Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The complexity of understanding human milk components and infant brain development Moukarzel, Sara 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2016_may_moukarzel_sara.pdf [ 2.49MB ]
Metadata
JSON: 24-1.0228326.json
JSON-LD: 24-1.0228326-ld.json
RDF/XML (Pretty): 24-1.0228326-rdf.xml
RDF/JSON: 24-1.0228326-rdf.json
Turtle: 24-1.0228326-turtle.txt
N-Triples: 24-1.0228326-rdf-ntriples.txt
Original Record: 24-1.0228326-source.json
Full Text
24-1.0228326-fulltext.txt
Citation
24-1.0228326.ris

Full Text

 The Complexity of Understanding Human Milk Components and Infant Brain Development  by  Sara Moukarzel  M.Sc., The University of Kansas Medical Center, 2011  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Human Nutrition)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   March, 2016  © Sara Moukarzel, 2016 ii  Abstract Understanding which and how human milk components contribute to infant brain development is complicated in part by their large diversity, complex arrangement in the milk matrix and potential interaction in metabolism. This research addressed the importance of studying the composition of minor milk lipids and of exploring their relationship with non-lipid milk components in infant brain development. More specifically, the milk fat globule membrane (MFGM), a complex tri-layer of cholesterol, glycoproteins, and polar lipids including ethanolamine plasmalogens (Pls-PE), naturally emulsifies milk triacylglycerols but is not currently added to infant milk substitutes. Clinical evidence suggests MFGM plays a role in cognitive function. Whether MFGM directly affects the developing brain is unclear. Due to analytical challenges, little is known about the fatty acid composition of human MFGM lipids, particularly Pls-PE. Pls-PE may be enriched in long-chain polyunsaturated fatty acids (LC-PUFA) such as docosahexaenoic acid (DHA), an important neural lipid during development. Additionally, milk contains different forms of water-soluble choline (WSC) compounds (free choline, glycerophosphocholine, phosphocholine) for which distinct biological roles are unknown, although choline as a molecule per se is an important structural component of the brain and a precursor of the neurotransmitter acetylcholine. After developing an analytical method for separation and recovery of milk Pls-PE, the first study demonstrated both human and cow milk Pls-PE are enriched in LC-PUFA including DHA compared to other phospholipids. Milk Pls-PE DHA does not seem to vary with maternal DHA intake. Using artificially-reared infant rats, the second study showed that developmental brain phospholipids and metabolites differ between rats fed formula with or without MFGM, with a closer phospholipid composition to mother-reared rats in rats fed MFGM. By analyzing human preterm and term milk samples for iii  WSC composition using mass spectrometry in the third study, we confirmed previous findings of the wide variability in WSC total content and composition in human milk and reported no significant association between individual WSC compounds. These studies provide new knowledge that milk contains novel components potentially relevant to the brain, and, while the mechanisms for improved cognition remain unclear, MFGM affects neonatal brain phospholipid composition.                    iv  Preface This dissertation was prepared according to the University of British Columbia Faculty of  Graduate and Postdoctoral Studies requirements.   The research in Chapter 2 was designed by Dr. Sheila Innis and me. The analytical method development section of the research was partially funded by a discovery grant from Mead Johnson. The cow milk samples were a research donation from the UBC Dairy Farm via Nelson Dinn. Roger Dyer, senior lab technician in our lab, Dr. Bernd Keller and I worked together to develop the analytical method, with technical assistance from our former technician Michael George. Recruiting participants, conducting study visits, collecting dietary data and milk samples were my primary responsibility, with assistance from former graduate student Lynda Soberanes. Lynda entered the dietary data into the nutrient analysis software. I was responsible for the lipid analyses of the milk samples. I was also responsible for the data analysis and preparation of the data for Chapter 2, with data interpretation done with Dr. Innis.    The research in Chapter 3 was designed by Dr. Sheila Innis, Dr. Cyrielle Garcia and me. Animal care and sample collection was done by members of the Innis lab, including Alejandra Weideman, Dr. Cyrielle Garcia, Guilaine Boyce, Janet King, Roger Dyer, and Vinodha Chetty. Rat cannulation was done by Alejandra Weideman, Dr. Cyrielle Garcia and Roger Dyer, with assistance from graduate students Vinodha Chetty and Guilaine Boyce. I prepared the experimental diets and assisted in sample collection. I was responsible for all lipidomic and metabolic sample analyses, with assistance from Roger Dyer and Dr. Bernd Keller in identifying brain metabolites by GC-MS. I was also responsible for the data analysis and preparation of the data for Chapter 3, with data interpretation done with Dr. Innis.    v  The research in Chapter 4 was designed by Dr. Sheila Innis and me. I was responsible for obtaining ethics approval, collecting milk samples, and analyzing milk fatty acids. Analysis of choline compounds in milk was done by Lynda Soberanes and myself, with guidance from Roger Dyer. I was responsible for the data analysis and preparation of the data for Chapter 4, with data interpretation done with Dr. Innis.    Ethics approval was required for this research and was approved by the Committee for Ethical Review of Research Involving Human Subjects at the University of British Columbia (B.C.) and the B.C. Children’s and Women’s Hospital; certificate number H12-03191 for Chapter 2, and H13-03393 for Chapter 4. Ethics approval to conduct the research in Chapter 3 was obtained from the University of British Columbia Animal Care Committee; certificate number A13-0257.              vi  Table of Contents Abstract ........................................................................................................................................... ii Preface............................................................................................................................................ iv Table of Contents ........................................................................................................................... vi List of Tables ................................................................................................................................. xi List of Figures .............................................................................................................................. xiv List of Abbreviations ................................................................................................................... xvi Acknowledgments……………………………………………………………………………..xviii Chapter 1. Introduction ................................................................................................................... 1  1.1 Background and Rationale……………………………………………………………..1            1.2 Literature Review-Section one ...................................................................................... 3               1.2.1 Lipids: Definition and Classifications .................................................................. 3               1.2.2 Human Milk: General Composition ..................................................................... 7               1.2.3 Human Milk Lipids .............................................................................................. 8                    1.2.3.1Human Milk Lipid Composition ............................................................... 8                    1.2.3.2 Formation of the Milk Fat Globule........................................................... 9               1.2.4 The Milk Fat Globule Membrane....................................................................... 11                    1.2.4.1 History of Molecular Organization of the MFGM ................................. 11                    1.2.4.2 MFGM Lipids: Arrangement and Composition ..................................... 11                           1.2.4.2.1 PL in Human and Cow Milk ..................................................... 12 vii                            1.2.4.2.2 PL Fatty Acids in Human Milk ................................................. 15                          1.2.4.3 MFGM Proteins Composition .......................................................... 17              1.2.5 Lipid Synthesis in the Mammary Gland ............................................................. 18                   1.2.5.1 Sources of Fatty Acids in Milk ................................................................ 18                   1.2.5.2 Triglyceride Synthesis in the Mammary Gland ....................................... 19                   1.2.5.3 PE and PC Synthesis in the Mammary Gland ......................................... 20              1.2.6 Essential Nutrients and Current Dietary Reference Intakes ................................ 23              1.2.7 Biological Roles of ω-3 and ω-6 Fatty Acids ..................................................... 23              1.2.8 Plasmalogen Digestion and Metabolic Fate ........................................................ 26              1.2.9 The Human Brain ................................................................................................ 28                    1.2.9.1 Brain Development: The Human and the Rat ........................................ 31                           1.2.9.1.1 Morphological Stages of Brain Development........................... 31                            1.2.9.1.2 Developmental Lipid Changes in the Brain ............................. 33                            1.2.9.1.3 Biological Functions of Plasmalogens in the Brain ................. 35                     1.2.9.2 Relevance of Several Dietary Components of MFGM to the Brain ...... 38                     1.2.9.3 Clinical Studies on MFGM.................................................................... 42            1.3 Literature Review-Section Two ................................................................................... 45             1.3.1 Choline in Human Milk ....................................................................................... 45             1.3.2 Recommendation for Infant Intake of Choline .................................................... 47             1.3.3 Biological Roles of Choline: Emphasis on the Brain ........................................... 48 viii              1.4 Research Rationale and Objectives ............................................................................. 52                1.4.1 Rationale and Objectives for Chapter 2 ............................................................ 52                1.4.2 Rationale and Objectives for Chapter 3 ............................................................ 53                1.4.3 Rationale and Objectives for Chapter 4 ............................................................ 55 Chapter 2. Long chain Polyunsaturated Fatty Acids: Where are They in Milk Lipids? ............... 57              2.1 Introduction ................................................................................................................ 57             2.2 Methods....................................................................................................................... 59             2.3 Statistical Analysis ...................................................................................................... 60             2.4 Description of the Developed Analytical Method ...................................................... 60            2.4.1 Solvents ............................................................................................................. 60            2.4.2 Analytical Method Development....................................................................... 61              2.5 Results for Analytical Method ................................................................................... 62              2.6 Results for Mature Human Milk Pls-PE Fatty Acids................................................. 66              2.7 Results for Cow Colostrum and Mature Milk Pls-PE Fatty Acids. ........................... 80              2.8 Discussion .................................................................................................................. 83 Chapter 3. The Milk Fat Globule Membrane and the Brain ......................................................... 89             3.1 Introduction ................................................................................................................. 89             3.2 Methods....................................................................................................................... 91               3.2.1 Cannulation ........................................................................................................ 91               3.2.2 Animal Maintenance .......................................................................................... 92 ix                3.2.3 Feeding Schedule ............................................................................................... 93             3.2.4 Animal Diets ........................................................................................................ 93             3.2.5 Tissue Collection .................................................................................................. 94             3.2.6 Analytical Methods .............................................................................................. 95             3.2.7 Statistical Analysis ............................................................................................... 95            3.3 Results .......................................................................................................................... 96           3.4 Discussion ................................................................................................................... 107 Chapter 4: Human Milk Choline- A Potential Player in Infant Milk Lipid Handling ................ 113           4.1 Introduction ................................................................................................................. 113           4.2 Methods....................................................................................................................... 116             4.2.1 Biochemical Assessment .................................................................................... 117             4.2.2 Statistical Analysis ............................................................................................. 118           4.3 Results ......................................................................................................................... 118           4.4 Discussion ................................................................................................................... 122 Chapter 5. Conclusion ................................................................................................................. 126           5.1 Summary of Specific Contributions............................................................................ 126           5.2 Strengths and Limitations ........................................................................................... 127           5.3 General Comments...................................................................................................... 130            5.4 Future Directions ....................................................................................................... 133 Bibliography ............................................................................................................................... 138 x  Appendices .................................................................................................................................. 160            Appendix A. Supplemental analytical methods ............................................................... 160 ……A.1 Procedure for Lipid Extraction ................................................................................. 160 ……A.2 Procedure for Methylating Fatty Acids of Different Lipid Fractions ...................... 160         A.3 Procedure for Methylating Total Fatty acids in Milk ............................................... 162         A.4 Procedure for Determining Protein Content in Brain Samples ................................ 163         A.5 Procedure for Preparing Brain Samples for GC-MS analysis. ................................. 164         A.6 Procedure for Cannula Preparation .......................................................................... 165            Appendix B. Supplemental List of Total Milk Fatty Acids for Individual Participants on 3 Separate Days.............................................................................................................................. 166            Appendix C. Fatty Acid Composition of Experimental Formulas and MFGM phospholipids .............................................................................................................................. 167           xi   List of Tables Table 1.1 Description of common fatty acids ................................................................................. 4 Table 1.2 Macronutrient composition of human, cow and rat colostrum and mature milk ............ 7 Table 1.3 Lipid classes in human and cow colostrum and mature milk ......................................... 9 Table 1.4 Distribution of major phospholipids in mature human and cow milk .......................... 13 Table 1.5 Distribution of major phospholipids in human and cow milk ...................................... 14 Table 1.7 Examples of proteins identified in the human MFGM by functional characteristic .... 18 Table 1.8 Total phosphatidylethanolamine fatty acids of the brain and the heart ........................ 25 Table 1.9 Fatty acids of selected phospholipids in human brain gray matter and myelin ............ 30 Table 1.10 Lipid composition of human brain gray matter and myelin ....................................... 30 Table 1.11 Major stages of human brain development and associated timeframes ...................... 32 Table 1.12 Characteristic periods of rat brain development ......................................................... 33 Table 1.13 Temporal changes in rat brain lipids with development ............................................. 35 Table 1.14 Selected studies linking MFGM components to postnatal brain development .......... 39 Table 1.15 Composition of choline compounds in mature term and preterm milk ...................... 47 Table 2.1 Description of the solvent gradient used for HPLC analysis ........................................ 62 Table 2.2 Intra-assay and inter-assay variability of Pls-PE and diacyl-PE recovery.................... 64 Table 2.3 Comparison of fatty acids in PE and PC before and after separation and recovery using SPE following by HPLC ............................................................................................................... 66 Table 2.4 Maternal characteristics ................................................................................................ 67 Table 2.5 Fatty acids in Pls-PE, diacyl- PE, PC and total lipid of mature human milk ............... 69 Table 2.6 Dietary polyunsaturated fatty acids intakes of breastfeeding women .......................... 71 xii  Table 2.7 Fatty acid composition of Pls-PE, diacyl-PE and PC from three mature milk samples collected from the same woman (#1) on three separate days ....................................................... 74 Table 2.8 Fatty acid composition of Pls-PE, diacyl-PE and PC from three mature milk samples collected from the same woman (#2) on three separate days ....................................................... 75 Table 2.9 Fatty acid composition of Pls-PE, diacyl-PE and PC from three mature milk samples collected from the same woman (#3) on three separate days ....................................................... 76 Table 2.10 Fatty acid composition of Pls-PE, diacyl-PE and PC from three mature milk samples collected from the same woman (#4) on three separate days ....................................................... 77 Table 2.11 Macronutrient composition of human milk samples within a women across days .... 78 Table 2.12 Total fatty acids in milk collected from each participant on 3 separate days ............. 79 Table 2.13 Fatty acids in Pls-PE, diacyl- PE, PC and total lipid of cow colostrum ..................... 81 Table 2.14 Fatty acids in Pls-PE, diacyl- PE, PC and total lipid of mature cow milk .................. 82 Table 3.1 Nutrient and oil composition of experimental milk formulas ....................................... 94 Table 3.2 Total body and brain weights of mother-reared rats (MR), rats fed formula with MFGM (MFGM-fed) and rats fed formula without MFGM (No-MFGM) .................................. 97 Table 3.3 Developmental changes in brain lipids of mother-reared rats (MR) ............................ 98 Table 3.4. Protein content in brains of mother-reared rats (MR), rats fed formula with MFGM (MFGM-fed) and rats fed formula without MFGM (No-MFGM) ............................................... 98 Table 3.5 Changes in brain lipids in rats fed formula with or without MFGM .......................... 101 Table 3.6 Polyunsaturated fatty acid composition of brain PE and PC in mother-reared rats ... 102 Table 3.7 Net DHA content in brain PE in mother-reared rats (MR), rats fed formula with MFGM (MFGM-fed) and rats fed formula without MFGM (No-MFGM) ................................ 103 Table 3.8 Differences in brain metabolites between rats fed formula with or without MFGM . 105 xiii  Table 3.9 Polyunsaturated fatty acid composition of brain PE and PC of rats fed formula with or without MFGM ........................................................................................................................... 106 Table 4.1 Composition of choline compounds in mature term and preterm milk ...................... 115 Table 4.2 Content of water-soluble choline compounds in preterm and donor milk samples .... 119 Table 5.1 Strengths and limitations of the analytical method for the separation and recovery of selected milk PL. ......................................................................................................................... 127                   xiv  List of Figures Figure 1.1 Schemaic of glycerophospholipids ................................................................................ 5 Figure 1.2 Schematic of sphingomyelin ......................................................................................... 6 Figure 1.3 Schematic of the diacyl-phospholipid and plasmalogen structure ................................ 6 Figure 1.4 Simplified schematic representation of the milk fat globule formation and secretion in the mammary epithelial cell. ......................................................................................................... 10 Figure 1.5 Summary of major suggested synthesis pathways for selected human milk lipids ..... 22 Figure 1.6 Schematic of a neuron ................................................................................................. 29 Figure 1.7 The structures of water soluble choline-containing compounds in human milk ......... 46 Figure 1.8 Overview of phosphatidylcholine synthesis pathways ................................................ 52 Figure 2.1 HPLC Chromatogram showing no detectable phospholipid in the presence of triglyceride .................................................................................................................................... 63 Figure 2.2 HPLC chromatograms showing separation of Pls-PE from other lipids as authentic standards and in mature human milk ............................................................................................ 65 Figure 2.3 Relationship between DHA in milk total lipid and DHA in Pls-PE, diacyl-PE and PC....................................................................................................................................................... 70 Figure 2.4 Variability of DHA in milk Pls-PE, diacyl-PE and PC with individual women ......... 73 Figure 2.5 HPLC chromatogram showing Pls-PE and diacyl-PE peaks with no TEA in the mobile phases ................................................................................................................................ 85 Figure 4.1 Overview of choline metabolic pathways ................................................................. 116 Figure 4.2 Box plots of the distribution of total lipid and protein content (g/dL) ...................... 119 in preterm and donor milk samples ............................................................................................. 119 Figure 4.3 Scatter plots of the relationship between the different WSC forms in preterm milk 121 xv  Figure 4.3 Percent distribution of the different WSC compounds in individual PM samples.... 122                       xvi  List of Abbreviations AI: adequate intake ACh: Acetylcholine ADA: adrenic acid (22:4 ω-6) ALA: α-linolenic acid (18:3 ω-3) ANOVA: analysis of variance  ARA: arachidonic acid (20:4 ω-6) BBB: blood-brain barrier BF3: boron-trifluoride CDP: Cytidine-5′diphosphocholine DMA: dimethyl acetal DHA: docosahexaenoic acid (22:6 ω-3) DHAP: dihydroxyacetone phosphate  DRI: Dietary Reference Intakes Eth: Ethanolamine EPA: eicosapentaenoic acid (20:5 ω-3) ER: endoplasmic reticulum  F-Cholesterol: Free Cholesterol FC: Free Choline GC-MS: Gas chromatography mass spectrometry GLC: gas liquid chromatography  GPC: Glycerophosphocholine  GM: gray matter Glycero-3-P: Glycerol-3-phosphate HPLC: high performance liquid chromatography  Kcal: kilocalorie  LA: linoleic acid (18:2 ω-6) LC-PUFA: long chain polyunsaturated fatty acids MCFA: medium chain fatty acid MFGM: milk fat globule membrane MUFA: monounsaturated fatty acid NMR: nuclear magnetic resonance PC: phosphatidylcholine    PE: phosphatidylethanolamine  PhosC: Phosphocholine PI:  phosphatidylinositol   PL: Phospholipid Pls: Plasmalogen PS: phosphatidylserine   SA: sialic acid SAFA: saturated fatty acid SAM: S-adenosylmethionine SD: standard deviation  SEM: standard error of the mean SPE: Solid phase extraction xvii  Sph: Sphingomyelin TG: triacylglycerol  TFA: total fatty acids VLDL: very low density lipoprotein WSC: water-soluble choline wk: weeks xviii  Acknowledgments  I would like to first thank my supervisor Dr. Sheila Innis for her guidance, patience, and unconditional dedication to research, without which this work would not have been possible. I am grateful for the long hours we spent raising questions, re-evaluating dogmatic ideas, and planning how best we can contribute to moving the field of human milk research forward. Dr. Innis sadly passed away few days before my dissertation defense. Beyond memories of the long waiting queue by her office door to discuss data, her common phrases “all good”, “you’re getting there” and “sit back and enjoy the chaos” still echo in my mind as a reminder of her resilience.  