 23.0 12.0 1.7 20.4 13.3 8.3 <1.5 3.4 7.2 <1.5 <0.5 94.8 44 20.0 1.4 <1.5 14.0 11.0 2.9 21.1 4.7 3.9 <1.5 <3 6.1 <1.5 <0.5 193.0 45 30.0 4.2 5.1 54.0 39.0 2.0 67.3 34.5 19.9 3.1 11.1 31.9 <1.5 0.5 198.9 46 5.9 1.0 <1.5 6.2 3.8 0.5 8.9 10.0 6.5 <1.5 <3 8.3 <1.5 3.2 112.0 47 26.0 1.1 4.7 11.0 6.3 0.1 22.7 4.8 3.7 <1.5 <3 11.6 <1.5 <0.5 242.0 48 18.0 2.1 12.0 27.0 20.0 <0.1 6.7 2.7 2.0 <1.5 4.9 9.0 <1.5 0.9 231.0 49 13.0 1.2 1.5 6.0 3.7 0.2 10.3 4.8 4.6 <1.5 <3 6.6 <1.5 <0.5 117.0 50 28.0 4.0 14.0 69.0 31.0 2.2 31.5 98.5 31.5 1.4 9.8 36.5 <1.5 0.6 77.8 51 19.0 1.6 12.0 26.0 24.0 0.9 0.6 10.6 4.6 2.5 3.5 13.5 <1.5 <0.5 185.0 52 36.0 1.9 <1.5 21.0 4.3 4.3 44.5 14.1 10.8 1.6 <3 11.4 <1.5 23.5 109.0 53 7.6 0.7 3.2 7.4 3.8 0.1 10.0 1.7 1.7 <1.5 <3 4.4 <1.5 <0.5 108.0 54 3.4 0.6 <1.5 4.3 2.1 0.1 9.2 0.8 0.9 <1.5 <3 2.9 <1.5 <0.5 73.1 55 8.1 0.6 <1.5 11.0 1.2 1.0 11.1 5.6 3.3 <1.5 <3 3.9 <1.5 <0.5 136.0 56 29.0 5.6 5.5 69.0 37.0 6.5 40.5 64.7 20.1 1.7 7.4 19.8 <1.5 1.2 82.9 57 22.0 1.2 3.1 14.0 13.0 0.3 5.9 6.8 4.4 1.3 <3 16.8 <1.5 <0.5 153.0 58 33.0 3.5 25.1 51.0 45.0 0.4 16.4 13.8 5.9 <1.5 9.2 20.8 <1.5 <0.5 309.0 59 39.0 4.3 12.0 62.0 57.0 9.5 24.7 29.8 10.8 <1.5 10.2 18.1 <1.5 1.9 281.0 60 11.0 68.0 73.0 89.0 0.1 18.6 28.6 11.8 2.1 10.1 32.5 <1.5 1.4 509.0 190

Appendix 18 Raw data set of cow salivary nucleotide precursors concentrations (µM). Subject 1 2 3 4 5 6 7 8 Creatinine 38.1 27.2 20.6 50.5 26.3 34.8 52.5 25.8 Pseudouridine <0.2 0.2 0.2 0.5 0.4 <0.2 <0.2 0.3 Uracil 1.6 5.7 4.1 5.8 5.6 1.6 2.8 3.4 Hypoxanthine 0.2 <0.2 0.3 0.2 0.4 <0.2 <0.2 <0.2 Xanthine <0.7 0.7 <0.7 <0.7 0.9 <0.7 <0.7 0.2 Adenine <0.1 0.2 0.1 0.1 0.6 <0.1 0.1 0.1 Adenosine 0.1 0.7 0.2 0.4 1.3 <0.03 0.1 0.1 Deoxyadenosine 0.1 0.1 0.1 0.2 0.3 <0.02 0.0 <0.02 Deoxyuridine <0.5 <0.5 <0.5 <0.5 0.6 <0.5 <0.5 <0.5 Inosine 0.2 0.9 0.4 1.3 3.0 <0.06 0.2 0.7 Guanosine 0.1 1.1 0.2 1.3 3.9 0.0 0.2 0.4 Deoxyinosine 0.4 0.5 <0.4 0.6 1.9 <0.4 <0.4 <0.4 Deoxyguanosine 0.6 0.5 0.1 0.8 2.0 <0.05 0.3 0.2 Thymidine 0.7 0.9 0.3 0.8 1.7 <0.3 0.5 0.3 Dihydrouracil <3 4.9 6.5 <3 <3 <3 4.1 <3 Uridine 0.2 1.5 0.5 2.3 4.0 0.1 0.3 0.4 Orotate <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 Urate 27.3 34.6 48.0 40.6 30.0 34.7 43.9 26.3 Thymine <1.5 <1.5 <1.5 <1.5 <1.5 <1.5 <1.5 <1.5