I was fortunate to be also supervised by Dr. Rajavel Elango, who stepped in to help during a very critical period of my dissertation writing. I would have not been able to finalize my dissertation document without his guidance and positive energy, and I thank him for his valuable mentorship.  I would also like to thank my supervisory committee members, Drs. David Kitts and Tim Green for their patience and support throughout my studies, and mostly for their endorsement of the discovery nature of this research.   There is no doubt that group work at the Innis lab from brainstorming to project execution was essential for the completion of this research. I am very grateful for the assistance of Roger Dyer, Janette King, Alejandra Weideman, Dr. Cyrielle Garcia, Guilaine Boyce, and Vinodha Chetty, particularly for the long hours of animal care. Roger Dyer has patiently helped me develop my analytical skills, one pipette tip at a time! For his long emails, morning chats, and emergency phone call replies, I am very thankful! A special thanks for Dr. Bernd Keller for sharing his expertise and for always explaining complex chemistry concepts in simple terms. Also, thanks to my fellow colleagues Dr. Kelly Mulder, Lynda Soberanes and Jie (Jessica) Yang for their support (and coffee breaks) over the years.   I would like to acknowledge the University of British Columbia for providing trainee funding support through the Leonard S. Klinck Memorial Fellowship and The Four-Year Doctoral Fellowship.   On a personal note, my deepest gratitude goes to my family for their unconditional love and support.    1  Chapter 1. Introduction 1.1 Background and Rationale Milk, a complex biological fluid produced by the mammary gland, has evolved over 150 million years to be the sole source of nutrition and immuno-protection for the infant, capable of sustaining life in an often pathogen-manifested and nutrient-poor extra-uterine environment (1, 2). Despite the species-specific evolution of milk characteristics to meet nutrition needs and species-dependent levels of developmental maturity at birth, it is fascinating that the mammary gland preserved common milk characteristics across species. These include the unique mammary-gland-specific synthesis of lactose, lipid structures and their arrangement in milk fat globules, and different forms of water-soluble choline (WSC) compounds. Lipids in milk form globules with a triacylglycerol (TG) core surrounded by a milk fat globule membrane (MFGM), an emulsifier consisting of three layers of potentially-bioactive phospholipids (PL), glycoproteins and cholesterol (3). Choline compounds in milk are predominately in water-soluble form and can be in free form (free choline) or phosphorylated with or without a glycerol backbone for which distinct biological functions are not yet understood (4). Although milk substitutes such as infant formula are available, they currently contain lipids without the MFGM and choline is almost entirely in free form, for which their implications on infant development are not completely understood. A major limitation in understanding the composition and potential importance of MFGM lipids and WSC, when this research was initiated five years ago, was in the analytical methods, mainly designed to measure total and average milk fatty acid and choline composition.  Milk lipids and choline are implicated in brain development, but emphasis has been on the role of the omega (ω)-6 and ω-3 fatty acids in the accumulation of brain docosahexaenoic 2  acid (DHA), an important functional component of neural lipids (5, 6). There has been little interest in the molecular forms in which the ω-6 and ω-3 fatty acids are provided to the infant (eg. MFGM PL) or the molecular forms themselves in brain development. Similarly, research on choline has been extensive relating total dietary choline to brain choline composition (i.e phosphatidylcholine (PC) and acetylcholine (Ach)), with little differentiation between the different dietary choline forms (7, 8).  The overall objective of this dissertation was to gain a better understanding of the unique lipid composition of human milk, particularly MFGM plasmalogens and to address the potential role of supplementing MFGM in the diet in infant brain development. This dissertation also includes an exploratory section that addresses the composition of WSC in milk. This research is important to understand whether the lipid structures in human milk contribute to infant development, beyond the roles of ω-3 and ω-6 fatty acids and to emphasize the need to integrate other milk components, such as WSC when addressing lipid-related functional outcomes likely to be influenced by different dietary components. The following literature review is divided into two main sections (1.2 and 1.3). The first provides background on milk lipids and fatty acids focusing on the MFGM, which is followed by review of early postnatal brain development and current understanding of the effects of milk lipids on brain lipid changes, particularly PL and DHA during infancy. The second section provides background on choline and its metabolism, current knowledge on choline compounds in milk and the role of choline in brain development.   3  1.2 Literature Review-Section one 1.2.1 Lipids: Definition and Classifications  The term “lipids” describes organic compounds that are soluble in organic solvents (e.g. chloroform, hexane) and contain hydrocarbon groups as primary parts of the molecule. Compound classes covered in this definition include fatty acids (FAs) and their esters, acyl- and alkenyl-glycerols, sterols, and isoprenoid hydrocarbons.  Fatty acids have a methyl end (-CH3), a methylene carbon chain (CHn) and a carboxyl end (COOH). They are distinct in their structure based on their chain length, degree of unsaturation, and the position of the first double bond, when present, from the methyl end in unsaturated fatty acids. Fatty acids with six or fewer carbons, eight to fourteen carbons, and 16 or more carbons are generally classified as short, medium and long chain respectively, with ω-3 and ω-6 fatty acids with 20 or more carbons sometimes referred to as very long chain. Saturated fatty acids (SAFA) do not have a double bond, monounsaturated fatty acids (MUFA) contain one double bond, and polyunsaturated fatty acids (PUFA) contain two or more double bonds.  Unsaturated fatty acids can be further categorized into groups of ω-X (or n-X), where X is the number of the first double bond from the methyl end. One of the common nomenclatures of a fatty acid is in the form of A:B ω-X, where “A” is the number of carbons and “B” is the number of double bonds. For example, 18:1 ω-9 is an 18-carbon MUFA, with the double bond at the ninth carbon from the methyl end. Table 1.1 summarizes the names of fatty acids commonly found in nature, including milk, with their linear formulas.   4  Table 1.1 Description of common fatty acids Nomenclature Common Name Linear Formula Common SAFA 12:0 Lauric acid CH3 (CH2)10 COOH 14:0 Myristic acid CH3 (CH2)12 COOH 16:0 Palmitic acid CH3 (CH2)14 COOH 18:0 Stearic acid CH3 (CH2)16 COOH Common MUFA 18:1 ω-9 Oleic acid CH3 (CH2 )7 CH=CH(CH2)7 COOH Common PUFA 18:3 ω-3 α-linolenic acid  CH3 CH2(CH=CHCH2)3(CH2)6 COOH 20:5 ω-3 Eicosapentaenoic acid CH3 (CH2CH=CH)5(CH2)3 COOH 22:6 ω-3 Docosahexaenoic acid CH3(CH2CH=CH)6CH2CH2 COOH 18:2 ω-6 Linoleic acid  CH3 (CH2)4CH=CHCH2CH(CH2)7 COOH 20:4 ω-6 Arachidonic acid CH3 (CH2)4 (CH=CHCH2)4 (CH2)2 COOH  SAFA, saturated fatty acids; MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids It is well known that the structure of a fatty acid contributes to its distinct function in biological systems, beyond fatty acid oxidation for cellular energy production (9). One example is the decrease in cell membrane fluidity by substituting a cis-unsaturated fatty acid with a trans-fatty acid, two exact molecules with the only difference being the geometrical orientation of the hydrogen atoms around the double bond (10). Fatty acids in mammalian cells can be present in TG, a molecule with a glycerol backbone and three esterified fatty acids at the stereo-specifically denoted  (sn)-1, sn-2 and sn-3 positions, or in more complex forms, including 5  glycerophospholipids (phospholipids, PL) and sphingolipids. PL, most commonly the diacyl-PL, consist of a glycerol backbone, with fatty acids esterified at the sn-1 and sn-2 positions and a phosphate with a hydroxylated polar head group at the sn-3 position (Figure 1.1). Usually, the sn-1 fatty acid is saturated or monounsaturated, whereas the sn-2 fatty acid is generally polyunsaturated. Sphingolipids, however, have the nitrogen-containing sphingosine as the backbone, one fatty acyl chain, and a side group, such as phosphocholine, a monosaccharide, or a sialylated oligosaccharide to form sphingomyelin (Sph, Figure 1.2), a cerebroside, or a ganglioside respectively.  The fatty acid in a sphingolipid is saturated or monounsaturated, particularly 18:0, 18:1 ω-9, 24:0 and 24:1 ω-9.  Quantitatively, PL and Sph are the most prominent in mammalian cells.  Cholesterol and its esters are not covered within the scope of this research work. Figure 1.1 Schemaic of glycerophospholipids   6  Figure 1.2 Schematic of sphingomyelin  Plasmalogens are a specific class of PL, distinguished by a vinyl ether moiety at the sn-1 position of the glycerol backbone, rather than an ester bond as is found in diacyl-PL (Figure 1.3). The polar head group in plasmalogens is usually choline or ethanolamine, and the sn-2 position is typically a long chain unsaturated fatty acid such as 20:4 ω-6 or 22:6 ω-3, with levels of these unsaturated fatty acids being higher in plasmalogens than in the corresponding diacyl-PL (11-13). Plasmalogens are found in the membranes of animal species, both vertebrate and invertebrate, and anaerobic bacteria, but are absent in plants, fungi and aerobic bacteria (14). An in-depth discussion on these lipids in human milk and in the brain is presented in sections 1.2.4.2.1 and 1.2.9.1.3 of the literature review.        Diacyl-phospholipid Plasmalogen Figure 1.3 Schematic of the diacyl-phospholipid and plasmalogen structure Diacyl-phospholipid Plasmalogen phospholipid 7  1.2.2 Human Milk: General Composition Human milk contains macro and micronutrients, as well as non-nutritive immunoprotective and developmental components (i.e immunoglobulins, stem cells, growth factors), able to support growth and development of the exclusively-breastfed infant. Milk composition changes throughout lactation, providing the infant with nutrients needed at specific stages of growth and development (15). Accordingly, milk can be categorized into three types: colostrum, transitional and mature (16). Colostrum is a thick yellow fluid secreted by the mammary gland for a few days before parturition to approximately 4 days postpartum. It is characterized by a high concentration of proteins, including lactoferrin and immunoglobulins (15, 17). Milk secreted at 5-14 days postpartum is considered transitional and is characterized by a shift from the high protein in colostrum to higher lactose and fat that is characteristic of mature human milk produced after about 14 days postpartum (18). Mature milk has a thinner liquid consistency than colostrum. Table 1.2 provides the average macronutrient content of human colostrum and mature milk, with cow and rat milk included for reference. Table 1.2 Macronutrient composition of human, cow and rat colostrum and mature milk  1 Sample taken at postpartum day 0. Fat content drops to 7.7 at day 5.  Only one paper is available on rat colostrum fat content, which was estimated from total esterified fatty acids, using an average triglyceride molecule weight (MW:245).   Human milk from Prentice, 1996 (18). Cow milk from Kehoe, 2007 (19).  Rat milk from Jensen, 1995 (20) and Nicolas, 1991 (21).    Human  Cow  Rat g/100 ml Colostrum Mature  Colostrum Mature  Colostrum1 Mature Fat      2.9 4.2  6.7 3.8  23 8.8 Lactose      5.3 7.0  2.6 4.8  2.6 3.8 Protein      2.0 1.1  15 3.3  7.8 8.1 8  It should be noted that Table 1.2 only provides average macronutrient content of milk, but the macronutrient content of milk varies, particularly for fat within a feed, with variability in composition of fatty acids occurring within and across women (22). Foremilk, secreted during the first 2-3 minutes of a feed is low in fat unlike hind milk, which is secreted later during a feed, and contains around four folds the amount of fat in foremilk (23). Milk DHA, for example, varies from 0.1-2.8 % total fatty acids across women (24). 1.2.3 Human Milk Lipids 1.2.3.1 Human Milk Lipid Composition  Lipids in milk are predominantly comprised of TG, secreted as globules emulsified in the milk aqueous phase due to their envelopment with an amphipathic lipid-containing globule membrane, the MFGM (3). Although low as a percentage of total milk lipid quantity, it is important to acknowledge that milk provides the infant with a dietary source of MFGM- lipids, including PL, cholesterol and other minor lipids (Table 1.3). With a two-fold increase in triglyceride between colostrum and mature milk, the percentage of PLs from total milk lipid decreases from 1% in colostrum to 0.3-0.5% in mature human milk, but the absolute content of PL remains constant (3). It is noteworthy to highlight that existing data on milk lipid classes, particularly as percent of total lipid is from the 1980-1990’s (25, 26) and is limited by the inaccuracy of the analytical methods at that time, including thin-layer chromatography and indirect estimation from total esterified fatty acids (27).     9  Table 1.3 Lipid classes in human and cow colostrum and mature milk    Human1                 Cow  Colostrum Mature  Colostrum Mature Triglycerides  Phospholipid  Cholesterol  Diglycerides  Monoglycerides Free fatty acids  Cholesteryl esters 97.6 1.1 1.3 - - - - 98.9 0.6 0.5 - - - -  97.2 0.72 0.53 1.01 0.06 0.26 0.05 97.4 0.56 0.30 1.72 0.03 0.18 0.04 Values are percent total lipid by dryweight. 1Values are from d3 and d42 in human and d3 and d180 postpartum in cow for colostrum and mature milk, respectively Bitman and Wood, 1990 (26) and Bitman et al, 1983 (25).  It is well-known that TG in milk are an important energy source for the breastfeeding infant, providing around 50% of calories and a source of essential fatty acids LA and ALA (20, 28). The sustenance of other lipids in milk, including PL through evolution raises the question of the importance of these molecules, beyond providing a physical carrier system for TG in milk. The summary below aims at highlighting the unique formation of milk fat globules in the mammary gland and the unique structure of the MFGM, in which PL are present, to begin to appreciate the potential functional role(s) of these lipids. An in-depth description of milk fat globule synthesis and secretion can be found in previous reviews (29-31).  1.2.3.2 Formation of the Milk Fat Globule  Triglycerides, synthesized in the endoplasmic reticulum (ER) of mammary gland epithelial cells, coalesce together to form microlipid droplets (< 0.5 um diameter), which are then released from the ER, surrounded by a single layer of PL and proteins (32) (Figure 1.4). Larger cytoplasmic lipid droplets (> 1 um diameter) form by droplet-droplet fusion and migrate to the cell apical surface, likely via cytoplasmic microtubule interactions. At the cell apical surface, lipid droplets become gradually coated with the cell plasma membrane, until completely released 10  into the alveolar lumen. During secretion, cytoplasm fractions known as cytoplasm crescents, may become entrapped between the lipid globule and its outer membrane. Therefore, unlike any other biological membrane, the MFGM is comprised of three layers, with or without a cytoplasm crescent: an inner monolayer of polar lipids and a bilayer of polar lipids (i.e glycolipids, PL), cholesterol, glycoproteins and proteins (29). Figure 1.4 Simplified schematic representation of the milk fat globule formation and secretion in the mammary epithelial cell.               Size of organelles, milk globules and membranes are not to scale. (a) nucleus, (b) Golgi Apparatus, (c) Smooth endoplasmic reticulum, (d) Rough endoplasmic reticulum, (e) mitochondria, MLD: microlipid droplet, CLD: cytosolic lipid droplet, MFG: milk fat globule. Image modified from Mammary Epithelial Cell by Database Centre for Life Science (DBCLS), licensed under Creative Commons 3.0.  MFG MLD CLD (a) (b) (c) (e) Basal plasma membrane Apical plasma membrane MFG Inner monolayer Outer bilayer  Cytoplasmic crescent crescent Triglyceride core (d) 11  1.2.4 The Milk Fat Globule Membrane 1.2.4.1 History of Molecular Organization of the MFGM  Our current knowledge on the molecular organization of the MFGM is the result of a century of experimentation, transitioning understanding of the MFGM as an adsorbed layer of milk-serum-derived compounds to a true biological membrane. Briefly, in 1959, Bargmann and Knoop (33) as described by Keenan et al (34), provided the first evidence that milk fat globules are secreted from mammary epithelial cells by envelopment in their apical plasma membrane. Between 1954 and 1967, it became clear that the MFGM contained enzymes, typically found in cellular membranes (34). The striking similarity of the PL composition between the MFGM and mammary cellular membranes was found next (35), with a more sophisticatedly understanding that this “biological membrane” in milk contained proteins asymmetrically organized along the lipid bilayer occurring in the early 1970s (36, 37). Today, our understanding is that the MFGM is certainly a tri-layered biological membrane, with a PL monolayer originating from the cytosol (i.e ER) and the PL bilayer originating from the apical plasma membrane (32).  1.2.4.2 MFGM Lipids: Arrangement and Composition Lipids of the MFGM consist of PL (diacyls and plasmalogens), sphingolipids, (i.e Sph), glygosphingolipids (i.e cerebrosides and gangliosides) and cholesterol. Their arrangement within the membrane is highly structured, forming two membrane phases: a lipid-ordered and a lipid-disordered phase (32). Lipid rafts and clustered Sph form the lipid-ordered phase of the MFGM, which is characterized by a high degree of rigidity relative to other sections of the MFGM (32). Lipid rafts refer to Sph and cholesterol that aggregate symmetrically along sections of the MFGM bilayer. Interestingly, lipid rafts have also been recently shown to be enriched in ethanolamine-plasmalogens (Pls-PE) (38). Sph molecules can also cluster together on the outer 12  membrane of the MFGM bilayer. It has been suggested that the lipid-ordered phase may contribute to the protection of bioactive milk components against gastric digestion (3, 39). The lipid-disordered phase, however, is composed of PL: PC, PE, PS and PI, whose arrangement within the lipid disordered phase is also specific. PC is mainly present in the outer leaflet and PE, PS and PI in the inner leaflet of the MFGM bilayer. The PL composition of the MFGM monolayer and arrangement of plasmalogens in the MFGM, however, are less understood. Regardless, the specific arrangement of PL between the outer and inner membrane intrigues interesting questions about their potential functional roles, particularly because of the known difference in their amounts and fatty acid composition in milk.  1.2.4.2.1 PL in Human and Cow Milk Although similar in type, milk PL differ in proportion between the human and the cow (Table 1.4). Particularly, plasmalogens in mature human milk comprises about 40% of total ethanolamine PL, compared to only 10% in mature cow milk (40). However, the information available to date shows considerable variability in the amounts of different PL in human and cow milk as evident, for example, in the wide standard deviations for means presented in Table 1.4, especially for plasmalogens, PE, PS and PI. This is likely related to the variability and limitations of analytical techniques, which will be discussed. Of interest, the identification and inclusion of Pls-PE in milk PL data is recent, with only a few reports since the year 1990 (40-42). van Beusekom et al (42) were the first to report on Pls-PE in milk, estimating Pls-PE at 14-15% of total milk PL and 64-68% of total milk PE in humans. However, they had major analytical limitations, including inability to resolve Sph from lyso-PC and Pls-PE from Lyso-PE by high performance liquid chromatography (HPLC) and indirect estimation of PL classes using their respective total fatty acids by gas chromatography (GC). Accordingly, their estimates were not in 13  accordance with common literature, reporting Sph for example at 48-57% and diacyl-PE at 7% of total PL. It was not until the year 2012 that the presence of Pls-PE in human milk was confirmed (40, 41) (Table 1.4), with quantification of PL using 31P NMR spectroscopy.  Table 1.4 Distribution of major phospholipids in mature human and cow milk               Adapted from Garcia 2012 (40).  Values are means ± SD for n=15 cows and n=22 humansSph, sphingomyelin; PE, diacylphosphatidylethanolamine;  Pls-PE, ethanolamine plasmalogen; PC, phosphatidylcholine;  PS, phosphatidylserine; PI, phosphatidylinositol  While it seems reasonable to review the data on differences in the PL distribution between colostrum and mature milk, it is questionable whether differences in PL found between colostrum and mature milk of humans and cows (Table 1.5) are accurate, or an artifact of analytical methodologies, given the variability highlighted earlier within one milk type.         Human  Cow  Sph 29.3  2.28 20.0  1.38 PE  18.8  3.37 31.8  2.88 Pls-PE 11.5  2.04 4.91  2.53 PC 24.1  3.01 28.7  2.73 PS 8.42  2.34 9.97  2.81 PI 3.67  1.10 3.72  1.87 14  Table 1.5 Distribution of major phospholipids in human and cow milk  Human  Cow  Colostrum Mature  Colostrum Mature Sph PE* PC PS PI 41 13 32 9 5 38 21 28 8 5  29 31 28 8 4 31 20 35 2 12 Values are percent total PL. * total PE, including Pls-PE. Values are from d3 and d42 postpartum in human milk, published in 1984 (43) and from d3 and d180 postpartum in cow milk, published in 1990 (26) Sph, sphingomyelin; PE, diacylphosphatidylethanolamine; Pls-PE, ethanolamine plasmalogen; PC, phosphatidylcholine;  PS, phosphatidylserine; PI, phosphatidylinositol One of the major challenges contributing to the variability in milk PL, including plasmalogens, is in the analytical methods. A comprehensive review on these challenges in plasmalogen analysis has been recently published (44).  Briefly, lipids generally need to be separated from other milk components, before further analysis. This is usually done by Folch lipid extraction using saline, chloroform and methanol (45), or modified methods with added acid to recover acidic lipid molecules (i.e PI and PS) (46). Considering the amphipathic nature of PL, their extraction from an aqueous biological system, such as milk is not trivial. Addition of acid to recover acidic PL jeopardizes intact plasmalogen recovery, with the well-known susceptibility of plasmalogen to oxidation (44, 47). In addition, most early studies in milk PL, their quantities and changes with lactation used thin layer chromatography (TLC) to separate PL and analysis of inorganic phosphorus in lipid extracts for their quantification (25, 26, 43, 48-50).  Techniques based on TLC, however, are compromised by the large amounts of TG in milk, inability to separate PL using simple solvent systems, and imprecise quantitative recovery from TLC plates for subsequent analysis. More recently, techniques based on 31P NMR spectroscopy, HPLC and tandem mass spectrometry are becoming more common in milk PL analyses (40, 41, 15  51, 52), although none except Garcia et al (40, 41) developed methods for Pls-PE estimation in milk. Regardless, some general conclusions can be drawn.  PL are about two-fold higher, as a percent of total lipid, in colostrum than mature human milk (20), with mature milk containing ~150-470 µg/ml PL (0.4-1.4 % total lipid) (40). An alternative interpretation is that PL are relatively constant, with an increase in TG during maturation of milk. PE and PC are the major PL in human milk, but their relative proportions, along with other milk PL are inconsistently reported. 1.2.4.2.2 PL Fatty Acids in Human Milk  Because milk fatty acids are almost entirely in TG, fatty acids are frequently analyzed in the total milk lipid in studies focusing on milk fatty acid variability with diet, disease and across populations, and in studies relating milk composition to infant health outcomes. For example, studies conducted in Canada, the United States and European countries generally report as mean percent of total milk fat, 1-1.5 % ALA, 12-16 % LA, 0.4-0.7 % 20:4 ω-6 (arachidonic acid, ARA), and 0.15-0.40 % DHA (53-55), with a wide range of fatty acid concentrations among women, particularly for DHA, accounted for partly by the variability in dietary DHA intake. Milk from vegan women is reported to contain around 0.1 % DHA in total lipid (56), while milk from women consuming diets rich in DHA-containing foods have been reported to contain 1.4-2.8 % DHA (57, 58). DHA supplementation from different sources is known to increase DHA in milk (59-62). However, the fatty acid composition in human milk total lipid reflects that of the milk TG, and fatty acids in other lipid fractions including PL are diluted with the overwhelming presence of TG. The fatty acid distribution in milk PL, observationally and in response to interventions, is less studied.  16   Six studies on the fatty acid composition of human milk PL fatty acids are identified (43, 63-67), with the majority being observational from the late 1960-1980’s. A key finding from these early studies is that milk PL, particularly PE contain a higher proportion of PUFA than TG with a stereo-specific fatty acid distribution such that PUFA are predominantly at the sn-2 position of the PL. An example for mature human milk PE compared to PC is provided in Table 1.6.  Table 1.6 Mean composition of major fatty acids in mature human milk PE and PC Fatty acid PE PC 18:2 ω-6 20.8 21 20:3 ω-6 2.2 1.1 20:4 ω-6 8.4 2.4 22:4 ω-3 2.2 0.5 22: 5 ω-6 1.7 0.4 18:3 ω-3 2.9 0.8 20:5 ω-3 0 0 22: 5 ω-3 0.9 0.1 22:6 ω-3 1.6 0.3 Total saturates   34.3 53.3 Total monounsaturates 23.1 17.4 Adapted from Bitman el al, 1984 (43).  Values are g/100 g fatty acid. Note sum of % fatty acids as reported in original paper is not 100.   Data are from n=9 mature term milk samples, 21-84 days postpartum. PE, phosphatidylethanolamine; PC, phosphatidylcholine.  Surprisingly, data is available from only one intervention study on the changes in human milk PL fatty acids in response to maternal supplementation with DHA (i.e ω-3 polyunsaturated fatty acid enriched eggs) (64). Although baseline PL fatty acid composition was not presented, the authors reported higher enrichment of long-chain PUFA, including ARA and DHA in milk 17  total PL compared to milk TG (64). Similarly, a strikingly high percentage of PUFA in the mature milk PL of Japanese women was found, such that total ω-6 and ω-3 fatty acids were 27 % and 12 % in PE, 18 % and 4 % in PC, but only 15 % and 3 % in total milk lipid, respectively in Japanese women (65). Particularly, ARA and DHA were 13 % and 5 % in PE, but only 3 % and 0.6 % in PC, and 1 % and 1.1 % in total milk lipids, respectively. While it is important to acknowledge that these are percentages from a different total number of fatty acids (i.e two fatty acids in a PL compared to ~three in total lipid (i.e TG), the higher percentage of PUFA in PL raises the question of a difference in metabolic fate (i.e Does the DHA on the sn-2 position of PE in milk have a different functional role than DHA in TG?). 1.2.4.3 MFGM Proteins Composition When studying the composition of the MFGM and its role in brain development, it is important to address MFGM proteins, not only lipids. Comprehensive reviews on the major proteins in human and cow MFGM are available (68-70). Briefly, MFGM proteins representing only 1-4 % of total milk proteins, are often undetected in the milk proteome, with predominant presence of casein proteins (69). A better understanding of the complexity of the MFGM proteome, however, has been made possible as a result of the tremendous advancement in proteomic analytical techniques over the last two decades. More than 190 intracellular, extracellular and membrane associated proteins have been recently identified in the human MFGM and are mainly involved in lipid and energy metabolisms, cell communication and signal transduction, and immune function (71). A sample list of these proteins is provided in Table 1.7. Interestingly, the cow MFGM has similar protein components to the human MFGM, with a comprehensive list of these proteins available elsewhere (69, 70).  18  Table 1.7 Examples of proteins identified in the human MFGM by functional characteristic  Functional Characteristic  Examples of human MFGM proteins  Immune function     Energy production and metabolism   Lipid metabolism2  polymeric immunoglobulin receptor human leukocyte antigens mucin 11 lactoferrin1 lactadherin 1 alcohol dehydrogenase [NADP+] isocitrate dehydrogenase [NADP+]; cytoplasmic glutathione peroxidase 3 butyrophilin xanthine dehydrogenase/oxidase3 1-acylglycerol-3-phosphate O-acyltransferase  fatty acid synthase bile-salt stimulated lipase lipoprotein lipase apolipoproteins A-1, A-11, A-IV, B-100, D,  and E fatty acid-binding protein long-chain-fatty-acid-CoA ligase 1/3/4 perilipin-2 peroxiredoxin-6 Summarized from Liao et al, 2011 (71), 1 antimicrobial agents (72). 2 enzymes in this category are associated with milk fat globule formation/secretion, de novo lipid synthesis, lipid digestion, transport, activation, storage and degradation 3also has functions in oxidative metabolism of purines in the liver  1.2.5 Lipid Synthesis in the Mammary Gland 1.2.5.1 Sources of Fatty Acids in Milk  Fatty acids in human milk are either synthesized in the mammary epithelial cell (C16:0 or less) or taken up from plasma: from chylomicrons and very low density lipoproteins (VLDL)-TG or as 19  non-esterified albumin-bound fatty acids (C≥16) (28). Briefly, glycolysis-derived acetyl CoA is carboxylated to malonyl CoA via acetyl-CoA carboxylase in the mammary epithelial cell cytoplasm. This is the first and rate-limiting step of fatty acid biosynthesis, positively regulated by insulin. Six two-carbon units from malonyl-CoA are then sequentially added to one molecule of acetyl CoA, via the cytoplasmic fatty acid synthetase complex. The mammary-gland-specific medium-chain acyl thioesterase II then cleaves the newly-formed fatty acyl chain at C14 to generate myristic acid (C14:0) or before C14 to form C10:0 and C12:0 (73). It is noteworthy that in ruminants including the cow, acetate and ß-hydroxyacetate from the rumen bacteria fermentation of cellulose and ß-cellulose, are the major sources of carbon for fatty acid synthesis (28). C16:0 predominantly originates from de novo synthesis in the human mammary gland. Fatty acids > C16 are cleaved, however, from their sn-1 and sn-2 positions in plasma chylomicron and VLDL-TG by mammary lipoprotein lipase and taken up for further mammary gland lipid synthesis. Therefore, fatty acids in human milk may originate from the mammary gland itself or externally from dietary fatty acid intake, dietary fatty acid metabolism in the maternal liver, hepatic de novo fatty acid synthesis, and fatty acids from the maternal adipose tissue (28). 1.2.5.2 Triglyceride Synthesis in the Mammary Gland The major precursors for TG synthesis in the ER of the mammary epithelial cell are glycerol-3-phosphate, and monoglycerides, with the glycerol-3-phosphate pathway likely to be the major synthesis pathway in humans (74). Glycerol-3-phosphate is formed either from the phosphorylation of glycerol, or from the reduction of the glycolysis-derived dihydroxyacetone phosphate (DHAP).  Monoglycerides and glycerol are derived from chylomicron and VLDL-TG breakdown. Unique to the human mammary gland, an unsaturated fatty acid, commonly C18:1 20  ω-9 is acylated at the sn-1 position of glycerol-3-phosphate, followed by C16:0 at the sn-2 position to form 1,2-diacyl-phosphoglyceride (73, 75). This is followed by a dephosphorylation and an unsaturated fatty acid acylation at the sn-3 position, to form a triglyceride. Similarly, two fatty acids can be acylated to the monoglyceride to form a triglyceride. It is unusual that TG in milk are enriched with C16:0 at the sn-2 position, with a conservation of this stereo-specificity post-absorption (73).  1.2.5.3 PE and PC Synthesis in the Mammary Gland As previously mentioned, PL in milk are part of the tri-layered MFGM. Although not directly studied in humans yet, studies in animals indicate that PL in milk arise from de novo synthesis in the mammary gland, not by uptake from the circulation (76, 77). Similar to TG, synthesis of diacyl-PE (and PC) also occurs in the ER, with similar steps as previously described to form 1,2-diacyl glyceride from DHAP. Here, ethanolamine is phosphorylated via ATP-dependent ethanolamine kinase to form phospho-ethanolamine, which then forms CDP-ethanolamine via CTP: phosphoethanolamine cytidylyl transferase. The final step of diacyl-PE synthesis is catalyzed by ethanolamine 1,2-diacylglycerol ethanolaminephosphotransferase, which adds a phosphoethanolamine head group at the sn-3 position of a 1,2-diglyceride to form diacyl-PE (78, 79). Alternatively, PS decarboxylation can generate diacyl-PE at least in the cow (78, 80), but the contribution of this pathway to diacyl-PE synthesis in the human mammary gland is not known. Of relevance, it not known if ethanolamine is synthesized de novo and it may originate from the diet. PC is similarly synthesized via the CDP-choline pathway and alternatively, can be formed by the sequential methylation of PE, using S-adenosylmethionine (SAM) as the methyl donor (81).  21  Pls-PE synthesis is more complex, involves the ER, peroxisome and mitochondria, and is mostly studied in the liver with little knowledge in the mammary gland (82). Unlike in TG and diacyl-PE synthesis, the acylation of DHAP to form 1-acyl DHAP, as the first step of plasmalogen synthesis, occurs in the peroxisome not the ER. This is followed by the substitution of the fatty acyl group at sn-1 with a fatty alcohol, to form 1-alkylDHAP. The following reduction of 1-alkylDHAP to 1-alkylglycerol-3-phosphate takes place in both the ER and the peroxisome. Then, acylation of a fatty acid at the sn-2 position and addition of a phosphoethanolamine head group at sn-3 to form 1-alkyl-2-acyl-glycerol-3-phosphoethanolamine occur in the ER. The final step is to form the vinyl ether bond at sn-1 by desaturation, and this occurs in the mitochondria (47).  To summarize, milk TG and likely PL, including plasmalogens are synthesized de novo in the mammary gland (Figure 1.5, adapted from Braverman (47)), with their fatty acid derived from circulation and/or de novo synthesis. The TG and PL are in different compartments in milk, with PL present in the MFGM surrounding the TG core.          22  Figure 1.5 Summary of major suggested synthesis pathways for selected human milk lipids   To summarize, the MFGM is a complex membrane containing both lipids and proteins. The breastfeeding infant is provided with the entire MFGM and not the individual components separately. While in vitro studies are important for determining mechanisms of action of individual MFGM components, such as in immune and cognitive functions during infancy, it is crucial to acknowledge the complexity of the MFGM structure and composition. The functional role of the MFGM may extend beyond the roles of its individual components. It is plausible that an interaction between its different components and the structure in which it is presented to the infant contributes to a unique MFGM function. Recently, bovine MFGM became commercially 23  available, opening the possibility for animal studies and clinical trials that explore the role of the entire globule membrane in infant health.   1.2.6 Essential Nutrients and Current Dietary Reference Intakes Dietary nutrients required to support growth, normal cellular function and freedom from disease, but cannot be synthesized by the human are considered essential. The recommended intakes of most nutrients for the infant are based on average nutrient composition in human milk, with the assumption that human milk from healthy women must provide the ideal nutrition for adequate infant growth and development. The identification of nutrients for which intake recommendations are needed for infants is usually triggered by research in adults, showing essentiality or reduction in chronic disease risk. Particularly relevant to this research, the dietary reference intakes (DRI) for infants include recommendations as adequate intake (AI) for the essential fatty acids ALA (0.5 g/d for infants 0-12 mo) and LA (4.4 and 4.6 g/d for infants 0-6 mo and 6-12 mo, respectively). Some groups have made recommendations for DHA intake during infancy, including the European Food Safety Authority (100 mg/d for infants 7-24 mo) and the French Safety Agency ( 70 mg/d  for infants 6-12 mo) (83, 84). While researchers debate setting DRIs for DHA for infants, other aspects of milk lipids should be considered. Key to these projects is the form of dietary lipid in which the DHA and other polyunsaturated fatty acids are delivered to the human infant (PL vs TG). 1.2.7 Biological Roles of ω-3 and ω-6 Fatty Acids The essentiality of the dietary ω-6 fatty acid 18:2 ω-6 was first recognized in rodents suffering from scaly skin, tail necrosis and reduced growth (85), with either deficiency symptoms prevented or skin lesions, and weight gain restored with addition of LA or much smaller amounts of ARA to the diet (86-88).  The ω-6 fatty acid deficiency was later identified in 24  humans, with the typical deficiency symptoms of skin rash in infants fed fat-free milk and in infants supported with fat-free parenteral nutrition (89-91). The essentiality of the ω-3 fatty acid 18:3 ω-3 was acknowledged much later, in the early 1970s, with early work in rodents showing altered learning, behavior and visual function in rodents fed diets without ALA or other ω-3 fatty acids (92-94). Later in the 1980’s and after, the role of ω-3 fatty acids in the retina and visual function became more evident, with much of the work in this area done in monkeys fed fat as vegetable oil low in ALA (i.e safflower oil with <0.5% ALA compared to soybean oil with ~7% ALA) (95-97).  In humans, the essentiality and potential importance of ω-3 fatty acids for neurologic functioning were first reported in 1982 after a patient (6 yr old female) receiving parenteral (intravenous) fat as safflower oil, with no ALA for 5 months, suffered from paresthesia, numbness, weakness, and blurred vision (98). Symptoms were alleviated within 12 weeks when the lipid source was changed to soybean oil, which contains ALA (98).  Since this time, studies on the complex roles of ω-3 fatty acids in tissues have been extensive, with tremendous emphasis on the role of the fatty acids themselves independent of the molecules and structures in which they are present. More recently, for example DHA supplements became available in different lipid forms (i.e TG in fish oil; PL in krill oil), however, the bioavailability and differences in functional roles of ω-3 fatty acids-containing molecules continue to be investigated (99).    The ω-3 and ω-6 fatty acids are mainly in PL of cellular membranes, with the PL themselves being diverse and unique in tissue distribution and fatty acid composition. For example, PL in the human heart tissue are ~40% PC and 20% PE, with at least 30% of PC as plasmalogens (100). In the adult human brain, however, PE is much higher, at ~36%, and the plasmalogens are predominantly ethanolamine-containing. Pls-PE constitutes up to 70% of the 25  ethanolamine-containing lipids in myelin and 40% in neural membranes (101, 102). Table 1.8 provides an example of the tissue-specific fatty acid composition of PL, in the case here, PE. However, the methodological approach and identification of specific fatty acids likely explains at least in part, the differences between human and rat PE fatty acids.  Table 1.8 Total phosphatidylethanolamine fatty acids of the brain and the heart  Human1 Rat2  Brain Heart  Brain Heart Fatty acid g/100 g fatty acid 16:0 5.9 3.6  6.57 15.0 18:0 30.4 21.1  23.0 30.2 18:1 8.7 3.8  8.34 7.2 18:2 ω-6 0.5 2.8  0.28 5.0 20:4 ω-6 13.2 35.1  14.7 18.5 22:4 ω-6 8.3 0.8  n/a n/a 18:3 ω-3 0.5 0.5  n/a n/a 20:5 ω-3 n/a 0.7  n/a n/a 22:5 ω-3 n/a 2.1  0.15 3.3 22:6 ω-3 28.6 7.2  20.4 17.5 1 Brain from Svennerholm, 1968 (103), cerebral gray matter, 26 yr female; Heart from Rocquelin, 1989 (100) mean,n = 36, 11-80 yr 2 Brain from Letondor, 2014 (104) frontal cortex, mean, n=10; adult controls; Heart from Benediktsdottir, 1988 (105), reference aduts; source of fat=butter.   The ω-3 and ω-6 fatty acids are not only integral to the cell membrane structure, but they are also distinctly functionally relevant. Particularly, they are involved in the regulation of gene expression and cell signaling and as precursors of lipid mediators (106-109). For example, ARA and EPA are precursors for the synthesis of eicosanoids, key mediators of the inflammatory response, and more recently, EPA and DHA have been shown to convert into anti-inflammatory and inflammation resolving molecules (E and D-series resolvins). Roles of ω-3 and ω-6 fatty 26  acids in gene expression, more understood in the liver and adipose tissue (108, 110) may extend to the brain. For example, DHA has been shown to reduce despair behavior and improve working memory in distressed mice by likely acting as a ligand for the retinoid X receptor (RXR), a nuclear transcription factor known to be expressed in the brain (111). The mechanisms by which the ω-3 and ω-6 fatty acids function at the molecular level are fatty acid and tissue-dependent. Therefore, care should be taken in extrapolating in vitro data to the human, particularly because fatty acids in vivo are predominantly lipid-bound before activation (e.g part of membrane PL or circulating lipoproteins) and not free, as typically used in in vitro studies.  Nonetheless, relevant to postnatal brain development, in vitro studies have suggested that DHA has roles in synaptogenesis and synaptic transmission (112, 113). 1.2.8 Plasmalogen Digestion and Metabolic Fate The question of whether dietary plasmalogens contribute directly to brain development, whether through contributing to plasmalogens in the brain or through other direct role(s), requires addressing plasmalogen digestion, absorption and its metabolic fate. Unlike for TG, the literature on plasmalogen digestion and absorption is scarce. Intracellular lipolytic plasmalogen enzymes have been investigated since the late 1980’s, with emphasis on their role in maintaining cellular homeostasis and preventing toxic cellular levels of plasmalogen breakdown by-products (114). However little attention has been given to the role of similar enzymes within the digestive system and the potential implications on the functional role of dietary plasmalogens. The complex MFGM structure in which plasmalogens are present, likely in the lipid rafts and the MFGM monolayer, is speculated to protect milk lipids from gastric degradation (3). Nishimukai et al (115) have reported dietary plasmalogen is absorbed with 80% efficiency, with evidence suggesting plasmalogen disappearance from the GI tract reflects absorption rather than 27  degradation. Dietary plasmalogen vinyl ether bonds were not degraded after incubation for 1 hour at a pH of 3.6, mimicking gastric conditions, and disappearance (i.e assumed absorption) of plasmalogen from the closed loop of the upper small intestine was ~36% mol, similar to absorption of diacyl-PL. Similar to diacyl-PL digestion, pancreatic phospholipase A2 is known to cleave the sn-2 fatty acid of plasmalogen to release a lyso-plasmalogen and one fatty acid in the small intestinal lumen (116). This raises the question of the importance of an sn-2 long-chain-PUFA-enriched dietary plasmalogen if the sn-2 fatty acid is released during digestion. Interestingly, although cleavage occurs, evidence suggests preferential re-esterification of lyso-plasmalogens to choline plasmalogen occurs with long chain PUFAs in the enterocyte (117). Lyso-plasmalogenase, an enzyme that catalyzes hydrolytic cleavage of the vinyl ether bond of lyso-plasmalogen, forming a fatty aldehyde and glycerophosphoethanolamine or glycerophosphocholine has been found in rat intestinal mucosa (118). Intriguingly, it is only specific for sn-2 deacylated plasmalogens, with no hydrolytic activity with intact plasmalogen in liver, with unknown substrate specificity in the intestine (114, 118). It is reasonable therefore to question the relative activities of these enzymes in the human infant, what the plasmalogen digestion byproducts are (i.e fatty aldehyde and lyso-plasmalogen), how well they are absorbed, and the extent to which they are used to synthesize plasmalogens post-absorption for delivery to the tissues.  The relative abundance of plasmalogen digestion byproducts in the intestinal lumen and bile salt micelles of the human, including the infant is not known. Lyso-plasmalogens and fatty aldehydes are absorbed intact in rodents (119). In the enterocyte of the adult rat, lyso-plasmalogen is preferentially acylated with ARA and DHA at the sn-2 position to form Pls-PE, transported via lymph and not the portal vein (117, 120). Whether de novo synthesis of 28  plasmalogen in the enterocyte occurs to contribute to plasmalogen in the chylomicron rather than enterocyte membrane PL is not known. Interestingly, the final enzyme that dictates a plasmalogen formation rather than a diacyl-PL, 1-O-alkyl-2-acyl-sn-glycero-3-PE desaturase is known to be present in the hamster intestinal mucosa (121).  Of relevance, adult rats consuming a diet fortified with 1.47 g/day bovine brain PL ( ~ 292 µg plasmalogen) for seven days had plasma plasmalogen concentrations four times higher than rats fed a diet without PL or comparable amounts of PC, with higher plasmalogen concentrations in the liver but not any other organ, including the brain (115). Similarly, feeding plasmalogens to adult rats for nine weeks resulted in a statistically significant increase in erythrocyte Pls-PE, which was higher in ARA than Pls-PE of control animals. To summarize, knowledge on plasmalogen digestion and absorption, particularly in the human infant is incompletely understood. Evidence suggests dietary plasmalogen bypasses gastric digestion, undergoes intestinal digestion, but may be re-esterified, at least in part to maintenance plasmalogen in circulation.  1.2.9 The Human Brain Undermining normal changes in the developing brain and as a response to dietary stimuli is complicated by the complexity of the brain anatomy and high degree of site specificity for different brain functions, also impacting brain metabolism. As such, the brief summary below aims to highlight the complexity of the brain structure and composition, all of which raise the question of whether analyzing entire brain content for key metabolites or nutrients accurately reflects functionally relevant changes in the brain.  The neuron is the fundamental information processing cell type in the brain (Figure 1.7). The mature human brain consists of more than 100 billion neurons, forming at least 60 trillion 29  neuronal connections with the connection site between two neurons referred to as a synapse. A synapse can be axodendritic, axoaxonic, dendrodentritic or axosomatic depending on the site of connection between the two neurons. Structurally, the mature brain has a convoluted surface, with characteristic ridges and crevices, and a cross sectional view shows a distinct color gradient from a gray exterior (the gray matter) to a white interior (the white matter). The grey matter predominantly consists of neuron cell bodies, hence its gray color, whereas the white matter includes myelinated nerve fibers, a collection of singular long projections of a neuron called axons clustered together to form information processing networks across different brain areas. Given the difference in cellular substructures present in the brain grey and white matters, the lipid composition of the brain gray and white matter is expected to be profoundly different (Tables 1.9 and 1.10). Caution should be used if lipids are taken as a markers of development, as changes found in a section of the grey matter may not reflect that in the white matter and analysis of the entire brain may dilute effects seen in specific brain sections.    Figure 1.6 Schematic of a neuron  Original image by Nicolas Rougier and from the Wikimedia public domain http://upload.wikimedia.org/wikipedia/commons/7/72/Neuron-figure-notext.svg.  30  Table 1.9 Fatty acids of selected phospholipids in human brain gray matter and myelin   PE  PC  PS  GM M  GM M*  GM M Fatty Acids g/100 g fatty acid    16:0 5.0 6.0  46.4 35.1  1.7 1.4    18:0 25.3 28.2  11.0 14.6  33 47.7    18:1 4.8 38.1  29.2 40.1  9.6 30.5    18:2 ω-6 0.2 1.4  0.2 0.3  0.2 0.4    20:4 ω-6 19.7 3.1  8.1 3.0  3.2 2.2    22:4 ω-6 0.5 -  Tr Tr   Tr tr    22:5 ω-6† 12.3 2.8  Tr Tr   6.0 1.6    22:5 ω-3 4.4 1.9  Tr Tr   8.6 4.5    22:6 ω-3 23.3 4.5  Tr Tr   33.5 4.3 Adapted from O’Brien JS et al, 1965 (122). * data from white matter † It is likely that 22:5 ω-6 and 22:4 ω-6 are reversed in the original paper.  Note: 18:3 ω-3 and 20:5 ω-3 are not detected. GM, Gray matter; M, myelin; PE, phosphatidylethanolamine; PC, phosphatidylcholine; PS, phosphatidylserine; Tr, trace. Table 1.10 Lipid composition of human brain gray matter and myelin   GM M Total lipids 36.4 78 Total PL 20.3 31.7 PE 6.8 14.2 PC 10.8 12.1 PS 2.8 5.5 Total sphingolipids 5.1 24.7 Sphingomyelin 1.8 4.6 Cerebrosides 2.5 18.8 Ceramides 0.8 1.2 Cholesterol 7.9 18.6 Adapted from O’Brien JS et al, 1965(123). Values are g/100g dry weight from frontal lobes from one individual,10 mnths of age.  Note: Total PL does not include phosphatidylinositol; 3 g/100g in GM and M unidentified. GM, Gray matter; M, myelin; PL, phospholipids; PE, phosphatidylethanolamine; PC, phosphatidylcholine; PS, phosphatidylserine;    31  1.2.9.1 Brain Development: The Human and the Rat Despite tremendous advancement in our knowledge on behavioral development in the human, much is yet to be understood on the structural and molecular changes in the brain during prenatal and postnatal development (124), complicated by the difficulty of current techniques that raise ethical questions on human brain samples. Therefore, rats are commonly used to study the effects of fatty acids in the central nervous system. In determining the extent to which findings on rat brain development can be extrapolated to the human, it is essential to address the developmental differences in timeframes of development onset and progression, as well as differences in brain morphology and biochemistry between the two species.  1.2.9.1.1 Morphological Stages of Brain Development Human brain development is a complex dynamic process, beginning in utero as early as three weeks post-conception and encompassing a series of hallmark stages, where a nutrient “insult” has the potential to limit normal progression into the following developmental steps. Brain development begins with neurulation and proceeds with neuronal migration, synaptogenesis, pruning, myelination, and cortical thinning (125).  The different major stages are summarized in Table 1.11, with reviews on prenatal and postnatal human brain development available elsewhere (113, 124, 125).       32  Table 1.11 Major stages of human brain development and associated timeframes Development stage Brief description Timeframe Neurulation Transformation of the embryonic ectoderm into a neural plate, followed by bending and folding to form a closed neural tube 3rd-4th  wk  Neuronal proliferation Rapid proliferation of neuroblasts 4th-12th wk  Neural differentiation Differentiation of neuroblasts into specific neuronal cells or macroglia (i.e astrocytes and oligodendrites)  By 8th wk  Neural migration Migration of immature neurons along radial glia or parallel to the outer cortical surface to their final positions 12th – 29th wk, peaks between 12-20 wk; complete by 26-29 wk  Synaptogenesis Synapse formation between two neighboring neurons, occurring concurrently with dentritic and axonal growth and with myelination of the subcortical white matter  18 wk – adolescence; peaks between 34 wk to 24 months postnatal Myelination Surrounding of oligodendrite axons with myelin, a glycosphingolipid and cholesterol-rich plasma membrane with phospholipids rich in saturated and monounsaturated fatty acids, facilitating the propagation of nervous impulses across neurons 28 wk- adulthood The rat central nervous system is premature at birth, relative to the human. This can be simplistically observed in the physical features and reflexes of an infant rat, born with closed eyelids and folded ears and unresponsive to primary reflex stimulations, such as the ear and eyelid twitches in response to an external physical stimulus, eg gently touching the ear with a cotton swab. These reflexes develop at the earliest at 9 days of age in the rat (126, 127) compared to the eyelid twitch and extremities traction since birth in the human. Table 1.12  summarizes the characteristic periods of brain development in the rat, as described by Wells et al (128). 33  Table 1.12 Characteristic periods of rat brain development       1.2.9.1.2 Developmental Lipid Changes in the Brain The human brain increases in size from 25 % at birth (~380 g) to around 70% (~950 g) of the adult brain weight (~ 1400 g) at the end of the first year of life (129). A more dramatic increase in postnatal brain size occurs in the rat, whose brain weight at birth (~0.29 g) is around 18% of that of the adult (i.e 400 days old; ~2.1 g), but increases sharply to around 70% (~ 1.36 g) within the first 20 days of life (130). The morphological changes in the brain during development are accompanied by changes in brain lipids and fatty acid content. Brain growth requires a quantitative increase in DHA (and other fatty acids, such as 22:4 ω-6 (adrenic acid, ADA)), particularly during the last trimester of pregnancy until at least two years of age in humans (131). During this period, the amount of total brain DHA increases 25-fold, which parallels the increase in brain size, resulting in an increase in the proportion of total brain DHA (g/100 g fatty acids) and results in a lower % ARA (131). The net accumulation of DHA (nmol/g brain) occurs to the largest extent in the brain grey matter Pls-PE, diacyl-PE and PS, based on data published in humans at or before 1998 (11, 122, 123). Concurrently, the amounts of ARA and ADA remain constant in diacyl-PE. In Pls-PE, however, the amounts of ARA, ADA and DHA all increase with maturation (11). The fatty acids 24:0, 24:1 ω-9 and 26:1 ω-9 but not DHA significantly increase in Sph, reflecting myelination (11, 131). The preferential distribution of Period Time  Characteristic change (s) I Embryonic and fetal  life  Neuronal differentiation and proliferation II 0-10 days postnatal Changes in gray matter: axonal and dendrites outgrowth; increase in brain size  III 10-20 days postnatal Changes in white matter: myelination  IV 20 days-adulthood Decreased rate of myelination; maturation     34  DHA into PE and PS, not PC, Sph or PI suggest PE and PS DHA may have important roles in the brain (24).  