191

Appendix 19 Raw data set of sheep salivary nucleotide precursors concentrations (µM). Subject 1 2 3 4 5 Creatinine 7.5 9.8 11.7 7.4 10.7 Pseudouridine 0.4 0.2 0.2 0.2 0.2 Uracil 8.3 9.0 3.4 3.8 9.7 Hypoxanthine 1.1 1.3 1.1 0.9 0.7 Xanthine 2.3 1.7 0.9 7.8 0.9 Adenine 0.1 1.2 1.6 0.3 1.2 Adenosine 2.1 2.2 11.1 7.8 9.0 Deoxyadenosine 0.1 0.1 0.2 0.2 0.2 Deoxyuridine <0.5 <0.5 <0.5 <0.5 <0.5 Inosine 1.1 1.0 2.0 1.0 1.1 Guanosine 1.9 2.5 6.5 2.8 4.9 Deoxyinosine <0.4 <0.4 <0.4 <0.4 <0.4 Deoxyguanosine <0.5 <0.5 <0.5 <0.5 <0.5 Thymidine 0.3 <0.3 0.5 0.6 0.4 Dihydrouracil 6.3 <3 3.3 10.9 6.8 Uridine 1.5 2.2 3.7 1.9 2.9 Orotate <0.5 <0.5 <0.5 <0.5 <0.5 Urate 40.2 43.0 40.2 44.3 42.3 Thymine <1.5 <1.5 <1.5 <1.5 <1.5

192

Appendix 20 Raw data set of goat salivary nucleotide precursors concentrations (µM). Subject 1 2 3 4 Creatinine 22.3 23.4 30.8 25.4 Pseudouridine 1.6 1.5 1.6 1.7 Uracil 40.0 29.5 61.4 50.3 Hypoxanthine 12.8 14.6 29.8 32.4 Xanthine 25.4 14.3 31.2 29.1 Adenine 1.1 4.2 4.1 2.4 Adenosine 1.4 4.0 3.0 2.3 Deoxyadenosine 0.1 0.3 0.2 0.1 Deoxyuridine <0.5 0.5 0.8 0.7 Inosine 5.4 27.1 6.0 4.8 Guanosine 2.7 29.5 6.6 3.8 Deoxyinosine 0.6 0.7 0.9 0.7 Deoxyguanosine 0.3 0.5 0.8 0.3 Thymidine 1.8 2.9 5.1 1.9 Dihydrouracil 15.5 <3 <3 15.2 Uridine 7.2 48.5 13.0 5.6 Orotate 0.9 0.1 <0.5 <0.5 Urate 57.0 47.8 102.6 59.9 Thymine 3.6 2.2 10.4 4.6

193

Appendix 21 Raw data set of horse salivary nucleotide precursors concentrations (µM). Subject 1 2 3 4 5 Creatinine 20.1 29.8 28.9 19.7 30.8 Pseudouridine 1.2 4.4 2.3 2.1 4.9 Uracil 28.2 91.3 72.3 80.5 130.0 Hypoxanthine 0.6 1.4 6.9 19.2 17.8 Xanthine 2.4 18.9 34.0 50.8 48.9 Adenine 0.2 2.5 0.1 0.3 0.1 Adenosine 0.1 0.1 0.1 0.2 0.2 Deoxyadenosine 0.1 0.0 0.0 0.1 0.0 Deoxyuridine 6.2 17.9 13.2 11.3 24.5 Inosine 5.4 22.1 6.9 9.4 26.6 Guanosine 6.2 21.4 6.6 10.2 39.9 Deoxyinosine 5.7 16.7 9.9 9.0 18.8 Deoxyguanosine 2.8 11.6 4.3 7.4 15.2 Thymidine 7.6 20.2 17.8 11.6 25.5 Dihydrouracil <3 <3 <3 <3 <3 Uridine 20.1 61.6 29.6 22.3 94.4 Orotate 0.4 1.4 1.2 <0.5 1.3 Urate 71.4 286.0 193.0 116.0 394.0 Thymine 11.6 21.8 33.1 26.0 47.6

194

Appendix 22 Raw data set of one camel salivary nucleotide precursors concentrations (µM). Subject 1 Creatinine 24.4 Pseudouridine 1.3 Uracil 24.2 Hypoxanthine 14.6 Xanthine 9.3 Adenine 2.5 Adenosine 1.4 Deoxyadenosine 0.5 Deoxyuridine 4.8 Inosine 10.0 Guanosine 10.2 Deoxyinosine 2.9 Deoxyguanosine 2.6 Thymidine 5.5 Dihydrouracil <3 Uridine 17.6 Orotate <0.5 Urate 47.1 Thymine 8.9