In the rat, similar to the human, neurogenesis is complete prenatally, and synaptogenesis begins in utero and continues after birth (132). Consistent with the peak in human synaptogenesis from 34 wk gestation-24 months postnatal, continuing to 5-7 years, a synaptogenesis surge occurs in the rat during the second week postnatal (132).  Myelination is minimal prenatally, but begins the first 12-15 days of postnatal life in the rat. This is reflected by an increase in brain 20:1 and 24:1 fatty acids, prominent fatty acids in myelin, from 15 days to adulthood (i.e means of 132 µg 20:1 and 127 µg 24:1 in 15 day old rat brains, compared to 2137 µg and 2115 µg in the adult rat, respectively) (130). Conversely, the rapid quantitative increase in ARA (0.6 mg at birth to 4.2 mg) and DHA (0.6 mg to 4.8 mg) in the rat brain during the first 15 days of life, translating to an accumulation of more than 50% of adult rat brain content of these fatty acids (6.4 mg and 10.2 mg, respectively), coincides with the period of gray matter development (130). Despite the increase in brain myelination, the amount of DHA continues to increase, consistent with brain development.  Brain lipid changes during development also extend to changes in brain PL composition in both rats and humans. Most prominently, the proportion of PC decreases with a concurrent increase in PE, although both increase in net amounts (133-135).  The higher PC but lower PE in the human newborn is consistent with the slower development, compared to the rat (133). Table 1.13 summarizes the temporal lipid changes in the developing rat brain. It is important to acknowledge the analytical limitations at the time this work was done and published in 1967, (128).  With no recent data on developmental lipid changes in the rat brain, it is reasonable to reassess the accuracy of these data, before using them as reference values for intervention 35  outcomes on rat brain development. Of note, gangliosides, cerebrosides, and cholesterol also increase with development but are not discussed, within the scope of this research work. Table 1.13 Temporal changes in rat brain lipids with development   Concentration (µmoles/g wet weight)  3 days 6 days 12 days 18 days 24 days Phospholipids PC Diacyl-PE Pls-PE  PI PS  14.7 5.25 2.19 1.21 2.91  14.8 5.66 2.75 2.75 3.56  20.4 7.96 4.73 1.59 4.51  24.4 9.37 7.02 1.86 6.10  24.8 11.0 11.3 2.04 7.04 Sphingolipids Sph Cerebrosides Gangliosides  0.23 0 0.82  0.26 0 2.0  1.04 2.30 2.29  2.15 5.80 2.70  3.19 10.3 2.98 Cholesterol 10.7 12.6 22.6 32.2 38.3 Adapted from Wells, 1967 (128).  Values are µmoles/g wet weight of total brain; brains from 6-9 rats were pooled and analyzed for each time point.  PC, phosphatidylcholine; PE, phosphatidylethanolamine; Pls-PE: ethanolamine plasmalogen; PI: Phosphatidylinositol; PS, phosphatidylserine; Sph: Sphingomyelin.   1.2.9.1.3 Biological Functions of Plasmalogens in the Brain Plasmalogens are important membrane PL, as evident by the drastic consequences of genetic diseases of plasmalogen deficiency, including intellectual disability, retinopathy, skeletal abnormalities, respiratory problems and facial deformation (136). Besides being structural components of cell membranes, plasmalogens play a variety of biological roles, including membrane fusion, anti-oxidation and as source of long-chain polyunsaturated fatty acids, all of potential relevance to the brain. Much of these roles, however, are investigated within the context of aging, Alzheimer’s disease, neurodegenerative diseases and peroxisomal diseases (44, 102, 137, 138) and remain to be fully understood, particularly during infant postnatal development.  36  Plasmalogens facilitate rapid membrane fusion because they tend to form non-lamellar structures, compared to other membrane PL (137, 139). Conversely, membranes frequently undergoing vesicular fusion, such as synaptic membranes, are enriched in plasmalogens (140). Membrane fusion is essential for synaptic transmission (141) and hormone exocytotic secretion (142), both critical for normal brain function.  Because plasmalogens are the richest PL in long chain polyunsaturated fatty acids, their importance has been greatly attributed to their role as the largest endogenous providers of PUFAs (137), for cell signaling, regulation of gene expression, and lipid mediator synthesis. However, their importance may not only be limited to their significant quantitative contribution to endogenous PUFAs. It is intriguing that cytosolic phospholipase A2 for PE in neurons and astrocytes is Pls-PE-selective (143, 144), which provokes speculation for potential specificity of function of PUFAs derived from Pls-PE, compared to PUFAs from diacyl-PE. This becomes important to understand because during Pls-PE deficiency, the proportion of diacyl-PE from total PE in the brain increases, as a homeostatic mechanism to conserve the total amount of ethanolamine-containing PL in the brain (145, 146). This is accompanied by a remarkable increase in ARA esterification at the sn-2 position of diacyl-PE at the expense of lower DHA (145). It is well-known that DHA in the liver, brain and retina of patients with plasmalogen deficiency is drastically reduced, with an increase in 22:5 ω-6 in the brain of some patients based on studies published in 1999 or earlier (146, 147). The functional implications of these shifts in ethanolamine-PL proportions and their ω-3 and ω-6 fatty acid contents in the brain are unknown.   Plasmalogens have also been implicated in HDL-mediated cholesterol efflux (HDL-MCE), such that human macrophages and fibroblasts with impaired plasmalogen biosynthesis have reduced HDL-MCE (148). Interestingly, the amount of PUFA-enriched Pls-PE in the cell 37  membrane seems to predict the extent of cholesterol esterification in the cell, required for HDL-MCE, with a higher free-to-esterified cholesterol ratio in cells deficient in PUFA-enriched Pls-PE (149). While this may not be directly relevant to brain cholesterol per se, which is only synthesized in situ and present primarily in a free form, there are speculations that changes in cholesterol balance across the whole body may alter sterol recycling in the brain, with potential implications for neuron and myelin membrane integrity (150).  The role of plasmalogens as antioxidants is debatable, but if true this could be provide a plausible mechanism for the functional relevance of the MFGM plasmalogens in the infant gut, particularly in protecting long-chain PUFAs from oxidation. Early studies in 1987 showed that the vinyl–ether bond at the sn-1 position of plasmalogens is more susceptible to oxidation than the sn-1 acyl bond in diacyl-PL (151).  It is postulated, therefore, that plasmalogens act as sacrificial oxidants, protecting the sn-2 fatty acids, commonly a LC-PUFA, and other lipids from free radical oxidation and iron-induced PUFA peroxidation (152, 153). Areas in the rat brain with the highest plasmalogen content, including myelin, are the least susceptible to induced oxidative damage (154, 155), which suggests plasmalogens can also act as membrane antioxidants. Whether plasmalogens in the MFGM have antioxidant properties in the infant gut, protecting free fatty acids and other lipid digestion byproducts from oxidation is not known.   Plasmalogens have been used as markers of synaptogenesis and have been shown to increase beginning 32nd wk gestation to four months postnatal in the human brain (135). The extent to which milk plasmalogens contribute to plasmalogens in the developing brain is not known. A purified form of milk plasmalogens is not available to investigate their biological roles in vivo, with only a limited number of studies on plasmalogen digestion, including plasmalogen extracted from bovine brain (115, 117, 120).  38  1.2.9.2 Relevance of Several Dietary Components of MFGM to the Brain  Experimental studies in vitro have shown the involvement of MFGM components, including LC-PUFA, sialic acid (SA), gangliosides, sphingomyelin, and choline in brain function and development, with several reviews available (5, 156-159). Unlike dietary essential nutrients such as LA and ALA, MFGM components are synthesized de novo, which complicates understanding of the importance of their dietary forms for the brain under normal physiological conditions. Many studies in animals and even humans are usually referenced in support of the role of a dietary source of these components in brain function and development. However, most of these studies showing positive outcomes are done in disease conditions, when de novo synthesis of the respective compounds is limited, inefficient, or absent (160, 161). For example, dietary Sph has been shown to restore myelin structure in the brain of developing rats, unable to produce ceramides de novo (161). Ceramides are precursors for cerebrosides, important components of myelin and well-known biomarkers for myelination (162). When de novo synthesis of ceramides is limited, an alternative pathway becomes active and involves the use of Sph as substrate. It is intriguing that dietary sphingomyelin could impact brain glycolipids and the myelin structure, given that no evidence exists for transfer of sphingomyelin from lipoproteins into the brain. Nonetheless, this would be relevant to the human infant if the activity of de novo ceramide synthesis in the brain is low during normal postnatal development, which is not clear. Therefore, extrapolating evidence of benefit of a dietary compound, also synthesized de novo from in vitro studies or from in vivo models of disease should be made with caution, when addressing normal physiology. Interesting findings from a few studies, however, link dietary MFGM components to postnatal brain development under normal physiological 39  conditions, with none on PL (163-166). Examples of animal studies are summarized in Table 1.14.  Table 1.14 Selected studies linking MFGM components to postnatal brain development   Component Animal Model Intervention of interest Major findings1 Reference Sialic acid (SA)  Neonatal piglets 0, 140, 300, 635, or 830 ml/L SA2  from d 3-d 34  postnatal  1) Improved learning performance and memory 2) ~10% higher brain SA2 btw 0 and 140 mg/L groups.  3) 2-3 fold higher mRNA of genes involved in SA brain metabolism  (163)  Sprague-Dawley rats Oral or IP administration of 0 or 1-1.2 mg/d SA3 From d 15- d21 postnatal Higher SA in cerebral and cerebellar gangliosides and cerebral glycoproteins  (164) Gangliosides (GS)      Sprague-Dawley rats Semi-purified diet with 0.2% w/w GS or no GS from d 18-d 32 postnatal Increased total GS in brain, intestine and plasma (165)  Artificially reared Long-Evans rats 0 or 24 mg/L GS from d5-d18 postnatal No effect in activities on short-term special memory (166) Choline      Sprague-Dawley rats ~0 or 1.75 mg oral choline4 from d0 to d24 postnatal Improved scores on memory tests in adulthood (167) 1 findings expressed in intervention group compared to control group, unless otherwise noted. 2Protein-bound Sialic acid 3free crystalline form of sialic acid  4 choline chloride; btw: between; IP: intra peritoneal    Despite evidence of changes in the brain after supplementation with these compounds, the extent to which these findings apply to the human infant, when fed as part of the MFGM rather than independently, needs to be better understood. For example, although dietary SA has 40  been shown to contribute to SA in the brain and improve brain function (163, 164), experimental SA was provided in a free or protein-bound form, whereas in human milk, only 23% of SA is protein-bound, 3% is free, with the majority (~74%) bound to oligosaccharides (168, 169) in the milk aqueous phase. The proportion of SA, therefore, in the MFGM as protein or lipid- bound is minimal and whether it is biologically functional solely or as part of the molecules it is present in is not known. It is also important to assess the experimental design of studies used to extrapolate evidence of biological role of milk components. For example, despite accumulation of SA in the brains of young rats (15-21 d) fed SA intra-gastrically (164), these rats were not artificially-reared, and therefore received mother’s milk which contains SA, as their source of nutrition, before weaning. As such, benefits from supplementation with oral SA in addition to milk SA typically available to the infant cannot be extrapolated to benefits of physiological levels of milk SA, within the scope of this study design. A similar and more complicated dilemma exists when studying choline. Section 1.3 of the literature review will address choline, but of relevance here, choline present in human milk is almost entirely (98%) as water-soluble compounds (free choline, phosphocholine, glycerophosphocholine), with ~2% bound to PC in the MFGM. Studies addressing the role of choline in brain development (167, 170, 171) use a free choline form, whose biological role may be very different than choline in the MFGM PC.   Many dairy processing companies have recently started isolating complex lipids from bovine MFGM as functional food ingredients, with particular interest in infant formula. These include Synlait (New Zealand), Arla Food Ingredients Group P/S (Denmark) and Fonterra Co- operative Group limited (New Zealand), producing Lipidex Phospholipid-Rich Powder, Lacprodan PL-20, and Phospholipid Concentrate 700 respectively. Lipidex is a milk powder (24.1 g protein, 26.6 g fat and 41 g carbohydrate per 100 g), containing phospholipids (5-7%), 41  gangliosides (0.45%), ceramides (0.8%) and sialic acid (0.45%) (http://www.synlait.com/site/uploads/2011/10/Lipidex-Phospholiph-rich-Powder2.pdf). Lacprodan PL-20 contains 57 g protein, 7 g lactose, and 24 g lipids per 100 g powder, of out which 20 g are phospholipids: 5.4, 5.0, 1.6, 4.8, 2.4, and 0.8% for PC, PE, PI, Sph, PS, and other phospholipids, respectively. The powder also contains gangliosides (0.7%) and ceramides (1.3%) (172). It is claimed that Phospholipid Concentrate 700 contains “phospholipid levels 5000 times that of native milk, including high levels of sphingomyelin” (https://www.fonterra.com/global/en/NZMP+Ingredients/Our+Ingredients/Specialty/Complex+Lipids/) To date, results from three animal studies (173-175) and one clinical trial (176) have been published on the role of MFGM lipids in brain and cognitive development. In artificially-reared neonatal piglets fed formula containing 0% (control), 0.8%, or 2.5% Lacprodan PL-20 from 2-28 days postnatal, piglets in the supplemented groups performed better in tests of spatial learning, independent of supplement dose, compared to the control group (173). Mean brain weight was 5% higher in both supplemented groups (0.8 and 2.5%) than the control group, with no difference in mean body weight between the groups. MRI at day 28 postnatal followed by brain region volume estimation and morphometry analysis showed several areas in the right and left cortex and cerebellum, with more gray and white matter in the supplemented compared to control group (173). Significant differences in 25 hippocampal metabolites, related to energy and PC metabolism were found between the 2.5% supplemented group only and the control group. These include lactate, malate, and 1-Linoleoylglycerophosphocholine (173). Vickers et al (174) and Guillermo et al (175) supplemented infant rats with MFGM complex lipids (Fonterra; New Zealand), at 0% (control), 0.2% or 1% w/w of estimated milk (between days 10-22 postnatal) or 42  chow (between days 23-80 postnatal). Rats were not artificially-reared, but left with their dams during the pre-weaning period, with the supplement administered by oral gavage. During the post-weaning period, the supplement was added to the chow in gel form.  The supplement was 80% lipid, out of which 6% were gangliosides, 51% were PL and 24% were neutral lipids.  No significant differences were found in total ganglioside and PL brain concentrations, as well as their respective individual species (i.e PC, PI, PE, PS and SM) between the three groups (174). Supplementation, independent of dose, tended to improve acute learning but not long-term memory acquisition (174). Again, considering that the rats were all mother-reared, it is possible that any “dietary requirement” for the MFGM lipids was met from mother’s milk, and an added functional benefit may not be present. It is however intriguing that although no definitive changes in cognitive outcomes were found, supplementation did alter synaptic function, particularly in dopamine output, with speculations on either increased synapse activity or increased synaptogenesis (175). In a pilot study on human infants (2-8 wk old; n=29-30/group), randomized to receive either a formula with the same Fonterra supplement above or standard formula (additional 15 mg PL and 3 mg gangliosides/100 g formula in the supplemented formula) until 6 months of age, supplemented infants had significantly higher hand and eye coordination IQ scores, performance and general IQ scores on the Griffith Mental Development Scale at 6 months (176). Their scores were closer to those of a reference group of exclusively-breastfed infants (n=32) than to the infants consuming the standard formula, except for general IQ.  1.2.9.3 Clinical Studies on MFGM Vegetable oils are the major source of fat in infant formula and intravenous lipid solutions. It may seem counterproductive that milk fat, including the MFGM is discarded during 43  the manufacturing of infant formula, and fat is added in the form of vegetable oils. Despite de novo synthesis of MFGM components, it is reasonable to question their dietary importance, given their presence in human milk and variability across lactation.  The recent availability of food-grade bovine MFGM has now made possible clinical trials to investigate the role of the entire globule membrane in infant health (Arla Foods Ingredients Group P/S, Denmark and Büllinger SA, Belguim).  To date, findings from four clinical studies have been published, for a range of health outcomes and age groups (177-180). Billeaud et al (177) reported adding MFGM to standard infant formula is well-tolerated in infants aged 14 days-4 months, with weight gain similar to standard formula-fed infants. Timby et al (178) reported Swedish infants fed formula supplemented with bovine MFGM (n=71;4% total protein by weight from MFGM) from 2 to 6 months scored 4.0 (95% CI: 1.1, 7.0) points higher on cognitive testing at 12 months using the Bayley Scales of Infant and Toddler Development than infants fed standard formula (n=64), with the former scores being similar to those of a reference group of breastfed infants (n=70). In this study, the formula-fed infants were stratified by sex first, then randomly assigned to one of the two formulas, such that the intervention was blinded to both staff and parents. Of note, the two experimental formulas also differed in energy and protein contents, which were lower in the MFGM supplemented formula compared to the standard formula (60 and 66 kcal/dL; 1.20 and 1.27 g/dl protein, respectively) Total caloric and protein intakes however were not different between the two groups, explained by a compensatory increase in the volume of formula consumed by the MFGM-supplemented group. The cumulative incidence of acute otitis media during the six months intervention was also significantly lower in the MFGM group than the control group (1%; 9%, respectively), with a concurrent lower incidence of antipyretic use (25%, 43%, P=0.021) (180). Infants in the MFGM 44  group had lower total serum cholesterol at 4 and 6 months compared the control group, but not different to the reference group of breastfed infants (179). In older Peruvian infants, 6-12 months old, prevalence of diarrhea was significantly lower in infants receiving a MFGM-fortified complementary food (cornstarch, whey protein concentrate and MFGM) for 6 months compared to infant receiving a complementary food without MFGM (cornstarch and skim milk protein) (3.84%, 4.37%, respectively) (181). The likelihood of having bloody diarrhea was 46% lower in the MFGM-supplemented group, with no difference in incidence of anemia or in micronutrient status between the two groups (181). Healthy Belgian children, 2.5-6 yrs old, randomized to consume chocolate milk fortified with MFGMs (i.e 500 mg PL/day) for 6 months had a lower number of parent-reported days (d) with fever (>38.5°C) at 6 wks than control children (1.71 ±2.47 d, 2.60 ± 3.06 d, P=0.028) (182). No difference in incidence of diarrhea, constipation and cough was found between the two groups. In summary, clinical studies show promising results for a role of the MFGM in intestinal health during infancy and childhood. The finding on improved cognitive outcomes due to MFGM in infants (178) merits further investigation in order to understand which stage of brain development is affected and by which mechanism (s) of action.  To summarize, milk is a source of complex lipids, including plasmalogens in the MFGM. Brain development is a complex process, for which morphological changes are accompanied with changes in brain PL and fatty acids, particularly an increase in brain PE and DHA. The MFGM may be involved in brain development. Several gaps in the literature relevant to this dissertation exist: The fatty acid composition of milk plasmalogens is not known. The biological importance of milk plasmalogens and their fatty acids are not known. The role of the MFGM in infant brain development is not known.  45  1.3 Literature Review-Section Two Given the diversity in nutrients in human milk, it is unlikely that milk lipids alone play a role in development, including the brain. This is becoming more appreciated as we learn more on the complexity of human milk nutrients, such as their different forms and arrangements within the milk matrix. A good example to address this complexity in relation to the infant brain is the nutrient choline (trimethyl-beta-hydroxyethylammonium), a quaternary amine with three methyl groups covalently bonded to a nitrogen atom (183, 184). As a summary, choline is an essential nutrient present in human milk predominately as the water-soluble forms free choline (FC), phosphocholine (PhosC) and glycerophosphocholine (GPC). This is distinct from lipid soluble choline compounds PC and Sph, in the MFGM. Choline has definitive roles in development, particularly as part of PC and SM for the formation of new cell membranes including synaptogenesis and myelination, and the production of the water-soluble neurotransmitter acetylcholine (8, 156). Why the mammary gland supplies choline to the infant largely in water-soluble forms is intriguing, particularly because choline in circulation and body tissues is predominantly lipid-bound (185, 186). Of relevance to lipids, it may be difficult to attribute a role of a dietary PL, in this case, PC and Sph in the MFGM in development to the molecules themselves when one of their building blocks (choline) is known to be also involved in development. Therefore, an understanding of choline compounds in milk, including their variation and determinants is necessary to begin to understand how these compounds alone or with an interaction with the MFGM contribute to brain development.  1.3.1 Choline in Human Milk Choline in human milk is delivered to the infant in different forms, with the water-soluble choline compounds (WSC) contributing to around 90% of total choline compounds in milk 46  (Figure 1.7). The lipid-soluble PC and Sph in the MFGM on average account for the remaining 10%. The net choline content in milk, after accounting for the molecular weight of the molecules in which choline is present, originates almost entirely (96%) from WSC (185-189). However, whether the MFGM and WSC compounds have similar functional properties to the infant is not known.  Figure 1.7 The structures of water soluble choline-containing compounds in human milk   The current literature on human milk choline compounds describes a wide range of amounts for total choline and choline forms, with the majority of the studies done on term milk (Table 1.15). The wide variation in lipid-soluble compounds, PC and Sph has been discussed in previous sections but briefly, is secondary to the limitations in the analytical methods for their extraction and quantification as well as reporting data as % rather than absolute amounts. Our group has recently shown that the variability in WSC compounds in milk is not explained by collection and storage conditions (i.e, sampling, time, temperature) as previously speculated (4), raising the question of a potential role of maternal diet and/or differences in de novo choline-containing compounds synthesis in the mammary gland.      Free choline Phosphocholine Glycerophosphocholine 47  Table 1.15 Composition of choline compounds in mature term and preterm milk     Year/ Place Zeisel et al (187)  1986/USA Holmes-Mc Nary et al (188)  1996/USA Holmes et al (189)  2000/England Ilcol et al (185)  2005/Turkey Fisher et al (186)  2010/USA                                                    Term milk  Sample size, n 10 16 8 95 48 FC PhosC GPC PC Sph Total  73 ± 21 - - 140 ± 32 188 ± 31 -    116 ± 88 570 ± 544 362 ± 280 82 ± 24 124 ± 36 12541  210 ± 141 480 ± 198 410 ± 226 100 ± 28 100 ± 28 1280 ± 396  V V v 228 ± 97 551 ± 32 499 ± 155 104 ± 107 94 ± 88 1476 ± 468  83 ± 54 551 ± 322 388 ± 168 107 ± 47 67 ± 27 11961   Other milk types      Preterm milk Term colostrum Term colostrum  Sample size, n   -17 8 21  FC PhosC GPC PC Sph Total    98.0 ± 186 693 ± 487 379 ± 173 90 ± 54 104 ± 37 1364  210 ± 141 480 ± 198 410 ± 226 100 ± 28.2 100 ± 28.2 1280 ± 395 132 ± 96 93 ± 119 176 ± 59.5 146 ± 82.0 129 ± 59.5 676 ± 160  Values are means ± SD (µmol/L). SD is calculated from published standard error for means and sample size. 1 Total is calculated from means of individual choline compounds. FC, free choline; PhosC, phosphocholine; GPC, glycerophosphocholine PC, phosphatidylcholine; Sph, sphingomyelin.  1.3.2 Recommendation for Infant Intake of Choline  Choline was established as an essential nutrient by the Institute of Medicine in 1998, based on evidence of insufficient hepatic de novo synthesis in humans (190). Recommended intakes were set for all age groups as Adequate Intake (AI). The AI for infants (0-6 mo: 125 48  mg/d and 7-12 mo: 150 mg/d) was based on the average total choline concentration in milk, from healthy and well-nourished lactating mothers, assuming these women would provide milk with adequate nutrient composition that supports infant growth and development. Despite the different forms of choline in milk and their diverse concentrations (Table 1.14), the establishment of the AI was only based on the total choline content in milk independent of choline form, from two papers published in 1986 and 1996 and using data from a total of only 26 milk samples (187, 188).  It is not known whether the different forms of choline in the diet differ in bioavailability and metabolic fate. However, it is intriguing that formula-fed infants have plasma free choline concentrations around half that of breast-fed infants (10.8  2.42; 21.8  7.61 mol/L, respectively), despite the similar total choline content in infant formula and human milk (185). Interestingly, unlike human milk, choline in infant formula is predominantly in free form, raising the question of differences in bioavailability and physiological importance of the different forms of choline in the infant diet. 1.3.3 Biological Roles of Choline: Emphasis on the Brain  Choline is an essential nutrient whose diverse biological roles can be categorized into three areas: maintenance of cell membrane structural integrity and signaling functions when bound to membrane PC and Sph, donation of methyl groups (CH3) via betaine in the liver and kidney particularly for the generation of SAM (i.e involved in DNA methylation, PC synthesis and neurotransmitter metabolism), and cholinergic neurotransmission in central and peripheral nervous system in which choline is acetylated to acetylcholine (191, 192). The importance of choline for normal brain development during the perinatal period is known, although the underlying mechanisms remain incompletely understood especially in the 49  postnatal period (156, 193). During prenatal brain development, at least in rodents, choline is important for the closure of the neural tube (194, 195), angiogenesis (196), neurogenesis, particularly for synthesis of cholinergic neurons, and for preventing neuronal apoptosis (197-199). The impact of choline during fetal development extends to brain function during adulthood, such that the offspring of choline-supplemented pregnant rodents have improved visuospatial memory than non-supplemented rodents (167, 200, 201). In this case, memory function correlates with changes in brain cholinergic activity, including hippocampal levels of choline acetyltransferase and muscuranic receptor density (167). In humans, choline is transported across the placenta, with fetal and newborn free choline in plasma being much higher than typically found in adults (202). Unlike in rodents, evidence of alteration in human offspring brain development due to low choline supply in utero (i.e insufficient maternal dietary choline) is limited. Interestingly, however, Wu et al (203) reported cognitive scores at 18 months of age in humans correlate positively with maternal free plasma choline and betaine concentrations at 16 wk gestation. The mechanisms by which choline affects postnatal brain development, including differences based on the dietary form of choline in human milk are not fully understood. However, the postnatal developmental window mostly sensitive to choline is believed to be synaptogenesis in the hippocampus and basal forebrain (156, 204, 205), which in the human continues rapidly to about 5-7 years (124). Similar to prenatal choline exposure, choline supplementation of infant rats enhances visuospatial memory in adulthood (167). Choline is known to be essential for normal cholinergic nerve functioning, as a precursor for ACh. ACh is a major excitatory neurotransmitter, with roles in learning and memory such as long-term potentiation and new memory encoding (206). ACh is synthesized via choline acetyltransferase 50  at the presynaptic terminal. After binding to the postsynaptic membrane and propagating a nerve impulse, ACh is released back into the synaptic cleft, where choline is released by acetylcholine esterase and taken up by the presynaptic neuron for recycling. De novo synthesis of choline in the brain is insufficient to sustain its cholinergic functions, with plasma choline transport across the blood-brain-barrier (BBB) (carrier-mediated and saturable) being crucial for maintaining brain choline homeostasis (207). Choline may then be incorporated in membrane PC, as a potentially major choline storage pool for acetylcholine synthesis (8). Free choline, as well can be taken up by high-affinity sodium-dependent transport across the presynaptic membrane, to be used for ACh synthesis (207). Therefore, choline availability determines the rate of acetylcholine synthesis in the brain, with a well-known increase in acetylcholine release into the synapse with higher acetylcholine synthesis (8, 208).  In addition to its role as a precursor for ACh in the brain, choline is a precursor for synthesis of membrane PC, via the CDP-choline pathway. An alternative pathway for PC synthesis by methylation of PE occurs in the brain in conditions of low CDP-choline availability (209). However, the extent to which this pathway “drains” brain PE and ethanolamine pools is not known and should be considered, given that the rate of PE synthesis is dependent upon the availability of ethanolamine. Choline can also impact infant brain development by altering gene expression via DNA methylation. This has been mainly studied prenatally, such that choline deficiency inhibits cell proliferation via DNA hypomethylation in rodent fetal brain (8). Given that DNA methylation continues during early postnatal brain development, although much slower that in utero, (210) it is reasonable that choline may regulate gene expression postnatally.  An alternative plausible reasoning to why the human infant is provided with WSC rather than lipid-soluble forms may be related to metabolic efficiency. Having enough free choline and 51  phosphocholine in the liver might be important to ensure adequate PC synthesis via the CDP-choline pathway, thus limiting the use of PE and methyl groups as substrates for PC synthesis. Notably, the predominant PC synthesis pathway in the intestine in the CDP-choline pathway. (211, 212) .In the adult liver, the CDP-choline pathway produces an estimated 70% of newly-synthesized PC (213, 214). But, in the absence of adequate choline, its de novo synthesis of increases, with a three-step PE methylation pathway by the enzyme phosphatidylethanolamine methyltransferase (PEMT) and using SAM as methyl donor (Figure 1.8) (4, 184, 215-220). If true, the metabolic advantage of providing the infant with a water-soluble form of choline for hepatic PC synthesis could be a significant, as the liver is expected to have the highest requirement for PC due to its role in secretion of lipoproteins and bile lipids.   To summarize, choline plays important roles in brain function and lipid metabolism but to date, it is not known whether dietary choline forms, particularly in milk, contribute to different biological functions in the infant. The gap in the literature most relevant to this dissertation is that available data on human milk WSC is limited by the small number of studies with only one on preterm milk, with generally small sample sizes and different methodologies used, making comparisons difficult.         52  Figure 1.8 Overview of phosphatidylcholine synthesis pathways     1.4 Research Rationale and Objectives 1.4.1 Rationale and Objectives for Chapter 2 Rationale: Plasmalogens are quantitatively minor lipids in milk, present in the MFGM, and their biological functions during infancy are not understood. Milk lipids are almost entirely (~98%) TG, and analysis of total milk fatty acids is often used when relating infant dietary fatty acid intake to outcomes of development, such as DHA in cognitive function. Analyzing total milk fatty acids, however, reflects the fatty acids in the predominant TG. Functional DHA is enriched in PL and not TG in body tissues, with particular enrichment of brain DHA in Pls-PE compared to diacyl-PE and PC. Whether a similar fatty acid pattern is present in milk is not known. It is challenging to analyze milk plasmalogen fatty acids using current methodologies due to the overwhelming presence of TG and difficulty in separating plasmalogens from their diacyl 53  counterparts. Developing methodology to separate milk plasmalogens is necessary to enable investigation of their fatty acid composition. In addition, DHA in cow milk total lipid is negligible, compared to human milk yet both species accumulate high amounts of DHA in the brain (221-224)  Analyses of milk plasmalogen fatty acids from the cow, ideally also in colostrum would help elucidate their importance as potential sources of LC-PUFA. Objectives: 1) To develop an analytical method for the separation and recovery of milk Pls-PE for fatty acid analysis 2) To determine and compare the fatty acid composition of Pls-PE to diacyl-PE and PC fatty acids in mature human milk and cow colostrum and mature milk 3) To explore the potential role of maternal DHA intake in DHA in human milk Pls-PE, diacyl-PE and PC.  Null Hypotheses: 1) Separation and recovery of milk Pls-PE for fatty acid analysis will not be achieved. 2) The fatty acid composition of milk Pls-PE will not differ from that of diacyl-PE and PC in mature human milk and cow colostrum and mature milk. 3) Maternal DHA intake will not be related to DHA in human milk Pls-PE, diacyl-PE and PC. 1.4.2 Rationale and Objectives for Chapter 3 Rationale: Chapter 2 demonstrated that milk Pls-PE can be separated and recovered from other milk lipids. In addition, the major finding in Chapter 2 was that milk Pls-PE contains a higher proportion of LC-PUFA, including DHA, than diacyl-PE and PC raising the question of whether DHA in these PL or the PL themselves have functional roles relevant to brain development, 54  different than milk DHA in TG. The analytical method we developed was not intended for or suitable as preparative method for plasmalogens, for which a purified dietary source is not currently available. At the time this research was ongoing, the MFGM became commercially available, and interest in the MFGM as a bioactive component important for infant cognition began to emerge. The MFGM not only contains Pls-PE, but also contains diacyl-PE, cholesterol, choline compounds, gangliosides, and glycoproteins, all of which have roles in brain development. Infants fed formula with MFGM from 2-6 months performed better than unsupplemented infants on cognitive tests at 12 months and had cognitive scores closer to a reference group of breastfed infants (178). However, whether the MFGM has specific effects on the brain is not known. To address this, the purpose of this study was to determine if brain PL, fatty acid and metabolite compositions differ when MFGM is included in the infant diet, using artificially-reared neonatal rats and an exploratory group of mother-reared rats.   Objectives: 1) To use artificial feeding as experimental approach to feed neonatal rats formula with or without MFGM  2) To assess whether the MFGM contributes to infant brain development by comparing brain PL, fatty acid and metabolites compositions between brains of rats fed formula with or without MFGM. Null Hypotheses: 1) The use of artificial feeding will not succeed in maintaining and feeding neonatal rats formula with or without MFGM. 2) Brain PL, fatty acid and metabolite composition will not be different between brains of rats fed formula with or without MFGM 55  1.4.3 Rationale and Objectives for Chapter 4 Rationale: Chapter 3 demonstrated that brain lipids and metabolites differ due to feeding MFGM to neonatal rats, suggesting the MFGM has direct effects of the developing brain. The findings in Chapter 3 raise several questions including which component (s) in the MFGM mediate (s) these effects and how. Concurrently to this work, interest in milk choline was arising as our research group was investigating the role of perinatal choline supply in infant cognition (203). Although choline is present in PC of the MFGM, the majority of choline in milk is in water-soluble form as FC, PhosC and GPC, with limited and inconsistent data on their amounts in milk (ie. particularly preterm milk). Their biological importance is not known, although choline per se is involved in neuro-transmission and function via acetylcholine, PC and Sph.  Two main key points became apparent: 1) The potential importance of milk WSC compounds in brain development when studying the role of the MFGM in the brain needs to be acknowledged and addressed. 2) The first step towards elucidating the biological roles of the milk WSC is to determine their concentrations in term and preterm human milk using current methods. Of note, access to BC Children’s Hospital neonatal intensive care unit (NICU) gave us a unique opportunity to collect and analyze term donor milk. These milk samples are pooled and pasteurized at the human milk bank and fed at the NICU when mother’s own milk is not available or sufficient in volume. Whether their WSC content differs between preterm and term donor milk is not known. The data presented in Chapter 4 was the first we analyzed as part of this dissertation work in 2012, and other members of our research team are addressing the biological roles of milk WSC.    56  Objectives: 1) To determine the concentrations of WSC compounds in preterm and term donor human milk fed at BC Children’s Hospital NICU using isotope dilution liquid chromatography tandem mass spectrometry (LC-MS/MS) 2) To compare the concentrations of WSC compounds between the preterm and term donor milk samples Null Hypotheses: 1) The concentrations of WSC compounds will not be determined in preterm and term donor milk samples. 2) The concentrations of WSC compounds will not be different between the preterm and term donor milk samples.               57  Chapter 2. Long Chain Polyunsaturated Fatty Acids: Where are They in Milk Lipids? 2.1 Introduction Human milk, the only source of nutrition in exclusively breastfed infants, is a complex biological fluid that facilitates growth and development in a much more sophisticated way than simply providing energy and essential dietary nutrients. Among the complex milk components, the PL are structurally diverse, differ in fatty acids, and are present in milk in unique and unusual tri-layered membranes of PL known as the milk fat globule membrane (MFGM).The MFGM surrounds and emulsifies the milk TG (29), which are the largest component of milk lipid, with an average of 34 g/L representing >95% of the lipid in mature human milk (41). Although PL have been considered minor because of their ~1% contribution to total milk lipid (3), it is remarkable that the term infant is provided with an estimated 350 mg/d and up to 610 mg/d of PL, assuming an average milk consumption of 780 ml/d and a milk PL content of 450 mg/L (range: 140-780) (41, 225).   In recent years, research in infant fatty acid nutrition with milk or milk alternates has typically measured fatty acids in total milk lipids (53, 54, 56, 58, 64, 226, 227), perhaps explained in part by the low PL relative to TG, and complexity and inaccuracies of separating different PL using traditional lipid extraction and separation techniques. Nonetheless, from these studies, milk long chain ω-6 and ω-3 fatty acids (LC-PUFA), including arachidonic acid (ARA, 20:4 ω-6) and docosahexaenoic acid (DHA, 22:6 ω-3) have been implicated in the development and function of neurological and immune systems (6, 225, 228). Importantly, LC-PUFA are enriched in cell membrane PL rather than TG, with enrichment also showing specificity in PL class, organ and membrane (100, 103, 122, 123, 229, 230). The human brain, for example increases in size from 25 % at birth (~380 g) to around 70% (~950 g) of the adult brain weight (~ 58  1400 g) at the end of the first year of life, with this data published in 1978 (129). In addition, there is also a marked increase in DHA during this period, particularly in ethanolamine-containing PL, known as plasmalogens (Pls-PE) (122, 123, 129). The best known PL, the diacyl-PL, have a glycerol backbone with fatty acids esterified at the glycerol sn-1 and 2 carbons (C) through an ester bond.  Least well-understood, plasmalogens have the same structure except that their sn-1 fatty acid is attached to glycerol as vinyl ether, with the sn-2 fatty acid attached via an ester bond as in other PL (138). Human milk total PL are higher in LC-PUFAs than milk TG (41), but information on the amounts and types of PL in milk are limited and often inconsistent. Most earlier methods, based on TLC and simple solvent systems and then HPLC, were compromised by the large amount of TG in milk and were insufficient to separate and quantitatively recover PL, particularly Pls-PE. More recently and using 31P nuclear magnetic resonance (NMR) spectroscopy, total PL content in mature human milk was reported as 268 ± 87.6 mg/L (n=22 women), with 30 % sphingomyelin, 19 % diacyl-PE, 11% Pls-PE, 24% phosphatidylcholine (PC), 9% phosphatidylserine (PS) and 4 %phosphatidylinositol (PI) (40).  The extent to which milk PL and their fatty acids contribute to brain development is not known. The evolutionary persistence of PL in milk, including Pls-PE possibly enriched in LC-PUFA, raises the question of whether they and their fatty acids play a biological role beyond emulsion stabilization of TG in milk, as part of the MFGM. The first objective of this work, essential to enable later consideration of the importance of milk PL, was to develop an analytical method that enables recovery and separation of PL, including Pls-PE, to better understand their fatty acid composition. Once developed, we sought to analyze mature milk Pls-PE, diacyl-PE and PC fatty acids from women who had given birth after term gestation. 59  It is also intriguing that despite fundamental differences in dietary fatty acid intakes between carnivorous and herbivorous mammals, the fatty acid composition in brain gray matter is consistent across species, with ARA, 22:4 n -3 and DHA as major fatty acids and <2% LA and ALA found in other cells (221, 222, 230, 231). Cow milk for example is well-known for its low ω-6 and ω-3 fatty acids, with low LA, DHA and ARA in the total milk lipid, although cow brain shows similar high DHA and ARA (221-224) with high DHA in bovine retina (232, 233) as in humans. We thus included analysis of cow colostrum and mature milk Pls-PE, diacyl-PE and PC fatty acids, in addition to human milk, addressing potential fundamental roles of milk PL in providing ω-3 and ω-6 fatty acids.  2.2 Methods Mature human milk was collected from n=25 healthy women, providing their milk as the sole source of nutrition for term gestation infants. Women with infants ≤ 8 wk of age, not exclusively breast-fed or born < 37 wk gestation, or who were ≤ 19 years of age, or had any disease or medication likely to impact milk volume or composition were not enrolled. Mid-feed milk samples of ~10 ml were collected from each woman. Of the 25 participants, 4 women provided milk samples collected the same way on each of three separate days. Instructions on milk collection and collection vials were given, and information on maternal age, gestation length and infant delivery date, ethnicity and dietary intake were collected by skilled research staff. Dietary intake was assessed using an interview-administered food frequency questionnaire (FFQ), as in previous studies (234-236). On collection, milk samples were immediately frozen by the mothers in the vials provided to them, with the date recorded. Samples were then transferred to the lab within 3 days of collection and frozen at -800C. All aspects of the study were approved by the Committee for Ethical Review of Research Involving Human Subjects at 60  the University of British Columbia and the British Columbia Children’ s and Women’ s Hospital. Written informed consent was obtained for each participant before participation. Cow colostrum and mature milk were collected from multi-parous Holstein cows and provided by the UBC Dairy Education and Research Centre as a research donation. Nutrient intakes were analyzed using the ESHA Food Processor SQL (version 10.1.0, ESHA research, Salem, OR, 2009), with the Canadian nutrient file updated to include complete data on dietary polyunsaturated fatty acids. Total energy, fat, lipid, and carbohydrate were determined in each mature human milk sample using a Human Milk Analyzer (MIRIS, Sweden).   2.3 Statistical Analysis Subject characteristics were analyzed using descriptive statistics, and the normality of data distribution was tested using Kolmogorov-Smirnov test. Values are expressed as means ± standard deviations, or medians, interquartile ranges (IQR), and ranges as appropriate, with one-way analysis of variance following by Tukey’s post hoc test for normally distributed data and Mann-Whitney U tests for skewed data. The associations between mother’s total milk fatty acids, dietary fatty acids, and the milk Pls-PE, diacyl-PE, and PC fatty acids were analyzed using Pearson’s correlation coefficient.  Statistical analyses were performed using SPSS software (Version 20 for Windows), with the level of statistical significance set at P<0.05.   2.4 Description of the Developed Analytical Method 2.4.1 Solvents Chloroform, methanol, isopropanol, methylene chloride, diethyl ether, formic acid, hexane, acetic acid and triethylamine were obtained from Fisher Scientific (Toronto ON) or Sigma-Aldrich (Mississauga, ON) and were HPLC-grade.  61  The following authentic PL were used: PC (P3556, Sigma-Aldrich), PS (P0474, Sigma-Aldrich), PI (P5954, Sigma-Aldrich), Pls-PE (#257625, Avanti Polar Lipids) and PE (P8664 and 7943, Sigma-Aldrich).  2.4.2 Analytical Method Development Lipids were extracted from milk (5 ml) using a modified Folch extraction method (Appendix A.1) then PL isolated by solid phase extraction (SPE) by modifying a method by Kaluzny et al (237): The lipid extract was solubilized in 2 ml methylene chloride: isopropanol (2:1 v/v) and transferred to an amino propyl cartridge (Waters WAT054515), loaded on an SPE vacuum Manifold. Methylene chloride: isopropanol (2:1 v/v; 8 ml), hexane (6 ml), diethyl ether with 4% formic acid (5 ml), and methanol (5 ml) were slowly added at 1 drop/sec, with PL elution and recovery in methanol. Methanol was evaporated with a Nitrogen Multivap evaporator (Organomation, Berlin MA) and PL transferred to an HPLC vial for analysis.  The method for separation of PL was a modification by us of the method by Mawatari et al (238). Pls-PE, diacyl-PE and PC were separated by HPLC (Waters Alliance 2695 HPLC, Waters Corp, Mississauga ON) using two YMC Diol column 4.6 × 250 mm and 5 micron pore size (Chromatographic Specialties, Brockville ON), detected by an evaporative light-scattering detector (Waters 2424 ELSD, Waters Corp, Mississauga ON) and recovered manually. N2 was used as nebulizing gas at a flow rate of 2.3 L/min, at a temperature of 45°C and a drift tube temperature of 70°C. The elution program was a tertiary linear gradient solvent system with 82:17:1 (vol/vol/vol) hexane/isopropanol/acetic acid with 1% triethylamine (TEA) (mobile phase A), 92:5:3 (vol/vol/vol) isopropanol/acetic acid/water with 1% % TEA (mobile phase B), and methanol (Mobile phase C). Accordingly, the amounts of added acetic acid used here were lower than those originally used (238). PL were resolved in 75 minutes, with the solvent gradient 62  summarized in Table 2.1. The flow was maintained at 1.6 mL/min. The samples and the column were equilibrated at 50°C. Pls-PE, diacyl-PE and PC fatty acids were analyzed by routine GLC (Appendix A.2), with total milk fatty acids analyzed for comparisons (Appendix A.3) (54).  Table 2.1 Description of the solvent gradient used for HPLC analysis   Solvent A Solvent B Solvent C Time, min Percentage of solvent used, % 2 95 5 0 3 91 9 0 22 88 12 0 34 70 20 10 50 68 20 12 53 40 35 25 56 95 5 0                           Solvent A: 82:17:1 (vol/vol/vol) hexane/isopropanol/acetic acid                             Solvent B: 92:5:3 (vol/vol/vol) isopropanol/acetic acid/water                            Solvents A and B prepared with -1% Triethylamine (TEA)                           Solvent C: Methanol  2.5 Results for Analytical Method An HPLC method to separate Pls-PE from mammalian tissues, including muscle, heart and erythrocytes is available (238). However, using this method for analysis of Pls-PE in milk is complicated by the predominance of triglyceride in total milk fat, overloading the HPLC column and preventing detectable PL (Figure 2.1). We therefore used a pre-HPLC step to remove the milk TG by SPE. Two SPE cartridges were tested: a HybridSPE-Phospholipid cartridge (55257-U; Sigma-Aldrich, Mississauga ON) and an amino propyl cartridge (WAT020840; Waters Corp, Mississauga ON). SPE using the HybridSPE-Phospholipid cartridge is based on highly selective Lewis acid –base interactions between zirconia ions bonded to the hydridSPE particles and the phosphate moiety of PL, while eluting other lipids. The retained PL were eluted using 5% ammonia in acetonitrile, which resulted in an unacceptable carry-over of the ammonium salt in 63  the PL. Clean up steps with centrifugation of the PL and re-extraction jeopardized PL recovery; Accordingly, the HybridSPE was not used in following work.  Figure 2.1 HPLC Chromatogram showing no detectable phospholipid in the presence of triglyceride             Chromatogram of total lipids from human milk; TG, triglyceride, FFA, free fatty acids  SPE using the amino propyl cartridge is based on normal-phase retention, such that polar compounds can be separated from non-polar solutions, and structurally similar molecules can be eluted through successive washes with solutions of different degrees of polarity. The solvents and corresponding volumes we used (Methods Section) resulted in removal of triglyceride, with 97 % and 93% recovery for Pls-PE and diacyl-PE, respectively. The intra-assay and inter-assay coefficient of variations for each are summarized in Table 2.2. The recovery of PS and PI was inconsistent, likely reflecting inefficient extraction as well-known for acidic PL (27). Adding sulphuric or phosphoric acid to the SPE solvents would improve PS and PI recovery, however this would degrade plasmalogens, vulnerable to breakdown at the sn-1 position of glycerol. Sph, although a major component of MGFM is well-known for its saturated and monounsaturated 64  fatty acids, but low ω-6 and ω -3 fatty acid composition (43), thus Sph was not included in this work.  Table 2.2 Intra-assay and inter-assay variability of Pls-PE and diacyl-PE recovery   Intra-assay CV %  Inter-assay CV % Pls-PE 2.8  7.0 Diacyl-PE 6.0  8.0 Based on 4 replicates per day on 3 separate days, using direct GLC analysis  with C17:1 as internal standard.  CV, coefficient of variation. Pls-PE, ethanolamine-plasmalogen; diacyl-PE, diacyl-Phosphatidylethanolamine.  For HPLC analyses, two chromatography columns were used to achieve separation of Pls-PE and diacyl-PE, as well as PC and Sph, with TEA at 1% in mobile phases A and B (Figure 2.2). The collected PL fractions from each peak, for Pls-PE, diacyl-PE and PC were confirmed via their respective fatty acids by GLC. Comparison of the fatty acids of the recovered PL with no separation/recovery showed the entire procedure (SPE and HPLC) had negligible effect on the fatty acids (Table 2.3).           65   Figure 2.2 HPLC chromatograms showing separation of Pls-PE from other lipids as authentic standards and in mature human milk       PE, phosphatidylethanolamine; Pls-PE, plasmalogen-PE;  PC: Phosphatidylcholine; Sph: Sphingomyelin.  