195

Appendix 23 Raw data set of dog salivary nucleotide precursors concentrations (µM). Subject 1 2 3 4 5 6 7 Creatinine 6.3 5.3 <5 14.6 12.7 10.5 23.6 Pseudouridine 0.7 0.4 <0.2 0.8 1.1 0.4 0.4 Uracil 33.3 3.2 <1.5 12.7 36.6 12.0 1.6 Hypoxanthine 0.5 0.3 <0.2 <0.2 <0.2 0.2 0.3 Xanthine 4.2 0.9 <0.7 1.1 1.4 1.6 1.2 Adenine 9.5 1.2 0.1 5.7 14.6 19.3 5.6 Adenosine 35.8 19.0 0.8 28.9 9.7 19.9 4.6 Deoxyadenosine 18.2 3.6 0.0 1.9 9.0 4.8 0.3 Deoxyuridine 3.8 <0.5 <0.5 0.5 1.7 0.6 <0.5 Inosine 8.3 3.1 0.4 11.0 3.4 5.8 0.7 Guanosine 3.5 3.5 0.5 6.0 3.3 2.2 0.5 Deoxyinosine 4.8 <0.4 <0.4 1.9 4.1 1.0 <0.4 Deoxyguanosine 5.9 3.5 0.1 1.1 5.7 2.3 0.2 Thymidine 13.7 1.7 <0.3 1.6 8.1 3.4 0.4 Dihydrouracil 3.1 <3 6.3 <3 6.2 <3 4.4 Uridine 9.3 3.4 0.4 7.6 3.9 5.0 1.1 Orotate <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 Urate 62.3 50.0 34.0 45.0 36.0 58.2 42.6 Thymine 2.1 <1.5 <1.5 <1.5 8.4 <1.5 <1.5

196

Appendix 24 Raw data set of cat salivary nucleotide precursors concentrations (µM). Subject 1 2 3 4 5 Creatinine 27.7 28.4 29.0 20.2 27.0 Pseudouridine 0.8 2.2 1.5 0.4 1.2 Uracil 25.1 189.6 82.6 24.9 32.9 Hypoxanthine 11.6 68.2 2.2 0.5 13.4 Xanthine 11.4 115.0 7.8 2.1 17.9 Adenine 5.3 38.0 1.8 0.2 0.4 Adenosine 2.0 10.2 3.2 0.7 1.9 Deoxyadenosine 1.0 2.7 2.5 0.4 2.7 Deoxyuridine 1.6 2.9 24.8 4.5 14.8 Inosine 9.1 18.6 10.3 1.8 30.5 Guanosine 5.4 15.4 12.2 2.6 33.0 Deoxyinosine 2.1 3.8 27.2 5.2 21.3 Deoxyguanosine 1.5 3.6 20.6 3.4 16.4 Thymidine 4.6 8.5 47.5 12.7 29.0 Dihydrouracil <3 <3 <3 <3 <3 Uridine 5.5 14.1 13.1 3.3 40.1 Orotate 1.3 <0.5 2.8 <0.5 0.6 Urate 48.5 68.0 46.9 37.3 31.2 Thymine 15.5 75.4 38.2 16.2 11.5

197

Appendix 25 Gel electrophoresis of bacterial 16S rRNA gene amplicons

198

Appendix 26 Summary of the de novo assembly statistics of 16S rRNA deep sequencing Number of Total Number of Average Sequence Sample %GC Sequences Bases Length 1_BF01 19318 9860320 510 52.55 2_BF02 16462 8359642 507 52.35 3_BF03 12782 6512350 509 52.75 4_BF04 17402 8892343 510 52.54 5_BF05 12069 6168019 511 52.02 6_BF06 15098 7697329 509 52.44 7_BF07 12823 6515245 508 52.54 8_BF08 15376 7795346 506 52.36 9_BF9 14341 7277190 507 52.27 10_BF10 13141 6685145 508 52.16 11_BF11 14726 7489693 508 52.82 12_BF12 15997 8190104 511 52.48 13_BF13 14612 7470382 511 52.45 14_BF14 12835 6578916 512 52.77 15_BF15 13624 6965154 511 52.61 16_BF16 12170 6230373 511 52.5 17_BF17 14329 7326207 511 52.38 18_BF18 12382 6325951 510 52.46 19_BF19 13003 6645984 511 52.37 20_BF20 11444 5867277 512 52.69 21_BF21 10224 5238193 512 51.26 22_BF22 14973 7583317 506 52.57 23_BF23 8282 4129303 510 52.42 24_BF24 11690 5972391 510 52.56 25_BF25 12405 6311171 508 52.52 26_BF26 12787 6519538 509 52.1 27_FF01 13150 6707120 510 52.35 28_FF02 13416 6810800 507 52.32 29_FF03 11574 5903588 510 52.49 30_FF04 11057 5644652 510 52.29 31_FF05 11409 5802858 508 52.37 32_FF06 13552 6883423 507 52.45 33_FF07 9561 4887924 511 52.46 34_FF08 17875 9092127 508 52.49 35_FF09 14296 7282066 509 52.38 36_FF10 12675 6457842 509 52.8 37_BF11 9688 4953122 511 52.49 38_FF12 8304 4241905 510 52.42

Total 589945 300767108 509 52.43 Demultiplexed

199

Appendix 27 16S rRNA sequence dataset 16S rRNA curated sequence data are available in a FASTA format at: http://dx.doi.org/10.6084/m9.figshare.894974 and http://dx.doi.org/10.6084/m9.figshare.894968

200