Note than the difference in retention times for different peaks between A and B are due to replacement of the HPLC column for a new but similar column. The new column was used for the analysis of all milk samples. A. Phospholipids as authentic standards B. Phospholipids in mature human milk 66  Table 2.3 Comparison of fatty acids in PE and PC before and after separation and recovery using SPE following by HPLC  Values are g/100 g fatty acid for one sample analysis per phospholipid standard. Fatty acids are presented in the table in the time order of their detection by GLC.  PE standard, Sigma P7943; PC standard, Sigma P3556.  2.6 Results for Mature Human Milk Pls-PE Fatty Acids   Twenty-six women were recruited for this study, out of which one was excluded due to an unexpected preterm delivery (<37 wk gestation) and therefore 25 lactating women were  PE standard  PC standard No separation Post SPE +HPLC  No separation Post SPE +HPLC 12:0 0.01 0.03  0.06 0.12 14:0 0.04 0.06  0.24 0.22 14:1 0.00 0.00  0.02 0.02 16:0 18.1 18.6  29.7 29.8 16:1 ω-9 0.03 0.03  0.20 0.22 16:1 ω-7 0.41 0.50  0.81 0.83 18:0 26.7 27.2  14.6 15.3 18:1 ω-9 18.9 18.4  26.7 27.7 18:1 ω-7 0.63 0.68  0.84 0.88 18:2 ω-6 12.6 12.4  17.8 16.6 18:3 ω-6 0.10 0.12  0.15 0.16 18:3 ω-3 0.12 0.14  0.20 0.05 20:0 0.03 0.04  0.05 0.09 20:1 0.38 0.30  0.31 0.29 18:4 ω-3 0.04 0.07  0.00 0.00 20:2 ω-6 0.30 0.29  0.30 0.30 20:3 ω-9 0.12 0.12  0.07 0.07 20:3 ω-6 0.45 0.45  0.38 0.39 22:0 0.00 0.00  0.02 0.00 20:4 ω-6 13.4 13.2  4.10 4.01 22:1 ω-9 0.19 0.37  0.58 0.08 20:5 ω-3 0.02 0.00  0.12 0.12 22:4 ω-6 0.62 0.60  0.23 0.21 22:5 ω-6 3.53 3.37  0.97 0.91 22:5 ω-3 0.23 0.22  0.10 0.10 22:6 ω-3 3.00 2.85  1.29 1.22 67  enrolled in the study. The characteristics are presented in Table 2.4. One participant reported no meat, poultry, fish or milk products intake; all others reported no dietary restrictions.   Table 2.4 Maternal characteristics  Study participants n=25 Maternal age (y) 32 ± 4.01 Milk production (mo) 4.0 (1.1)2 Ethnic background (n) Caucasian Latin American Asian African  16 4 4 1 1 mean ± SD; 2 median (interquartile range) y, year; mo, months                  The mature human milk provided mean ± SD 66.4±14.7 kcal/dL, with 3.72± 1.52 g/dL fat, 1.09±0.12 g/dL protein, and 6.80±0.32 g/dL carbohydrate. The proportion of LC-PUFA in milk PL is significantly higher than in milk total lipid. Pls-PE had higher 20 and 22 carbon chains (C) compared to diacyl-PE and PC, notably higher arachidonic acid (ARA, 20:4 ω-6), adrenic acid (ADA, 22:4 ω-6), 22:5 n-3, and docoxahexaenoic acid (DHA, 22:6 ω-3) (Table 2.5). Conversely, linoleic acid (LA, 18:2 ω-6) and oleic acid (OA, 18:1 ω-9) constituted around 12% and 18% of the fatty acids in Pls-PE compared to the higher 26% for LA and 24% for OA in diacyl-PE. Consistent with the well-known differences in fatty acids between different PL in circulation, particularly plasma PC and PE (239-241), the polyunsaturated 20 and 22 carbon chains in Pls-PE were also higher than in milk PC, including higher ARA, ADA, DHA, 20:2 ω-6, 20:3 ω-6, and 22:6 ω-6 (Table 2.5). With the majority of the sn-1 PC fatty acids as saturated (~48% total PC fatty acids as C16:0 and C18:0) and the known esterification of PUFAs at the sn-68  2 position of PL, PC contained a higher proportion of OA and LA at the sn-2 position, compared to Pls-PE.                     69  Table 2.5 Fatty acids in Pls-PE, diacyl- PE, PC and total lipid of mature human milk   Ethanolamines     plasmalogen-PE diacyl-PE PC Total fatty acids 18:2 ω-6 12.4 ± 2.00a 25.5 ± 3.64b 26.4 ± 3.31b 14.0 ±2.73 18:3 ω-3 0.31 ± 0.11a 0.63 ± 0.19b 0.49 ± 0.25a,b 1.65 ± 0.64      20:2 ω-6 0.48 ± 0.21a 0.41 ± 0.15a 0.24 ± 0.08b 0.24 ± 0.07 20:3 ω-6 1.37 ± 0.38a 1.03 ± 0.29a,b 0.68 ± 0.22b 0.32 ± 0.08 20:4 ω-6 10.5 ± 1.71a 3.82 ± 0.92b 1.88 ± 0.54c 0.48 ± 0.08 22:4 ω-6 2.06 ± 0.55a 0.29 ± 0.22b 0.29 ± 0.44b 0.08 ± 0.03 22:5 ω-6 0.31 ± 0.11a 0.58 ± 0.20a 0.06 ± 0.04b 0.04 ± 0.01      20:5 ω-3 0.84 ± 0.63a 0.62 ± 0.31a 0.40 ± 0.26a 0.14 ± 0.09 22:5 ω-3 2.11 ± 0.70a 0.58 ± 0.20b 0.18 ± 0.07c 0.20 ± 0.13 22:6 ω-3 2.88 ± 1.49a 1.65 ± 0.94b 0.62 ± 0.21c 0.34 ± 0.16      16:0 16.7 ± 2.18 5.77 ± 1.23 22.3 ± 2.28 19.2 ± 2.33 18:0 16.3 ± 2.89 30.9 ± 3.48 26.2 ± 3.20 6.49 ± 1.29 20:0 0.63 ± 0.29 0.20 ± 0.07 0.28 ± 0.11 0.20 ± 0.07 22:0 0.40 ± 0.21 0.39 ± 0.15 0.25 ± 0.19 0.10 ± 0.04      16:1 ω-7 6.56± 3.16 0.52 ± 0.26 0.65 ± 0.13 2.49 ± 0.76 18:1 ω-9 18.0 ± 2.91 24.4 ± 6.18 16.2 ± 2.70 38.8 ± 3.83 18:1 ω-7 0.36 ± 0.21 1.70 ± 0.58 1.43 ± 0.33 0.81 ± 1.02 20:1 ω-9 3.46 ± 1.00 0.87 ± 0.40 0.48 ± 0.20 0.02 ± 0.07 24:1 ω-9 0.12 ± 0.08 0.06 ± 0.10 0.14 ± 0.10 0.06 ± 0.02 Values are % (g/100 g fatty acid); Data from n=25 mature milk samples PE, phosphatidylethanolamine; Pls-PE, plasmalogen-PE; PC: phosphatidycholineMeans of polyunsaturated fatty acids with different letters are statistically different (P<0.05), by ANOVA followed by Tukey’s test.  Fatty acids in Pls-PE, diacyl-PE and PC found in <0.1% and not presented in the table are: C8:0, C10:0, C12:0, C14:0, C14:1, C18:3 ω-6, C18:4 n-3, C20:3 n-9 and C24:0. Total milk lipid also contains C8:0, 0.17 ± 0.04; C10:0, 1.26 ± 0.26; C12:0, 5.75 ± 1.84; C14:0, 5.82 ± 1.35; and C14:1, 0.25 ± 0.13 %.   70  Given the high variability in Pls-PE DHA found among women (mean: 2.88; median: 2.40; range: 0.95-6.51%), we explored the potential role of dietary DHA in contributing to this variability. Total milk DHA is known to vary with maternal DHA intake (242, 243). We therefore first explored the relationship between Pls-PE DHA and total milk DHA. DHA in Pls-PE was significantly correlated with total milk DHA (r2=0.56), although this should be interpreted with caution, given the limited number of milk samples with total DHA >0.5% in this group (Fig 2.3). A similar relationship was also found between DHA intake and DHA in diacyl-PE and PC, both with r2=0.62, P<0.01. Figure 2.3 Relationship between DHA in milk total lipid and DHA in Pls-PE, diacyl-PE and PC  R2=0.62 71  We then explored the effect of maternal DHA intake on milk DHA in total fatty acids and variability in Pls-PE, diacyl-PE, and PC across women. The lactating women consumed 2478 ± 721 kcal/day, with 49.7 ± 7.90, 17.3 ± 3.00, and 33.1 ± 7.53 % energy from carbohydrate, protein, and fat, respectively. Fat intake, as g/day, was skewed with a higher mean than median (92.5 ± 44.0; median: 82.4; 5th-95th: 58-136 g/day), although the distributions of total saturates (35.0 ± 7.05%), monounsaturates (41.6 ± 5.00%), polyunsaturates (22.0±4.48%), and trans fat (1.40 ± 0.77%) were normal. The dietary intakes of DHA as mg and % total fat per day was skewed, as evident by the wide standard deviations and ranges of intake (Table 2.6).  Table 2.6 Dietary polyunsaturated fatty acids intakes of breastfeeding women  Mean ± SD Median (IQR) 5th-95th percentile 18:2 ω-6, g/d                  %2 17.6 ± 7.61 19.2 ± 3.72 15.7 (6.29) 19.1 (4.82) 8.65-29.6 13.6-25.1 20:4 ω-6, mg/d                   % 156±120 1 0.16±0.07 145 (0.07) 0.17 (0.10) 46-296 0.08-0.24 18:3 ω-3, g/d                 % 2.20±1.36 1 2.30±0.81 1.81(1.31) 1.02(0.91) 0.79-5.13 4.41-1.43 20:5 ω-3, mg/d                    % 114 ± 103 1 0.12 ± 0.09 1 81.0 (88.4) 0.11 (0.10) 6.60-279 0.02-0.27 22:6 ω-3, mg/d                    % 174 ± 150 1 0.19 ± 0.13 1 125 (123) 0 (0.12) 17-354 0.03- 0.39                            1 Indicates intake distribution is skewed.                   2 Percent from total fat; n=25 women.      Note that 18:2 ω -6 and 18:3 ω -3 intakes are in g/d and other polyunsaturated                    fatty acids are in mg/d.  72   Dietary DHA had no significant correlation with its respective percentages in total milk, Pls-PE, diacyl-PE or PC (P>0.05). At least for total milk DHA, a major limitation to not finding a significant correlation with DHA intake is the small range in total milk DHA (0.1-0.8%), among our study participants. Therefore, DHA in milk PL, including Pls-PE varied among women, independent of their intakes of DHA within the ranges of DHA intakes reported here and when DHA intake is estimated using an FFQ. We therefore explored next whether DHA in these PL vary within a woman on separate days to begin to elucidate determinants of milk PL DHA.  Despite the consistent higher DHA in Pls-PE than diacyl-PE or PC across days within a woman (Figure 2.4), DHA in these PL varied within a women particularly in Pls-PE and diacyl- PE. The fatty acid composition of Pls-PE, diacyl-PE and PC across days for each woman are presented in Tables 2.7-2.10. A better understanding of the cause of variability in milk PL DHA within a woman across days and its implications on infant DHA intake, requires a better understanding of DHA intake assessment (i.e recent dietary intake versus average monthly intake as assessed by FFQ), variability in total milk lipid content and absolute content of these PL in milk within a woman. Using the total milk lipid content and total fatty acid composition (Tables 2.11 and 2.12) of each milk sample per women on 3 separate days, we found DHA content to be variable across days within a woman (#1: 119, 222, 360 mg/d; #2: 185, 81, 118 mg/d; #3: 167, 211, 195 mg/d; #4: 58, 107, 97 mg/d, on 3 separate days respectively). Given the variability in total lipid content across milk samples within a women (Table 2.11) and the variability in total milk DHA, it becomes clear that understanding the implications of the variability in DHA (g/100 g fatty acid) in milk PL, particularly Pls-PE requires addressing the variability in PL content within a women across days, potentially also within a day, in future studies. 73  Figure 2.4 Variability of DHA in milk Pls-PE, diacyl-PE and PC with individual women      DHA in mature and mid-feed human milk collected on three separate days over two wks.  #1, 2, 3 and 4 represent 4 different women.        Phosphatidylcholine (PC) DHA (g/100 g fatty acid) Plasmalogen phosphatidylethanolamine  ( Pls- PE) Diacyl phosphatidylethanolamine (Diacyl-PE) #4 #1 #2 DHA (g/100 g fatty acid) DHA (g/100 g fatty acid) DHA (g/100 g fatty acid) #3 74  Table 2.7 Fatty acid composition of Pls-PE, diacyl-PE and PC from three mature milk samples collected from the same woman (#1) on three separate days    Pls-PE  Diacyl-PE  PC*  d1 d2 d3  d1 d2 d3  d1 d2 d3 18:2 ω-6 12.7 12.9 14.7  27.7 27.3 26.3  31.2 30.3 31.8 18:3 ω-3 0.29 0.34 0.29  0.49 0.59 0.56  0.82 0.77 0.46             20:2 ω-6 0.58 0.49 0.41  0.30 0.30 0.32  0.22 0.18 0.17 20:3 ω-6 1.22 1.21 1.05  0.97 0.88 0.88  0.82 0.68 0.60 20:4 ω-6 7.18 6.61 9.88  3.46 3.59 4.11  3.29 2.76 2.18 22:4 ω-6 2.46 0.33 1.44  0.27 0.17 0.32  0.48 0.34 0.11 22:5 ω-6 0.38 0.17 0.21  0.17 0.18 0.01  0.09 0 0             20:5 ω-3 0.43 1.46 1.89  0.60 1.04 1.29  0.10 0.81 0.51 22:5 ω-3 0.38 1.61 2.36  0.66 0.56 0.76  0.66 0.37 0.21 22:6 ω-3 2.11 2.46 3.58  1.55 2.02 2.29  1.29 1.28 0.94             16:0 9.91 7.80 5.01  3.85 3.06 4.14  16.9 18.7 19.0 18:0 18.2 17.4 7.43  32.5 31.5 27.8  21.6 21.5 22.7 20:0 3.60 3.46 3.19  0.11 0.05 0.12  0.22 0.22 0.11 22:0 0.55 0.31 0.24  0.04 0 0.03  0.42 0.36 0             16:1 ω-7 6.56 9.35 11.3  0.54 0.30 0.31  1.40 0.78 0.67 18:1 ω-9 21.1 19.2 21.9  24.1 26.0 27.6  17.0 17.7 18.3 18:1 ω-7 8.49 11.3 12.0  1.68 1.88 1.79  1.21 1.48 1.36 20:1 ω-9 3.60 3.46 3.19  0.81 0.12 1.14  0 0 0.12 24:1 ω-9 0.13 0 0  0 0 0.03  0.12 0 0 Values are g/100 g fatty acid. d: day. PE, phosphatidylethanolamine; Pls-PE, plasmalogen-PE; PC: phosphatidycholine * PC also contained 1.30 and 1.0 % C14:0 at d1 and d2, respectively.       75  Table 2.8 Fatty acid composition of Pls-PE, diacyl-PE and PC from three mature milk samples collected from the same woman (#2) on three separate days    Pls-PE  Diacyl-PE  PC†  d1* d2 d3  d1 d2 d3  d1 d2 d3 18:2 ω-6 7.89 9.52 11.0  22.0 22.7 21.7  28.6 29.4 31.5 18:3 ω-3 0 0 0  0.39 0.27 0.13  0 0.29 0.36             20:2 ω-6 0.55 0.78 0.51  0.40 0.41 0.42  0.22 0.17 0.23 20:3 ω-6 1.53 1.68 1.53  1.18 1.35 1.27  0.74 0.84 0.76 20:4 ω-6 10.1 9.00 10.3  3.19 3.37 3.20  1.50 1.67 1.77 22:4 ω-6 1.97 1.83 1.58  0.23 0.19 0  0.09 0.20 0 22:5 ω-6 0.33 0.23 0.24  0.23 0.15 0.20  0.08 0 0             20:5 ω-3 1.82 0.84 0.67  0.75 0.91 0.78  0.07 0.12 0.17 22:5 ω-3 1.98 2.37 2.07  0.52 0.69 0.51  0.11 0.22 0 22:6 ω-3 4.34 3.40 3.27  2.23 2.31 1.78  0.67 0.81 0.84             16:0 9.28 8.59 12.4  5.01 4.87 5.49  22.4 21.5 22.4 18:0 25.1 23.2 25.4  32.2 31.7 34.1  23.8 23.6 21.7 20:0 1.93 2.53 1.05  0.18 0.23 0.17  0.09 0.16 0 22:0 0.33 1.22 0.43  0.07 0.08 0.07  0.07 0.32 0             16:1 ω-7 7.12 8.01 4.77  0.46 1.01 1.03  0.51 0.92 0.35 18:1 ω-9 17.8 20 19.5  27.1 25.8 25.4  18.2 16.5 17.8 18:1 ω-7 1.08 1.13 1.31  2.02 1.99 1.79  1.44 1.31 1.31 20:1 ω-9 3.98 1.35 3.50  1.19 1.62 1.37  0 0 0 24:1 ω-9 0 2.37 0.07  0.13 0.09 0.14  0 0.07 0 Values are g/100 g fatty acid. d: day. PE, phosphatidylethanolamine; Pls-PE, plasmalogen-PE; PC: phosphatidycholine  * Pls-PE at d1 also contained 1.13% C14:0 † PC also contained 1.22 and 0.89 % C14:0 at d1 and d2, respectively.  Values not adding to a 100 because of minor fatty acids including 14:0, 14:1 ω-9; 18:3 ω-6,18:4 ω-3, 20:3 ω-9 and 24:0.      76  Table 2.9 Fatty acid composition of Pls-PE, diacyl-PE and PC from three mature milk samples collected from the same woman (#3) on three separate days    Pls-PE  Diacyl-PE  PC  d1 d2 d3  d1 d2 d3  d1 d2 d3 18:2 ω-6 11.1 12.6 14.5  22.5 24.4 22.1  24.7 30.7 29.2 18:3 ω-3 0.21 0.25 0.31  0.44 0.66 0.44  0.26 0.58 0.51             20:2 ω-6 0.09 0.34 0.39  0.37 0.28 0.31  0.19 0 0.17 20:3 ω-6 1.58 1.06 1.27  1.38 1.07 0.95  0.95 0.52 0.64 20:4 ω-6 14.5 12.2 12.1  4.75 5.32 4.36  2.16 2.54 2.41 22:4 ω-6 2.29 1.65 1.95  0.52 0.46 0.57  0.12 0.12 0.20 22:5 ω-6 0.38 0.25 0.33  0.33 0.30 0.23  0.07 0.07 0.11             20:5 ω-3 0.45 1.13 1.25  0.86 0.97 1.39  0.34 0.36 0.58 22:5 ω-3 3.08 2.06 2.68  0.91 0.89 0.92  0.24 0.26 0.30 22:6 ω-3 4.99 3.01 3.72  3.11 2.41 2.27  0.95 0.86 0.97             16:0 16.6 18.1 10.9  6.23 5.08 3.90  23.0 21.8 22.2 18:0 15.2 17.4 18.0  26.7 29.3 33.8  23.4 23.9 23.6 20:0 0.16 0.85 1.18  0.26 0.18 0.20  0.35 0.05 0.40 22:0 0.03 0.75 0.38  0.08 0.20 0.23  0.18 0 0.09             16:1 ω-7 7.04 4.99 9.85  0.68 0.65 2.19  0.65 0.78 1.15 18:1 ω-9 13.7 17.1 16.0  28.7 24.7 22.8  18.6 15.2 13.7 18:1 ω-7 0.40 1.58 1.77  1.44 1.48 2.23  2.14 1.14 1.21 20:1  6.11 3.19 2.97  0.08 1.15 1.01  0.56 0 0.45 24:1 ω-9 0.19 0.07 0  0 0.04 0  0.05 0 0 Values are g/100 g fatty acid. d:day. PE, phosphatidylethanolamine; Pls-PE, plasmalogen-PE; PC: phosphatidycholine. Values not adding to a 100 because of minor fatty acids including 14:0, 14:1 ω-9; 18:3 ω-6,18:4 ω-3, 20:3 ω-9 and 24:0.       77   Table 2.10 Fatty acid composition of Pls-PE, diacyl-PE and PC from three mature milk samples collected from the same woman (#4) on three separate days    Pls-PE  Diacyl-PE  PC  d1 d2 d3  d1 d2 d3  d1 d2 d3 18:2 ω-6 14.5 15.0 19.6  26.6 25.9 27.8  30.8 30.6 31.4 18:3 ω-3 0 0.02 0.06  0.39 0.51 0.71  0.34 0.42 0.58             20:2 ω-6 0.61 0.84 0.58  0.40 0.49 0.35  0.22 0.26 0.21 20:3 ω-6 0.98 1.12 1.34  0.73 0.71 0.78  0.58 0.52 0.62 20:4 ω-6 11.6 14.3 15.1  4.58 4.37 4.28  2.58 2.29 1.88 22:4 ω-6 2.64 3.48 2.40  0.57 0.58 0.51  0.29 0.21 0.30 22:5 ω-6 0 0.40 0.32  0.22 0.23 0.24  0.23 0.21 0.25             20:5 ω-3 0.83 0.07 0.34  1.08 0.15 0.06  0.09 0 0.07 22:5 ω-3 1.90 2.25 1.89  0.73 0.60 0.58  0.23 0.21 0.25 22:6 ω-3 1.20 1.50 1.81  0.77 0.79 0.95  0.34 0.31 0.49             16:0 15.6 10.9 9.78  4.48 3.71 3.78  23.2 21.5 21.7 18:0 12.2 18.0 16.7  27.6 25.9 25.2  19.5 20.7 19.6 20:0 1.16 1.02 0.92  0.31 0.19 0.18  0.20 0.14 0.16 22:0 0.69 1.02 0.50  0.08 0.03 0.06  0.16 0.13 0.14             16:1 ω-7 5.69 1.57 1.11  0.68 0.83 0.99  0.90 0.49 0.53 18:1 ω-9 24.3 22.1 22.6  27.3 31.3 30.2  18.2 20.0 20.6 18:1 ω-7 1.36 0.94 0.73  1.62 1.88 1.87  1.26 1.27 0.20 20:1  3.58 4.78 3.80  1.53 1.70 1.29  0 0 0.17 24:1 ω-9 0 0 0  0 0 0  0 0.05 0.03 Values are g/100 g fatty acid. d:day. PE, phosphatidylethanolamine; Pls-PE, plasmalogen-PE; PC: phosphatidycholine Values not adding to a 100 because of minor fatty acids including 14:0, 14:1  ω-9; 18:3 ω-6,18:4 ω-3, 20:3 ω-9 and 24:0.      78  Table 2.11 Macronutrient composition of human milk samples within a women across days    Fat protein  carbohydrate   energy  g/dL  Kcal/dL #1-d1 3.5 0.7 6.7  63 #1-d2 6.0 0.9 7.0  88 #1-d3 6.0 0.7 6  83       #2-d1 4.3 1.0 6.8  73 #2-d2 2.9 1.0 6.8  60 #2-d3 3.7 0.7 5.1  59       #3-d1 3.8 0.9 6.7  67 #3-d2 3.9 0.8 6.4  66 #3-d3 3.0 0.8 6.5  58       #4-d1 3.4 0.9 6.8  64 #4-d2 8.2 0.8 6.0  104 #4-d3 6.9 0.7 5.8  91      # 1, 2, 3, and 4 represent individual women; d: days     79  Table 2.12 Total fatty acids in milk collected from each participant on 3 separate days    Participant #1  Participant #2  Participant #3  Participant #4   d1 d2 d3  d1 d2 d3  d1 d2 d3  d1 d2 d3 16:0 17.1 19.4 16.7  18.6 24.9 16.6  19.5 22.2 20.1  21.4 19.7 18.7 16:1 ω-9 2.15 2.90 0.75  2.21 3.20 2.12  1.93 2.91 2.34  2.26 1.89 1.88 18:0 6.16 6.16 5.22  6.76 7.99 6.12  8.02 9.01 6.84  6.16 6.92 5.66 18:1 ω-9 41.6 39.3 46.5  44.2 33.8 46.1  42.2 37.4 36.5  40.2 43.2 41.6 18:1 ω-7 1.68 1.58 1.62  0.19 2.01 1.67  0.18 1.59 1.55  1.59 1.43 1.40 18:2 ω-6 16.7 15.2 15.4  14.2 11.2 15.8  10.7 9.7 14.1  14.9 14.3 16.6 18:3 ω-6 0.07 0.12 0.07  0.04 0.15 0.16  0.09 0.04 0.12  0.05 0.07 0.08 20:0 0.20 0.18 0.06  0.18 0.20 0.17  0.36 0.30 0.23  0.22 0.23 0.26 18:3 ω-3 1.53 1.46 1.40  1.62 1.24 1.17  0.84 1.21 1.69  1.02 1.26 1.84 20:1 0.51 0.39 0.51  0.06 0.45 0.38  0.83 0.54 0.47  0.46 0.56 0.57 18:4 ω-3 0.02 0.02 0.01  0.05 0.00 0.03  0.02 0.00 0.04  0.00 0.01 0.01 20:2 ω-6 0.21 0.17 0.21  0.20 0.24 0.22  0.21 0.14 0.19  0.25 0.29 0.24 20:3 ω-9 0.03 0.09 0.12  0.02 0.07 0.03  0.04 0.11 0.03  0.08 0.09 0.03 20:3 ω-6 0.23 0.29 0.24  0.32 0.38 0.34  0.32 0.25 0.32  0.23 0.20 0.25 20:4 ω-6 0.43 0.45 0.50  0.43 0.38 0.42  0.48 0.46 0.57  0.43 0.45 0.45 22:0 0.07 0.07 0.07  0.06 0.08 0.06  0.23 0.11 0.15  0.15 0.13 0.19 22:1 0.15 0.14 0.17  0.13 0.16 0.13  0.13 0.10 0.10  0.14 0.09 0.10 20:5 ω-3 0.19 0.22 0.37  0.12 0.08 0.15  0.17 0.15 0.26  0.05 0.06 0.05 24:0 0.06 0.04 0.07  0.04 0.06 0.05  0.19 0.08 0.10  0.12 0.07 0.11 22:4 ω-6 0.05 0.03 0.08  0.05 0.08 0.05  0.09 0.06 0.09  0.07 0.08 0.08 22:5 ω-6 0.04 0.04 0.06  0.03 0.06 0.03  0.02 0.06 0.06  0.05 0.04 0.05 22:5 ω-3 0.16 0.17 0.26  0.21 0.13 0.14  0.22 0.26 0.35  0.13 0.12 0.13 22:6 ω-3 0.34 0.37 0.60  0.43 0.28 0.32  0.44 0.54 0.65  0.17 0.13 0.14 Values are g/100 g fatty acid; # 1,2, 3, and 4 represent individual women; d, days; fatty acids are presented in the table in the time order of their detection by GLC; a summary table with all fatty acids, including C8:0 to C14:0 is available in appendix B.80   2.7 Results for Cow Colostrum and Mature Milk Pls-PE Fatty Acids.  Intrigued by whether a similar pattern of LC-PUFA enrichment in milk PL can be found in milk of other species and to begin to underscore the importance of milk PL as potential source of LC-PUFA, we determined the fatty acid composition of total milk lipid, Pls-PE, diacyl-PE and PC in cow colostrum and mature milk. Similar to mature human milk, C20 and C22 ω-6 and ω-3 fatty acid, particularly ARA and DHA in Pls-PE were also higher than in diacyl-PE and PC in cow colostrum and mature milk (Tables 2.13 and 2.14). The lower ARA in diacyl-PE of cow colostrum and mature milk was accompanied with higher OA and not LA, as found in mature human milk diacyl-PE. Surprisingly, Pls-PE ARA was at least three folds higher in cow colostrum than mature milk (13.4 ± 1.74 and 3.85 ± 1.02%, respectively), which consisted of almost double the percentage of OA in cow colostrum (31.2± 2.15, 17.0 ± 2.44 % in cow mature milk and colostrum, respectively). The enrichment of Pls-PE in DHA and ARA compared to diacyl-PE and PC in mature mammalian milk and the significantly higher ARA and DHA in Pls-PE of cow colostrum than mature milk prompt the question of whether colostrum in humans is also a source of Pls-PE enriched in C20 and 22 ω-6 and ω-3 fatty acids.         81  Table 2.13 Fatty acids in Pls-PE, diacyl- PE, PC and total lipid of cow colostrum    Ethanolamines     plasmalogen-PE diacyl-PE PC total fatty acids 18:2 ω-6 11.6 ± 5.61a 16.4 ± 2.81a 10.6 ± 2.81a 2.71 ± 0.26 18:3 ω-6 0.15 ± 0.08a 0.16 ± 0.09a 0.17 ± 0.14a 0.07 ± 0.02 18:3 ω-3 0.54 ± 0.13a 1.30 ± 0.42b 0.74 ± 0.12c 0.44 ± 0.19      20:2 ω-6 0.15 ± 0.10a 0.10 ± 0.03a 0.17 ± 0.02a 0.04 ± 0.01 20:3 ω-6 3.39 ± 1.40a 2.18 ± 0.47b 1.51 ± 0.43c 0.28 ± 0.05 20:4 ω-6 13.4 ± 1.74a 6.07 ± 0.92b 2.58 ± 0.48c 0.50 ± 0.08 22:4 ω-6 1.25 ± 0.24a 0.41 ± 0.14b 0.37 ± 0.29b 0.08 ± 0.03 22:5 ω-6 0.22 ± 0.14a 0.08 ± 0.06b 0.03 ± 0.02c 0.03 ± 0.02      20:5 ω-3 2.47 ± 1.29a 1.71 ± 0.89a 0.48 ± 0.21b 0.18 ± 0.08 22:5 ω-3 5.40 ± 1.21a 2.35 ± 0.45b 0.91 ± 0.31c 0.38 ± 0.16 22:6 ω-3 0.77 ± 0.31a 0.39 ± 0.17b 0.13 ± 0.05c 0.09 ± 0.05      16:0 14.7 ± 6.82 16.0 ± 3.85 36.7 ± 2.73 41.8 ± 3.45 18:0 7.18 ± 3.45 16.7 ± 3.19 11.4 ± 2.60 7.97 ± 1.34 20:0 0.51 ± 0.22 0.17 ± 0.09 0.09 ± 0.09 0.18 ± 0.04 22:0 0.30 ± 0.14 0.07 ± 0.07 0.11 ± 0.06 0.16 ± 0.05      16:1 ω-7 12.7 ± 5.28 2.63 ± 1.20 2.10 ± 0.55 3.05 ± 0.67 18:1 ω-9 17.0 ± 2.44 30.78 ± 9.05 26.3 ± 2.08 22.1 ± 3.60 18:1 ω-7 5.99 ± 3.06 1.16 ± 1.03 1.22 ± 0.62 0.23 ± 0.05 20:1 ω-9 0.84 ± 0.08 0.51 ± 0.12 0.08 ± 0.08 0.18 ± 0.04 24:1 ω-9 0.14 ± 0.04 0.04 ± 0.02 0.04 ± 0.04 0.05 ± 0.03 20:3 ω-9 0.64 ± 0.36 0.24 ± 0.12 0.12 ± 0.06 0.05 ± 0.03 Values are % (g/100 g fatty acid). Data from n=10 cow colostrum samples PE, phosphatidylethanolamine; Pls-PE, plasmalogen-PE; PC: phosphatidycholine. Means of polyunsaturated fatty acids with different letters are statistically different (P<0.05), by ANOVA followed by Tukey’s test.  Fatty acids in Pls-PE, diacyl-PE and PC found in <0.1% and not presented in the table are: C8:0, C10:0, C12:0, C14:0 (except in PC; 3.63 ± 0.98%), C14:1, C18:4 ω-3, and C24:0.  Total milk lipid for cow colostrum also contains C8:0, 0.68 ± 0.15; C10:0, 1.62 ± 0.38; C12:0, 2.60 ± 0.49; C14:0, 11.9 ± 1.18; C15:0, 0.95 ± 0.24; C19:0, 0.20 ± 0.04; and C14:1, 1.17 ± 0.24 %.   82  Table 2.14 Fatty acids in Pls-PE, diacyl- PE, PC and total lipid of mature cow milk    Ethanolamines     plasmalogen-PE diacyl-PE PC total fatty acids 18:2 ω-6 12.7 ± 2.05a 14.7 ± 1.83a 9.70 ± 1.43a 2.37± 0.39 18:3 ω-6 0.13 ± 0.09a 0.10 ± 0.06a 0.11 ± 0.03a 0.05 ± 0.01 18:3 ω-3 0.42 ± 0.12a 0.97 ± 0.14b 0.57 ±  0.24a,b 0.03 ± 0.01      20:2 ω-6 0.18 ± 0.19a 0.07 ± 0.02a 0.12 ± 0.03a 0.03 ± 0.01 20:3 ω-6 2.79 ± 0.78a 1.33 ± 0.57b 0.94 ± 0.26c 0.12 ± 0.04 20:4 ω-6 3.85 ± 1.02a 1.77 ± 0.35b 0.97 ± 0.26c 0.18 ± 0.04 22:4 ω-6 1.08 ± 0.44 0.34 ± 0.18 0.15 ± 0.05 0.03 ± 0.01 22:5 ω-6 0.02 ± 0.04a 0.02 ± 0.01a 0.01 ± 0.01a 0      20:5 ω-3 0.32 ± 0.16a 0.34 ± 0.06a 0.15 ± 0.06b 0.04 ± 0.01 22:5 ω-3 1.12 ± 0.35a 0.66 ± 0.12b 0.01 ± 0.01c 0.08 ± 0.01 22:6 ω-3 0.61 ± 0.52a 0.06 ± 0.05b 0.05 ± 0.03b 0.01 ± 0.01      16:0 9.26 ± 3.23 12.1 ± 1.80 39.7 ± 2.90 38.5 ± 3.42 18:0 11.8 ± 3.61 11.4 ± 1.38 7.59 ± 1.05 9.67 ± 1.38 20:0 0.36 ± 0.40 0.08 ± 0.06 0.09 ± 0.08 0.14 ± 0.02 22:0 0.39 ± 0.22 0.16 ±  0.06 0.09 ± 0.05 0.07 ± 0.01      16:1 ω-7 15.8 ± 3.64 1.91 ± 0.34 1.44 ± 0.44 2.45 ± 0.29 18:1 ω-9 31.2 ± 2.15 51.2 ± 2.61 29.6 ± 3.08 23.8 ± 3.88 18:1 ω-7 5.57 ± 1.17 0.72 ± 0.47 0.96 ± 0.24 0.77 ± 0.20 20:1 ω-9 0.62 ± 0.30 1.05 ± 0.34 0.09 ± 0.14 0.23 ± 0.08 24:1 ω-9 0.02 ± 0.52 0.02 ± 0.02 0.02 ± 0.02 0.02 ± 0.01 20:3 ω-9 1.03 ± 0.56 0.16 ± 0.06 0.07 ± 0.06 0 Values are % (g/100 g fatty acid). Data from n=10 mature cow milk samples PE, phosphatidylethanolamine; Pls-PE, plasmalogen-PE; PC: phosphatidycholine Means of polyunsaturated fatty acids with different letters are statistically different (P<0.05), by ANOVA, followed by Tukey’s post hoc test. Fatty acids in Pls-PE, diacyl-PE and PC found in <0.1% and not presented in the table are: C8:0, C10:0, C12:0, C14:0, C14:1, C18:4 n-3, and C24:0.  Note that total lipid for mature cow milk also contains C8:0 (0.43 ± 0.29), C10:0 (2.68 ± 0.40), C12:0 (3.98 ± 0.62), C14:0 (12.1 ± 1.40), and C14:1 (1.54 ± 0.43) %.    83  2.8 Discussion  With much emphasis over the last few decades on the role of fatty acids in human brain development, particularly the omega-3 fatty acid DHA and its controversial dietary essentiality, our studies here are prompted by a more specific outlook to understanding dietary lipids beyond fatty acids per se. Our overall aim was to expand consideration of fatty acids from a simplistic view of an average fatty acid composition in the infant diet to include the complex lipids in which fatty acids are provided to tissues, including the brain.  Particularly in human milk, as the only source of nutrition to the exclusively-breastfed infant capable of sustaining adequate growth and development, fatty acids are predominantly associated with TG (20). Therefore fatty acids in milk are often studied in isolation of the form in which they are delivered in to the infant. As more is being learned on the MFGM as a potentially bioactive milk component, consisting uniquely of three layers of PL (3), an understanding of the functional role of these milk PL, including differences in content and function of their respective fatty acids relative to those in TG, becomes necessary. The need for this understanding is clear, given the well-known differences in fatty acid functionality in circulation and tissues depending on the type of lipid molecules in which they are present. For example, TG-bound DHA in chylomicrons or VLDL is used as a storage form of energy in the adipose tissue or as energy source in the skeletal muscle. However, a PL-bound or non-esterified DHA in circulation may be more relevant to the brain, with functional roles extending beyond a structural component of synaptic membranes to include regulation of neurotransmitter metabolism (5, 244, 245).  The first step of this work to expand knowledge on the complex milk lipids in which fatty acids are delivered to the infant was to develop methodology for the separation and recovery of milk PL, with emphasis on Pls-PE. We are interested in milk plasmalogens as potential sources 84  of polyunsaturated fatty acids, following recent confirmation of presence of Pls-PE in human milk and enrichment of total milk PL in DHA, in milk samples low in total DHA (41). Furthermore, DHA is predominantly present in Pls-PE in the brain, with both synaptic DHA and Pls-PE increasing during infant development (5). We succeeded in separating and recovering milk Pls-PE, by modifying the HPLC method of Mawarati et al (238) and using a pre-HPLC step of SPE to separate PL from other milk lipids. Unlike other biological fluids and tissues, milk is unique in its predominant TG composition (~98% of total lipid), compared to for example, 24% triglyceride in human plasma lipid and <1% in erythrocytes (246). Using total milk lipid for HPLC compromises PL detection and hampers PL species recovery, as shown in Figure 2.1 (above). Separating and recovering PL by SPE was successful for milk Pls-PE, diacyl-PE and PC with >90% accuracy and minimal change in fatty acids, but not for PS and PI. As acidic and amphipathic lipids, extracting PS and PI from milk using organic solvents is challenging, and if extracted, PS and PI require an acid wash for elution from the SPE cartridge. The addition of an acid wash step to our SPE method degraded Pls-PE, a major study focus here. We therefore ensured accurate recovery of Pls-PE without an acid wash at the expense of poor recovery of PS and PI with future work needed to accurately recover these PL.  By using two HPLC diol columns, diacyl-PE and Pls-PE were separated, close to baseline resolution.  In theory, using an authentic Pls-PE standard and a standard concentration curve, quantification of Pls-PE in the milk samples would have been possible. However, a major obstacle was the co-elution of TEA with Pls-PE, hindering accurate quantification. TEA in the mobile phases is used as chromatography enhancer, without which peak shapes would be distorted, as shown in in Figure 2.5. Therefore, given the difficulty in accurately extracting PS and PI and in quantifying Pls-PE, this method is considered non-quantitative, but enables 85  addressing our question on differences in fatty acid composition between milk PL, particularly ethanolamine-containing PL.  Figure 2.5 HPLC chromatogram showing Pls-PE and diacyl-PE peaks with no TEA in the mobile phases  Using this method, we then showed that Pls-PE contains a higher proportion of C20 and 22 ω-3 and ω-6 fatty acids than diacyl-PE and PC in mature human and cow milk, as well as cow colostrum. It is well-known that PL in milk, similar to PL in circulation and tissues, contain a higher proportion of ARA and DHA, compared to triglyceride (41).  Our work here is to the first to show that milk Pls-PE is enriched in these LC-PUFAs, more than their diacyl counterparts. It is intriguing that even in the cow with no dietary intake of ARA and DHA and with a well-known low ARA and DHA in milk total lipid, their milk PL were enriched in LC-PUFAs.  Given the low amounts of Pls-PE and other PL relative to TG in milk (40), the quantitative contribution of these PL to total ω-3 and ω-6 fatty acids to the infant, has been considered minimal, with little distinction in ω-3 and ω-6 functionalities between different lipid species in infant development. For example, our results suggest that less PL is needed to provide equivalent amounts of LC-PUFA, including ARA and DHA compared to milk TG, assuming one LC-PUFA per molecule of PL or TG. Theoretically, 50 Pls-PE molecules would be needed to provide the infant with 10 ARA fatty acids (10% ARA in Pls-PE and 2 fatty acids/PL), compared 86  to 688 TG molecules (0.48% ARA in total milk lipid and 3 fatty acids/TG). Despite the much lower proportion of ARA and DHA in total milk lipid than PL, it may be debated that milk TG would remain the major quantitative contributor to these fatty acids. For instance, assuming the infant consumes on average 780 ml/d of milk, with an average of 37 g/L fat (0.34% DHA), then daily intake of DHA from total fat would be 110 mg/d. The daily amount of DHA consumed from milk Pls-PE, however, would be 216 µg, even with DHA as 2.88% of total Pls-PE fatty acids. However, a major assumption here is that PL and TG are equivalent dietary sources of LC-PUFA, with similar metabolic fatty acid handling, which could not be the case. Two key steps in lipid handling should be considered here: lipid digestion and fatty acid uptake by extra-hepatic tissues.  The unique stereo-specific fatty acid distribution of human milk TG, with palmitate (C16:0) predominantly at the sn-2 position of the glycerol backbone and LC-PUFA at the sn-1 and 3 positions supports the presence of a different role of LC-PUFA in TG than PL, in which they are esterified on the sn-2 position. The stereo-specificity of milk TG is maintained post-digestion, with gastric and pancreatic lipases preferentially hydrolyzing the sn-1 and sn-3 fatty acids of TG, followed by TG synthesis in the enterocyte mainly via the 2-monoacylglycerol pathway (73). Extra-hepatic tissue uptake of TG fatty acids, mainly in chylomicrons and VLDL, is mediated by lipoprotein lipases with preferential activity towards sn-1 fatty acids and C16 and C18 from the sn-3 position (247). Consequently, DHA and other LC-PUFA in TG may be cleared differently, with possible re-distributing to other lipids (i.e, Pls-PE, diacyl-PE, lyso- PL, unesterified fatty acids) during de novo hepatic synthesis.  Dietary diacyl-PL digestion is more understood than that of plasmalogen and is different from triglyceride digestion. The major enzyme for PL digestion is pancreatic phospholipase A2, 87  which cleaves the sn-2 fatty acid. 2-lyso-PL in the enterocyte can then be re-acylated to form a PL or hydrolyzed to release the sn-1 fatty acid and a glycerol-phosphate-moiety. Data on the digestion of milk Pls-PE in humans is scarce. In rats, however, dietary plasmalogens undergo minor degradation in the stomach and intestinal lumen, are hydrolyzed to 2-lysoplasmalogen and a fatty acid, then preferentially re-esterified with ARA and DHA at the sn-2 position and absorbed via the lymph (117, 120). Whether humans possess a digestive enzyme capable of cleaving the sn-1 fatty acid of Pls-PE, while preserving the sn-2 fatty acid is not known. This would be relevant to the brain, with evidence of preferential uptake of DHA bound at the sn-2 position of 1-lyso-PL than other sources, including non-esterified DHA (245, 248). Stable isotope tracer studies in neonatal piglets and baboons show higher efficacy of dietary PL than TG-bound ARA and DHA for ARA and DHA brain accretion (249, 250). Regardless of the complex issue of the source of DHA up taken into the brain as recently reviewed (245), it is clear that milk PL and TG fatty acids are not equivalent in metabolic fate and therefore their dietary forms need to be addressed separately, beyond their quantitative contribution to infant total intake.   The mechanisms by which dietary Pls-PE may affect brain development may extend beyond the role of their fatty acids. The oxygen atom attached to glycerol by the characteristic ether bond at the sn-1 position of plasmalogens acts as a sacrificial oxidant, speculated to protect the sn-2 fatty acid from oxidation (47). Whether the ARA and DHA in plasmalogen are more resistant to oxidation in the intestine and have higher bioavailability is not known. It is clear that this research raises fundamental questions regarding the structure-function characteristics of lipids in infant development, with future studies needed to better understand whether the dietary 88  fatty acid form alone, the PL  type alone, or an interaction of both impart distinct properties to human milk, capable of sustaining infant development, including the brain. In conclusion, our studies enabled the development of an improved method for the separation and recovery of milk Pls-PE and diacyl-PE, showing LC-PUFA including DHA in are enriched in milk PL, particularly in Pls-PE of human and cow milk. Future studies need to address the biological roles of milk PL and their fatty acids.                   89  Chapter 3. The Milk Fat Globule Membrane and the Brain 3.1 Introduction Milk has evolved over millions of years to provide the offspring with nutrients and non-nutritive components necessary for thriving in the extra-uterine environment. Nutrients in milk for infants have often been researched in isolation from one another, despite the potential interaction between them in impacting infant growth and development.  Human milk lipids are complex in their structure and diverse in their composition, with TG emulsified in the milk aqueous phase due to their envelopment in a unique tri-layered milk fat globule membrane (MFGM). Unlike any other biological membrane, the MFGM is comprised of three layers of polar lipids (i.e, glycolipids, PL), cholesterol, glycoproteins and proteins (29).With lipids providing around 50% of the calories to the breastfeeding infant and triglycerides contributing to the majority of milk lipids (~98%), emphasis has been on the role milk lipids as energy source and source of essential ω-6 and ω-3 fatty acids. Little attention has been given to the potential biological role of the MFGM and its lipids, including the ethanolamine-containing PL, diacyl-phosphatidylethanolamine (diacyl-PE) and plasmalogen-PE (Pls-PE). These PL in milk are enriched in LC-PUFA, including arachidonic acid (ARA; 20:4 ω-6) and docoxahexaenoic acid (DHA; 22:6 ω-3) which are implicated in central nervous system and immune system function (6, 225, 228). Notably, infant formulas and enteral/parenteral nutrition products are based on vegetable oil preparations with no MFGM. Interest in the MFGM as a functional dietary component has risen, with the recent commercial availability of bovine MFGM. Of relevance to our work on the brain, infants fed formula supplemented with bovine MFGM from 2 to 6 months scored 4.0 (95% CI: 1.1, 7.0) points higher on cognitive testing at 12 months using the Bayley Scales of Infant and Toddler Development than infants fed standard 90  formula, with the former scores being similar to those of a reference group of breastfed infants (178). In artificially-reared neonatal piglets, supplementation with  MFGM-phospholipids (PL) resulted in better performance in tests of spatial learning in infancy, with an increase in cortical and cerebellar gray and white matter and differences in brain metabolites between supplemented and control animals (173).  Brain development is a complex process beginning in utero and continuing during childhood. Individual components of the MFGM, including PL, DHA, and sialic acid have been implicated in brain development, but the extent to which their dietary forms within the MFGM or the MFGM itself contribute to brain development is not known. Accessing the human brain for analytical samples raises ethical concerns, which requires using alternative research approaches to better understand whether or not the MFGM affects the human infant brain. The rat has been used as a model to understand the role of milk fatty acids particularly the essential fatty acids and DHA, in brain development, usually by manipulating the fat and fatty acid content of the maternal diet to alter those in milk (92, 251-254). However, this approach is not possible for addressing the role of the MFGM, with also no knowledge on differences in MFGM composition and biological determinants between the human and the rat. Gastrostomy rearing (artificial-rearing) is a well-established model in nutritional studies, allowing complete control of dietary intake during the early period of postnatal growth and development, independent of the effect of maternal physiology on milk composition (255).  This research was designed with the objective of determining whether feeding MFGM to artificially-reared infant rats affects brain lipid and metabolite composition during early postnatal development.  91  3.2 Methods  All animal procedures were conducted as approved by the Animal Care Committee of the University of British Columbia and conformed to the guidelines of the Canadian Council on Animal Care. Pregnant Sprague Dawley rats (gestational age 10-13 days; n=3-4 per experiment) (Charles River Laboratories) were housed individually in a temperature- and humidity-controlled animal facility with a 12-hour light: dark cycle and adlibidum access to water and food (commercial rodent chow; Harlan-Teklad; www.harlan.com). The offspring was born at term (21-22 days gestation), with a litter size of 7-10 pups, and remained with the rat dams for 5 days.  Beginning at day 5 postnatal, infant rats were gastrostomy reared, using a technique commonly referred to as the pup-in-the-cup technique: Rats are fed by gastric tubes with each rat kept in a floating container in a temperature-controlled water bath. A reference group of rat pups (mother-reared, MR) were also included and were kept with their dams. Of note, comparisons of changes in brain lipids of our experimental groups with MR rats are only exploratory, acknowledging the complex composition of milk beyond presence and absence of the MFGM (eg, growth factors, hormones) which cannot be currently duplicated in formula (256, 257). - 3.2.1 Cannulation At day 5 postnatal, the infant rats were cannulated, with optimization of methods from previous work (258) and with experience here. The rats were anaesthetized using halothane and not isofluorane, to ensure they do not regain consciousness during the procedure. A 7cm, PE-50 silica tubing, with a wire inside it, was used to perform the cannulation. The wire, along with the tubing, was inserted into the mouth, moved through the esophagus and pulled out of the stomach, abdominal wall, and skin. To avoid scratching of the esophagus with wire, the wire was extended out of one end of the PE-50 tubing by about 2 cm, and the tubing was lubricated with a mucous-92  like gel (MUKO; Source; Missisauga ON). Once the stomach was pierced, the 7 cm PE-50 tubing was replaced with a 25 cm PE-10 tubing, with a plastic circle at its end, as an anchor for the PE-10 tubing to the stomach (Appendix A.6).  The PE-10 tubing was gently pulled out of the stomach, until the plastic circle touches the stomach wall. The cannula and wire were then passed through a 1 cm fold of skin on the back of the neck, to increase the stability of the cannula during rat movement and handling.  The entire procedure took around 5 min/animal. After wire removal, the cannulas were connected to gastric feeding tubes, connected to micro-peristaltic pumps (Ismatec-IPC; Wertheim, Germany), via cassettes. 3.2.2 Animal Maintenance The rat pups were maintained in 500 ml plastic cups, containing corn cob and paper fiber (Biofresh, Absorption Cor) as contaminant-free bedding, and floated on a water bath incubator, maintained at 40-42oC. Diligent care was taken during rat pup maintenance activities (daily from 7:00 am to 10:00 pm) to ensure minimal risk of infection and illness, and to avoid unbiased care (i.e, more physical contact with one rat versus the other, when not justified), as these may affect experimental outcomes. Typical physiological maintenance included stimulating the rats to urinate and defecate, by gently rubbing the ano-genital region with a moistened cotton swab, every 3-5 hrs. Body weight was measured every time the rats were stimulated, and individual adjustments to milk volume were made for rats occasionally losing weight. Hypothermia was avoided by monitoring and adjusting the water bath incubator temperature to 40oC twice a day, gently covering the rats with bedding when in the plastic cups, and placing them on a heating pad when outside the plastic cups. Subcutaneous injections of 100-200 µl of lukewarm sterile saline solution (Pedialyte, Abbot) were made, in rare cases of rat dehydration. A topical antibiotic cream (Polysporin, Johnson & Johnson) was applied on the stomach and neck area if signs of local inflammation (i.e redness and swelling) were observed. In extreme conditions of hypoxia 93  (evident by rat gasping), significant hypothermia (evident by a change of skin color to pale blue), and detachment of the cannula, rats were monitored meticulously and euthanized under anesthesia if necessary.   3.2.3 Feeding Schedule The rat pups were fed for the first 10 min of a 30-min cycle (i.e 10 min feeding and 20 min rest), repeatedly for 24 hrs a day. The total milk volume and milk flow rate were adjusted daily, in the morning and evening, at 25% of average body weight. The gastric tubes were flushed with warm water and refilled with milk three times a day (i.e morning, early afternoon and evening) to ensure patency and a consistent flow of milk across tubes. The accuracy of the pumps in dispensing a pre-set volume of milk was measured daily, and the pumps were re-calibrated if an error in volume, greater than 5%, was found.  3.2.4 Animal Diets The infant rats were randomly assigned to either a milk formula with MFGM or without MFGM on day 5 and maintained on the same diet until sacrificed, on days 10, 13, 15 or 18 postnatal (n=4-6/group). A reference group of rat pups were kept with their dams and sacrificed at days 5 (as baseline on cannulation day), 10, 15 and 18.  The two milk formulas were prepared as nutritionally adequate rat milk substitutes, following previous methods with optimization (258) and provided similar amounts of carbohydrate, protein, fat, vitamins and minerals/L (Table 3.1). Formula supplemented with MFGM (6 g/L; Lacprodan® MFGM-10, Arla Foods Ingredients, Denmark) provided PE, PI, PS, PC and Sph as 2391 ± 139, 730 ± 46.0, 218 ± 14.0, 1742 ± 98.0, 1227 ± 51.0 µg/100mg respectively (based on 4 replicate analyses). Of note, Lacprodan® MFGM-10 also contains gangliosides, sialic acid, 94  immunoglobulins, and lactoferrin, as claimed by manufacturers. The complete fatty acid composition of the milk formulas and MFGM PL is summarized in Appendix C. Table 3.1 Nutrient and oil composition of experimental milk formulas     Formula with MFGM Formula without MFGM Carbohydrate, g/L Protein, g/L Fat, g/L      Coconut oil      Corn oil      Soybean oil      Captex MCT2 MFGM3, g/L Vitamins4, g/L Minerals5, g/L 30 1171 140 20 41 26 53 - 4 9 30 1171 140 20 41 26 53 6 4 9 Riboflavin6, mg/L Niacin6, mg/L Pyridoxal6, mg/L Inositol6, mg/L Creatine6, mg/L Ethanolamine6, mg/L 16 26 14 914 70 34 16 26 14 914 70 34 The milk formulas provided 180 kcal/100 ml.  1 Total Protein consisted of casein and whey (70, 47 g/L, respectively).  2 MCT: medium chain triglycerides (Captex®, Aic, USA). 3 MFGM: milk fat globule membrane (Lacprodan® MFGM-10, Arla Foods Ingredients, Denmark). 4Harlan-Teklad #40060; 5Harlan-Teklad AIN93G. 6 added as a supplemental vitamin mix  3.2.5 Tissue Collection At 5, 10, 13, 15 or 18 d of age, rat pups were anesthetized using isoflurane, and blood was collected by cardiac puncture, then the rats killed by cutting through the diaphragm. The 95  cerebrum was rapidly removed, weighed, flash frozen in liquid nitrogen and stored at -80°C until analyzed.  3.2.6 Analytical Methods Total lipids were extracted from homogenized frontal lobe (Appendix A.1), then lipid classes (free cholesterol, PE, PI, PS, PC and Sph) were separated using a high performance liquid chromatograph (HPLC) with a quaternary linear gradient solvent system, and quantified with an evaporative light-scattering detector (73). PE and PC were recovered using a fraction collector, then their fatty acids were separated and quantified as their respective methyl esters and dimethyl acetals (Appendix A.2), by capillary column gas liquid chromatography (GLC) (259).  The frontal lobe protein content was determined using a modified Bradford method (Appendix A.4).  Targeted metabolomics on the frontal lobe samples was performed by Gas Chromatography-Mass Spectrometry (GC-MS) using a Quattro micro TM-GC mass spectrometer (Agilent 6890N Network GC system; Agilent Technologies. CA, USA). A 30 m * 0.25 mm *0.25 um film thickness column (Agilent HP5MS), with helium as carrier gas was used, and samples were analyzed under the following set conditions: injected sample volume of 1 µl; injector mode as splitless; flow rate of 1.5 ml/min; injector temperature of  250 ⁰C, temperature gradient held at 80 ⁰C for 2 min, then increased to 312⁰C at a rate of 8C/min and held at 312 C for 17 min; total analysis time of 48 min. Mass spectra were acquired at m/z 75-650 at a rate of 120 scans/minute.  3.2.7 Statistical Analysis  Results from lipidomic analyses are expressed as means ± standard deviation (SD). Differences betweenthe experimental groups were determined using independent student t-test. 96  Differences in body and brain weight across the three groups were determined by one-way analysis of variance (ANOVA).Statistical analyses were performed using SPSS software (version 20 for Windows), with the level of statistical significance set at P <0.05.  MarkerLynx software was used to perform Principal Component Analysis (PCA) for differences in the overall brain metabolite patterns, for metabolites associated with mass spectra of m/z between 75 and 650, between the groups and to identify the metabolite peaks with significantly different area counts between the groups. The spectra for each peak was obtained using MassLynx software (version 4.1), then used to identify the nature of the metabolites using the online NIST MS library (version 2.0). The nature of the identified metabolites was confirmed by comparing their retention times and mass spectra to those of their respective authentic standards, derivatized using the same protocol as described in Appendix A.5.     3.3 Results  Rats fed diets with or without MFGM had similar body and brain weights at different time points, with no significant differences with MR rats (Table 3.2).         97  Table 3.2 Total body and brain weights of mother-reared rats (MR), rats fed formula with MFGM (MFGM-fed) and rats fed formula without MFGM (No-MFGM)   MR  MFGM-fed No-MFGM    grams  d5            Body  Brain1 17.0 ± 1.15 0.54 ± 0.09  14.8 ± 1.82 NA 14.4 ± 1.87 NA d10 Body Brain 24.3 ± 3.52 0.80 ± 0.10  21.1 ± 0.31 0.68 ± 0.05 20.9 ± 1.15 0.88 ± 0.10 d13 Body Brain NA   34.8 ± 5.02 0.92 ± 0.03 30.3 ± 6.54 0.93 ± 0.05 d15 Body Brain 45.4 ± 5.79 1.25 ± 0.18  36.4 ± 7.88 0.94 ± 0.08 39.1 ± 4.28 1.03 ± 0.05 d18 Body Brain 38.9 ± 6.24 0.92 ± 0.03  37.9 ± 1.19 1.05 ± 0.07 36.1 ± 2.68 1.00 ± 0.05 means ± SD; 1 values correspond to cerebrum wet weight n=4 per group; d, day of organ collection NA, not applicable No significant difference between groups using ANOVA  Given that the literature on rat brain PL during the early postnatal period is from the early 1960’s with different methodologies used nowadays (128), we quantified brain PL for a reference group of mother-reared (MR) rats at 5, 10, 15 and 18 days postnatal (Table 3.3). Not expecting significant differences in brain PL at d13, a reference group of pups at d13 was not included, unlike in the experimental groups. Brain lipids in MR rats, as well as experimental rats are expressed by brain weight (µg/50 mg) and not by unit of protein content (µg/mg protein) because of the difficulty in interpreting findings, given the variability in protein content, found at different time points (Table 3.4). For example, an increase in brain protein content from 1.7 to 2.7 mg/50 mg between d10 and d13 in MFGM-fed rats would decrease 98  the amount of lipid relative to protein between the two time points, complicating interpretation of temporal changes in the brain. Table 3.3 Developmental changes in brain lipids of mother-reared rats (MR)    d5 d10 d15 d18  µg/50 mg brain wet weight Phospholipids     PE  280 ± 56.9 282 ± 33.7 473 ± 31.5 521 ± 69.4 PI  74.5 ± 15.6 56.1 ± 7.28 88.9 ± 1.17 106 ± 20.3 PS 50.1 ± 8.17 36.3 ± 8.79 55.3 ± 0.57 59.1 ± 11.8 PC  454 ± 99.1 431 ± 50.9 631 ± 59.2 657 ± 73.5 Total 859 ± 176 805 ± 97.1 1249 ± 89.8  1343 ± 164      Sph 30.3 ± 12.6 42.1 ± 11.5 60.8 ± 5.29 44.7 ± 8.22 F-Cholesterol 165 ± 26.5 225 ± 30.8 441 ± 47.8 501 ± 113 means ± SD; n=4 per group d, days postnatal  F, free; PE, phosphatidylethanolamine; PI, phosphatidylinositol PS, phosphatidylserine; PC, phosphatidylcholine; Sph, sphingomyelin Table 3.4. Protein content in brains of mother-reared rats (MR), rats fed formula with MFGM (MFGM-fed) and rats fed formula without MFGM (No-MFGM)    MR  MFGM-fed No-MFGM    mg/50 mg wet weight d5 2.5 ± 0.70  NA NA d10 2.3 ± 0.88  1.7± 0.38 1.7 ± 0.61 d13 -  2.7 ± 0.81 2.8 ± 0.68 d15 2.3 ± 0.88  2.3 ± 0.92 2.4 ± 0.44 d18 2.9 ± 0.28  2.5 ± 0.54 2.7 ± 0.35 means ± SD; n=4 per group; d: day of organ collection;  NA, not applicable  Changes in Brain PL Total brain PL increased from 859 at d5 to 1343 µg/50 mg  at d18, with the  most evident increase found in PE, PC and PI (86, 45, and 42 % increase, respectively) (Table 3.3). PE was the main PL accumulating in the brain, with almost a doubling in its amount from d5 to d18. The proportion of plasmalogens (Pls-PE) from total brain PE, and therefore brain Pls-99  PE content, also increased from an estimated 13% at d5, to 21% at d10 and 26% at d15 and d18. Consequently, the PC-to-PE ratio decreased from 1.62 ± 0.02 at d5 to 1.30 ± 0.05 at d18. PC and PE became statistically difference from d5, at d15 (P=0.03 and P=0.02, respectively. Although no significant difference in PS and PI were found across different days, this is likely due to the well-known inaccuracy of extraction methods in recoveries these PL. The accumulation of brain free cholesterol is consistent with the older literature, likely due to cholesterol de novo synthesis in the brain (128). While present in all cell membranes, the amount of Sph in the brain is closely associated with myelination, which in the rat has been reported to have the highest rate between d12 and d18 postnatal (128). Here, we found a statistically insignificant increase in Sph from d5 to d15 with no further increase at d18.  Temporal changes in brain lipids were also found in rats fed formula with or without MFGM, with similar amounts of total brain PL at all time points (Table 3.5). The no-MFGM group tended to be significantly higher in total brain PL than MFGM-fed rats at d18 (1479 ± 131, 1262 ± 144 µg/50 mg, respectively, P=0.05). This was explained by significantly higher brain PE (P=0.03) and PS (P=0.01) in the no-MFGM than in the MFGM-fed group, with PI also tending to be higher in the no-MFGM group (P=0.05). Interestingly, although total PE was similar in both groups at d10, Pls-PE accounted for 23% of total PE lipids in the MFGM-fed group compared to only 11% in the no-MFGM groups (P=0.02). The proportion of Pls-PE from total PE was similar between the groups at d13, d15 and d18, ranging from ~18-22%.  There was no significant difference between the two groups in brain cholesterol at any time point, except for higher cholesterol in the no-MFGM group at d18, approaching statistical significance (499 ± 60.2, 408 ± 36.5 µg/50 mg, respectively, P=0.05). MFGM-fed rats had 100  approximately double the amount of Sph at d10 than no-MFGM rats (34.3 ± 10.1, 16.3 ± 1.67, P=0.01), but this difference was not maintained at d13 onwards.  Interestingly, the brain lipid composition of MFGM-fed rats was closer to that of MR rats than the no-MFGM group. No differences in brain lipids were found between the MFGM-fed and MR rats, except for higher PE at d10 in the MFGM-fed group (P=0.04). However, rats in the no-MFGM group had lower Sph at d10 (P=0.03) and at d15 (P=0.02), and higher PS at d18 (P<0.01) than MR rats. PS at d15 and PE at d18 also tended to be higher in the no-MFGM than MR group (P=0.07 for both). At d18, brain PC-to-PE ratio in MFGM-fed rats was closer to that of MR rats (1.17 ± 0.04, 1.30 ± 0.03, respectively, P<0.01) than the no-MFGM group (1.06 ± 0.01, P<0.01). Therefore, feeding MFGM to infant rats during a period of brain development resulted in changes in brain lipids, which were closer in composition to those of MR rats than when MFGM was not fed.             101  Table 3.5 Changes in brain lipids in rats fed formula with or without MFGM     d51 d10 d13 d15 d18  µg/50 mg wet weight Phospholipids      PE           With MFGM 280 ± 56.9  366 ± 62.8 397 ± 48.3 493 ± 107 508 ± 49.1† Without MFGM 372 ± 135 452 ± 54.6 478 ± 131 612 ± 58.4 PI            With MFGM 74.5 ± 15.6  71.4 ± 15.1 81.6 ± 6.38 91.0 ± 34.4 92.0 ± 19.7 Without MFGM 70.6 ± 23 84.0 ± 6.46 89.8 ± 27.9 120 ± 15.9 PS           With MFGM 50.1 ± 8.17  48.3 ± 7.09 54.3 ± 3.11 62.2 ± 32.5 67.8 ± 15.6† Without MFGM 57.7 ± 18.4 54.1 ± 11.8 71.8 ± 19.6 98.2 ± 11.3 PC            With MFGM 454 ± 99.1  510 ± 71.2 580 ± 59.7 621 ± 92.9 594 ± 65.7 Without MFGM 515 ± 212 526 ± 56.5 578 ± 120 649 ± 11.5 Total           With MFGM 859 ± 176  996 ± 149 1113 ± 113 1267 ± 137.4 1262 ± 144 Without MFGM 1015 ± 121 1116 ± 72.0 1218 ± 290 1479 ± 131       Sph           With MFGM 30.3 ± 12.6  34.3 ± 10.1† 29.4 ± 5.39 42.3 ± 9.96 35.8 ± 11.0 Without MFGM 16.3 ± 1.67 32.1 ± 12.3 31.6 ± 16.2 42.1 ± 11.5 F-Cholesterol            With MFGM 165 ± 26.5  275 ± 45.7 366 ± 49.1 414 ± 81.9 408 ± 36.5 Without MFGM 210 ± 31.1 408 ± 94.9 389 ± 119 499 ± 60.2 means ± SD; n=4 per group per time point; d, days postnatal 1 reference values from MR rats at d5.  †significantly different between the two groups, by independent student t-test (P<0.05). F, free; PE, phosphatidylethanolamine; PI, phosphatidylinositol; PS, phosphatidylserine; PC, phosphatidylcholine; Sph, sphingomyelin  Brain DHA in Response to Feeding MFGM   It is known that DHA increases in the rat postnatal brain during development similar to the human, and is primarily present in gray matter PE and PS (i.e, synaptic membranes) (122, 128, 130, 132).  Consistently, DHA in PE constituted ~15-20% of fatty acids in MR rat brain, compared to ~2 % in PC at all time points (Table 3.6). PE DHA increased from 16% at 102  d5 to 20 % at d10 (P<0.01) and remained insignificantly different at d15 and d18 in MR rats. With PE increasing in rat brain during development and DHA predominantly being in PE, an increase in net brain DHA (µg/50 mg brain) secondary to the increase in PE would be expected and was indeed found (Table 3.7). As such, net PE DHA increased from 21.3 ± 3.45 at d5 to 48.3 ± 8.79 µg/50 mg at d18 in MR rats. Therefore, as suggested during the 1960’s-1970’s (128, 130), our data show the increase in brain PE during development in the mother-reared rat contributes to a net increase in brain DHA.  Table 3.6 Polyunsaturated fatty acid composition of brain PE and PC in mother-reared rats    d5 d10 d15 d18  g/100 g fatty acid in PE 18:2 ω-6 0.61 ± 0.32 1.12 ± 0.34 1.83 ± 0.40 1.67 ± 0.01 20:4 ω-6 19.8 ± 1.22 21.3 ± 1.35 19.6 ± 1.12 19.3 ± 1.48 22:4 ω-6 3.76 ± 1.91 4.78 ± 0.38 5.52 ± 0.37  6.10 ± 0.34 22:5 ω-6 2.56 ± 0.28 2.10 ± 0.12 1.50 ± 0.31 1.35 ± 0.34 20:5 ω-3 0.07 ± 0.06 0.06 ± 0.03 0.12 ± 0.07 0.07 ± 0.01 22:5 ω-3 0.36 ± 0.04 0.57 ± 0.05 0.53 ± 0.12 0.43 ± 0.03 22:6 ω-3 16.0 ± 1.83 20.1 ± 0.93 18.6 ± 2.18 19.6 ± 0.65  g/100 g fatty acid in PC 18:2 ω-6 1.22 ± 0.16 1.28 ± 0.12 1.81 ± 0.32 1.23 ± 0.21 20:4 ω-6 3.95 ± 0.89 4.22 ± 0.82 5.06 ± 0.68 6.69 ± 1.31 22:4 ω-6 0.46 ± 0.17 0.29 ± 0.16 1.28 ± 0.34 0.82 ± 0.17 22:5 ω-6 0.46 ± 0.17 0.26 ± 0.16 1.28 ± 0.32 0.82 ± 0.17 20:5 ω-3 0.08 ± 0.08 0.08 ± 0.04 0.09 ± 0.04 0.09 ± 0.03 22:5 ω-3 0.06 ± 0.01 0.08 ± 0.04 0.09 ± 0.04 0.09 ± 0.03 22:6 n-3 1.52 ± 0.18 1.54 ± 0.43 1.70 ± 0.16 2.20 ± 0.44   means ± SD; n=4 per group; d, days postnatal. 103  Table 3.7 Net DHA content in brain PE in mother-reared rats (MR), rats fed formula with MFGM (MFGM-fed) and rats fed formula without MFGM (No-MFGM)   d5 d10 d13 d15 d18  µg in PE/50 mg brain wet weight MR 21.3 ± 3.45 34.6 ± 1.37 NA 30.8 ± 6.24 48.3 ± 8.79 MFGM-Fed NA 18.1 ± 7.25 22.8 ± 1.58 23.2 ± 2.5 39.4 ± 5.69 No-MFGM NA 22.9 ± 12.7 25.4 ± 6.31 31.6 ± 12.5 33.9 ± 11.2 Means ± SD. NA, not applicable. No significant between experimental artificially-reared groups using independent student t-test.  Similar to MR rats, brain PE and not PC was enriched in DHA in the artificially-reared rats. We therefore focused our analysis on differences in PE DHA between the groups, with future work needed to explore these differences for DHA in PS. A similar distribution of polyunsaturated fatty acids in brain PE and PC between rats fed formula with or without MFGM was found (Table 3.9). DHA as a percent of total PE fatty acids, was remarkably similar between the groups at d10, d15 and d18, but was higher in the no-MFGM than MFGM-fed group at d13 (15.7 ± 0.73, 17.4 ± 0.36, respectively, P<0.01). Nonetheless, there was no significant difference in the net DHA content of brain PE at all time points between the two groups (Table 3.7). Interestingly here, despite having higher PE than the MFGM-fed group at d18, the no-MFGM group were not found to have higher net PE DHA than the MFGM-fed group (33.9 ± 11.2, 39.4 ± 5.69, respectively). Therefore, our data suggest no difference in net PE DHA accumulation in the artificially-reared rat due to feeding MFGM during early development.  Net DHA accumulation and changes in PL composition, including PE accumulation, have been traditionally used as markers of development in the rat and human (11, 122, 123, 128, 130, 135, 146). However, defining adequate brain function and development based on measures of brain lipids is complicated. Work by Rupoport and colleagues (260, 261) stimulates interest into examining brain function through more functional measures, such as brain DHA turnover 104  rather than net DHA amounts. It is intriguing that when DHA supply to the brain is limiting, DHA turnover in the brain decreases, such that net amounts of DHA are preserved (260, 261), potentially at the expense of altered DHA-related functions (eg. signal transduction for neurotransmitter release) (262-264).  We therefore extended our lipidomic study to include metabolomics, to better understand whether rats fed formula with or without MFGM differ in brain metabolites. No significant differences in brain metabolites, for which the mass spectra were acquired at m/z between 75-650, were found at d10 and d15 between the two experimental groups. At d13, MFGM-fed rats were significantly lower in 7 brain metabolites: 5 amino acids, glycerol-3-phosphate (G-3-P), and inositol (Table 3.8). Glutamate, G-3-P, and inositol were also significantly lower in the MFGM-fed group at d18, in addition to lactate and serine. Alanine tended to be significantly lower in the MFGM-fed rats at d13 and d18 (fold difference; P: 0.55; 0.08 and 0.49; 0.07, respectively), as well as glycine at d18 (0.67; 0.05). Of note, no consistent pattern of metabolite changes were found with age within a group.         105  Table 3.8 Differences in brain metabolites between rats fed formula with or without MFGM Brain metabolite Fold-difference1 P d13      aspartate      glutamate      glutamine      glycerol-3-phosphate      glycine      inositol      threonine   d18      glutamate      glycerol-3-phosphate      inositol      lactate      serine         0.63 0.62 0.61 0.29 0.52 0.35 0.34   0.63 0.36 0.62 0.69 0.44   0.004 0.007 0.01 0.008 <0.001 0.005 0.03   0.03 0.01 0.01 0.04 0.005                   d, day postnatal; 1 difference is for metabolites in MFGM-fed rats                              relative to no-MFGM rats.                              Differences between groups determined using independent sample t-test       106  Table 3.9 Polyunsaturated fatty acid composition of brain PE and PC of rats fed formula with or without MFGM  mean ± SD; n=4 per group; d, days postnatal; ND, not detected; PE, phosphatidylethanolamine; PC, phosphatidylcholine     d10  d13  d15  d18  MFGM No-MFGM  MFGM No-MFGM  MFGM No-MFGM  MFGM No-MFGM   g/100 g fatty acid in PE          18:2 ω-6 0.48 ± 0.34 1.22 ± 1.09  0.91 ± 0.12 1.68 ± 0.61  1.37 ± 0.56 1.11 ± 0.13  1.80 ± 0.33 1.38 ± 0.14 20:4 ω-6 22.1 ± 2.37 23.3 ± 1.44  21.6 ± 0.37 18.8 ± 1.02  21.8 ± 1.21 21.8 ± 1.53  20.4 ± 1.35 20.4 ± 1.10 22:4 ω-6 5.06 ± 0.43 5.92 ± 0.32  5.97 ± 0.40 6.02 ± 0.37  6.21 ± 0.59 6.48 ± 0.72  6.07 ± 0.37 6.43 ± 0.27 22:5 ω-6 2.98 ± 0.18 3.66 ± 0.44  3.12 ± 0.28 2.88 ± 0.40  3.42 ± 0.50 3.99 ± 0.78  2.68 ± 0.33 1.89 ± 1.51             20:5 ω-3 0.10 ± 0.04 0.05 ± 0.04  0.13 ± 0.25 0.08 ± 0.03  0.07 ± 0.08 0.08 ± 0.01  0.08 ± 0.02 0.05 ± 0.03 22:5 ω-3 0.35 ± 0.04 0.34 ± 0.06  0.34 ± 0.04 0.32 ± 0.01  0.40 ± 0.05 0.36 ± 0.05  0.32 ± 0.01 0.32 ± 0.03 22:6 ω-3 16.4 ± 1.14 16.5 ± 2.35  15.7 ± 0.73 17.4 ± 0.36  16.2 ± 2.11 15.1 ± 1.55  17.0 ± 0.96 16.1 ± 1.32 g/100 g fatty acid in PC          18:2 ω-6 1.61± 0.14 1.66 ± 0.15  1.96 ± 0.20 2.00 ± 0.24  1.68 ± 0.11 1.89 ± 0.25  1.69 ± 0.22 1.62 ± 0.16 20:4 ω-6 3.79 ± 0.58 3.82 ± 0.59  5.66 ± 0.18 5.08 ± 0.45  4.95 ± 0.27 6.38 ± 1.78  5.66 ± 0.87 6.30 ± 0.88 22:4 ω-6 0.35 ± 0.05 0.44 ± 0.13  0.60 ± 0.10 0.53 ± 0.08  0.48 ± 0.38 0.70 ± 0.16  0.59 ± 0.12 0.69 ± 0.11 22:5 ω-6 0.19 ± 0.09 0.39 ± 0.06  0.29 ± 0.03 0.31 ± 0.08  0.31 ± 0.19 0.38 ± 0.19  0.31 ± 0.04 0.37 ± 0.10             20:5 ω-3 ND 0.01 ± 0.01  0.04 ± 0.02 0.03 ± 0.02  ND ND  ND ND 22:5 ω-3 0.07 ± 0.03 0.29 ± 0.05  0.05 ± 0.01 0.05 ± 0.01  0.08 ±0.02 0.04 ± 0.04  0.05 ± 0.01 0.06 ± 0.04 22:6 ω-3 1.05 ± 0.20 0.76 ± 0.42  1.35 ± 0.11 1.54 ± 0.29  1.33 ± 0.22 1.16 ± 0.82  1.76 ± 0.69 1.87 ± 0.45 107  3.4 Discussion   In the present study, we sought to understand whether the MFGM has effects on the brain during development. Our results show that the brains of infant rats fed formula with MFGM differ in lipid and metabolite composition than brains of rats not fed MFGM, for which the functional relevance needs to be addressed.  The lipid research paradigm has shifted tremendously over the last decade, with much more interest now on the structural-functional properties of human milk lipids instead of absolute amounts of individual dietary fatty acids. Knowledge gained during the last 50 years have contributed to ensure infants on enteral feeds and parental nutrition are free of deficiency (i.e, nutritionally sufficient). Relevant here, enteral and parental products, including infant formula contain essential fatty acids, with often added DHA similar in composition to the human milk triglyceride (TG). TG in infant formula, however, is often emulsified by polysaccharides, such as carrageenan, and lecithins (265). This is different than in mammalian milk, with TG emulsified in the MFGM. Why the mammary gland invests in energy and substrates (e.g, phosphorus, ethanolamine, choline, proteins) to form the MFGM, as a three-layered envelope of PL and proteins around the TG is intriguing and incompletely understood.  The mammalian MFGM, mostly understood in the human and the cow, is a complex biological membrane consisting of PL (diacyls and Pls), sphingolipids, (i.e Sph), glygosphingolipids (i.e, cerebrosides and gangliosides), cholesterol and proteins (32, 69). Its complexity is not only related to its diverse components, but also to the specificity of their arrangement within the membrane. For example, the MFGM includes lipid rafts, as aggregates of Sph, cholesterol and Pls on its outer membrane (38). PC is mainly present in the outer leaflet and PE, PS and PI in the inner leaflet of the MFGM bilayer, with limited knowledge of the 108  composition of the innermost MFGM monolayer (32). More than 190 intracellular, extracellular and membrane associated proteins have been identified in the human MFGM (71). Clearly, the breast-fed infant is provided with the entire MFGM and not some of its individual components separately, as is the case in enteral feeds including formula.  The recent availability of commercial bovine MFGM provided the possibility for animal studies and clinical trials to explore the role of the entire globule membrane in infant health.   To our knowledge, this study is the first to investigate the effect of the entire MFGM on the developing brain in an animal model. Interestingly however, as our experiments were ongoing, results from one clinical trial of MFGM supplementation in infants, using the same MFGM powder we have used (MFGM-10, Arla, Denmark) was published (178). Swedish infants fed formula with MFGM (4% total protein by weight from MFGM) from 2 to 6 months scored 4.0 (95% CI: 1.1, 7.0) points higher in cognitive testing at 12 months using the Bayley Scales of Infant and Toddler Development than infants fed formula without MFGM, with the former scores being similar to those of a reference group of breastfed infants (n=80/group). These results implicated the MFGM in neurodevelopment. More interestingly, finding a difference in cognitive scores 6 months after MFGM supplementation was stopped suggests that a morphological change in the brain may have occurred, lasting enough to alter future cognitive function, at least during infancy. The supplementation period (i.e, 2-6 months) coincides with an active phase of brain development, characterized by both rapid synaptogenesis and myelination (113, 125). Components of the MFGM have been shown to be essential for these processes, including DHA, gangliosides, sialic acid and ethanolamine-PL (24, 157, 158). But whether the MFGM affects the brain simply by providing these components or by other more complex ways is not known and is difficult to elucidate using clinical studies. The limitations associated with 109  clinical research (eg, ethical concerns over access to infant brain tissue, environmental confounding variables) make it necessary to use alternative research methods to better understand whether and how the MFGM affects the brain.  In our study, feeding MFGM to infant rats did not change brain weight, which was also similar in the experimental groups to the MR group. However, lipid analyses revealed several differences in the brain secondary to feeding MFGM, particularly higher Pls-PE and Sph at d10 and lower PE (i,e, diacyl-PE) at d18. The decrease in PC-to-PE ratio, well-known to occur with brain development, was closer to that in MR rats, when MFGM was added to the formula. Accordingly, our data suggest the possibility of faster accumulation of PE when MFGM is not fed, which intriguingly was not associated with an increase in net DHA accumulation in brain PE.  Rapid gray matter growth, including synaptogenesis is characteristic of the rat brain during the first 12 days postnatal (128). Dorninger et al (145) have recently shown that when the adult murine brain is deficient in Pls-PE, total brain PE content is conserved by increasing the brain’s diacyl-PE content. Our results suggest that when MFGM is not fed, brain Pls-PE is reduced but total PE content is maintained, in membranes synthesized during rapid gray matter development (i.e, synaptogenesis around d10 postnatal). The implications of decreasing the Pls-PE-to-diacyl PE ratio in the neuronal membrane on its functionality during development are not known but could be significant. Plasmalogens, and not other membrane PL in synaptic membranes, are known to facilitate rapid membrane fusion (137, 139, 140). Membrane fusion is essential for synaptic transmission (141) and hormone exocytotic secretion (142), both critical for normal brain function. At least in vitro, Pls-PE is part of cell membrane lipid rafts together with Sph and cholesterol (38). Whether their ratios in the cell membranes in the brain is important for lipid raft formation is unclear. However, if so, this might explain the lower Sph also found at d10.  110  Finally, our finding of a lower PE-to-PC ratio when MFGM is not fed, was not expected, as the increase in brain PE is usually a positive marker of development (128, 133). Unlike in the human, myelination in the rat begins postnatally, at approximately 12 days postnatal as synaptogenesis continues. With PE increasing in the synaptic membranes and myelin, understanding which membrane is affected by feeding the MFGM is complicated, when a section of the brain is homogenized for analysis, as in this study. Further complicated, fatty acids in PE of the synaptic membranes are predominantly unsaturated (i.e, DHA), whereas PE in myelin contains mostly saturated and monounsaturated fatty acids (130, 266). We found no difference in net DHA accumulation from PE at d18, despite the higher diacyl-PE content, when MFGM is not fed. Whether this is explained by a difference in proportion of myelin in the brain with MFGM feeding merits further investigation. In fact, Liu et al (173) have shown several areas with more gray and white matter in the cortex of artificially-reared neonatal piglets, fed formula with MFGM lipids (Lacprodan PL-20, Arla, Denmark). The supplementation period from 2-28 days postnatal is equivalent to human infancy.  Alternatively, decreased PE DHA turnover may explain lack of difference in net DHA between the experimental groups. It is interesting that when MFGM lipids were supplemented to breastfeeding rats (i.e, not artificially-reared) until adulthood, synaptic function and not brain PL composition differed from control rats (174, 175). Dopamine output, known to be altered by omega-3 fatty acid supply particularly ARA and DHA (263), was increased with supplementation, suggesting either increased synapse activity or increased synaptogenesis due to feeding MFGM lipids (175).  Here, metabolite concentration differences were also found between MFGM-fed and no-MFGM rats. Why these were lower in the MFGM-fed rats is difficult to answer: Are the metabolites decreased with MFGM feeding due to increased use or 111  decreased need for them in the brain? Glucose is the conventionally accepted substrate for energy production in the brain, and lactate build up, from hydrogenation of pyruvate, is postulated to occur when the rate of oxidative metabolism is lower than the rate of aerobic glycolysis (production of pyruvate from glucose) (267). Higher lactate may indicate lower oxidative metabolism and therefore lower energy availability for brain function, when MFGM was not fed. Whether lactate is used as a primary energy source in neurons is still subject of debate (267).  Glutamate, aspartate, glycine and alanine are amino acids with neurotransmission functions (268). Glutamate, as excitatory neurotransmitter (NT), is also well-known as precursor of the inhibitory NT gamma-aminobutyric acid (GABA). The latter with glutamine ensures adequate glutamate turnover in the brain and protection against excitotoxicity (269, 270). Aspartate is similarly excitatory with neurotoxic potential (ref). Whether higher brain glutamate and aspartate with similar GABA concentration in the rats not fed MFGM indicates susceptibility to excitotoxicity is not known. Interestingly, both glutamate and aspartate are synthesized de novo in the brain, from intermediates in the tricarboxylic acid cycle and the BBB prevents their net entry into the brain (271-274). This suggests the effect of the MFGM found in brain tissue may be in part due to specific changes in brain metabolism rather than in other organs. Conversely, whether the MFGM contributes to the normal development of the BBB is another area to be explored, potentially explaining why glutamate and aspartate are lower in rats fed formula with MFGM.  Little is known on the functional roles of threonine and serine in the brain. However unlike glutamate and aspartate, their physiological brain concentrations are dependent on uptake from plasma (268, 274). Therefore, whether their lower levels in MFGM-fed rats are due to lower circulating plasma concentrations secondary to MFGM effects on other organs requires 112  better understanding. Inositol, likely myo-inositol, in MFGM-fed rats was almost 3 folds lower than in no-MFGM rats, the biggest fold difference among other metabolites. Inositol functions as cerebral osmolyte and is a key signaling molecule, as component of PI and inositol phosphates (275, 276). The inositol triphosphate (IP3)/Ca2+  signalling pathway is well-known to control several cellular processes, which in the brain, have been implicated in synaptic plasticity (277). Glycerol-3-P is an intermediate in glycolysis and precursor for de novo synthesis of PL. Lower glycerol-3-P in MFGM-fed rats suggests higher PL turnover, given than total PL content was not different between the experimental groups. In conclusion, this study has shown that feeding MFGM results in differences in several brain PLand metabolites during rat development, specifically at d18. The results suggest that the MFGM may havebiological roles relevant to the human brain. If true, the implications to feeding the formula-fed infant and infants on enteral feeds can be significant. We raise several questions that need to be addressed in future studies: Which component (s) of the MFGM mediate (s) brain development and how? Do Pls-PE in MFGM play a role in brain development? What is the functional relevance of these differences in brain lipid and metabolite composition to the human infant?         113  Chapter 4: Human Milk Choline- A Potential Player in Infant Milk Lipid Handling 4.1 Introduction  Choline (N,N,N-trimethylethanolammonium) was established as an essential nutrient by the Institute of Medicine in 1998, based on evidence of insufficient hepatic de novo synthesis in adults (190). Recommended intakes for infants were set as Adequate Intake for total choline (AI, 0-6 mo: 125 mg/d and 7-12 mo: 150 mg/d) based on its average content in milk from healthy women (187, 188), assuming the milk would be nutritionally adequate in supporting growth and development. Data used to set the AI were from two papers published in 1986 and 1996(187, 188), - that reported on only 26 milk samples combined, with analytical methods of limited accuracy and precision used (i.e radioenzymatic method and phosphorus assays).   Not addressed in current recommendations, human milk provides the infant with various forms of choline, with the three water-soluble choline compounds (WSC) free choline (FC), phosphocholine (PhosC) and glycerophosphocholine (GPC) contributing to around 90% of total choline compounds in milk. The lipid-soluble PC and Sph in the MFGM on average account for the remaining 10% (185-189). It is not clear whether the different choline forms in milk differ in bioavailability, metabolic fate, and functional role in infant development. However, it is intriguing that the mammary gland provides the infant with WSC, when the majority of choline compounds in circulation is lipid-bound (185, 186). Interestingly, plasma free choline concentrations in formula-fed infants are around half those of breast-fed infants (10.8  2.42; 21.8  7.61 mol/L, respectively), despite similar total choline content in infant formula and human milk (185). Instead, choline in infant formula is predominantly in free form. This raises the question of differences in bioavailability and physiological importance of the different forms of choline in the infant diet.  114  Choline has crucial metabolic and physiological roles, which importantly involve choline in different forms.  These roles can be categorized into three areas (Figure 4.1): maintenance of cell membrane structural integrity and signaling functions when in membrane PC and Sph, donation of methyl groups (CH3) via betaine, the oxidized form of choline in the liver and kidney particularly for the generation of S-adenosylmethionine (SAM) (i.e involved in DNA methylation and PC synthesis), and cholinergic neurotransmission in the nervous system when choline is acetylated to form acetylcholine (191, 192). Our lab is currently investigating the potential metabolic importance of WSC in the infant diet. PhosC and GPC may be important to ensure adequate PC synthesis via the CDP-choline pathway, thus limiting the use of PE and methyl groups as substrates for PC synthesis and sparing free choline for extra-hepatic functions (eg. neurotransmission). In developing this research, it became evident to us that the current literature on human milk WSC is limited and inconsistent, with the need to better understand WSC composition in milk, and therefore infant intake, as we address their biological importance.  The current literature on human milk choline compounds describes a wide range of amounts for total choline and choline forms particularly the WSC, with only one study on preterm milk (Table 4.1). The wide variability in milk WSC across studies or within a study, for example having an SD of 544 with a mean of 577 µmol/L (n=16) for PhosC (188) is in part due to study limitations, including small sample size. Comparison across studies is complicated by the different analytical methods used, including radioenzymatic methods (185, 187), HPLC (185), HPLC followed by GC-MS (188), NMR spectroscopy (189) and LC-MS (186). With only one published paper on WSC in preterm milk, n=8 (188) and no studies on pasteurized term donor milk, our objectives here were to determine and compare the WSC concentrations in preterm and term milks fed to preterm infants in our neonatal intensive care (NICU) using 115  advanced LC-MS/MS methodology and to stimulate interest in addressing infant choline intake beyond average total choline content, extending to the different forms of choline in milk.  Table 4.1 Composition of choline compounds in mature term and preterm milk     Year/ Place Zeisel et al (187)  1986/USA Holmes-McNary et al (188)  1996/USA Holmes et al (189)  2000/England Ilcol et al (185)  2005/Turkey Fisher et al (186)  2010/USA                                                    Term milk  Sample size, n 10 16 8 95 48 FC PhosC GPC PC Sph Total  73 ± 21 - - 140 ± 32 188 ± 31 -    116 ± 88 570 ± 544 362 ± 280 82 ± 24 124 ± 36 12541  210 ± 141 480 ± 198 410 ± 226 100 ± 28 100 ± 28 1280 ± 396  V V V 228 ± 97 551 ± 32 499 ± 155 104 ± 107 94 ± 88 1476 ± 468  83 ± 54 551 ± 322 388 ± 168 107 ± 47 67 ± 27 11961   Other milk types      Preterm milk Term colostrum Term colostrum  Sample size, n  17 8 21  FC PhosC GPC PC Sph Total    98.0 ± 186 693 ± 487 379 ± 173 90 ± 54 104 ± 37 1364  210 ± 141 480 ± 198 410 ± 226 100 ± 28.2 100 ± 28.2 1280 ± 395 132 ± 96 93 ± 119 176 ± 59.5 146 ± 82.0 129 ± 59.5 676 ± 160  Values are means ± SD (µmol/L).  SD is calculated from published standard error for means and sample size.  1 Total is calculated from means of individual choline compounds.  FC, free choline; PhosC, phosphocholine; GPC, glycerophosphocholine; PC, phosphatidylcholine; Sph, sphingomyelin.         116  Figure 4.1 Overview of choline metabolic pathways    4.2 Methods The milk samples used in this study were collected as part of a nutrient quality assessment study for the milk fed at BC Children’s Hospital NICU in Vancouver, to determine the extent to which current standard fortification protocols and the variability in milk lipid and energy contents contribute to infant malnutrition (manuscript in preparation). Milk samples were collected daily for a duration of one month, between 7:00 am and 9:00 am, coinciding with the timing when nurses typically prepare milk feeds. It is standard of practice for the NICU nurse to place the milk vials by the bedside, allowing milk to reach room temperature. Once the desired temperature was achieved, indicated by the nurse, fresh or thawed milk samples (500 µl) were taken from the vials using single-use sterile syringes, only when the available milk exceeds the volume needed to feed the infants. Type of the milk (preterm milk (PM) or donor milk (DM)) 117  was recorded. Personal identifiers for the mother or the infant were not collected and samples were given random codes for laboratory analyses. Samples were placed on ice and transferred immediately to the nutrition lab within close proximity to the hospital. The study protocol was approved by the University of British Columbia – Children’s and Women’s health centre of BC research ethics board.  4.2.1 Biochemical Assessment Milk samples were immediately analyzed in duplicates for total lipid content (g/dl) using the creamatocrit method (278). Aliquots for the analysis of total protein (100 µl) and choline-containing compounds (20 µl) were then taken and samples stored at – 80 degrees for future analysis. Total protein content was determined by the Bradford method (279). Results for total lipid and protein content are presented here, for descriptive purposes. FC, PhosC and GPC were analyzed using isotope dilution liquid chromatography tandem mass spectrometry (LC-MS/MS), using a Waters ACQUITY UPLC system connected to a Quattro Micro tandem MS configured with an electrospray source (Waters Corporation, MA, USA). The LC included a 2.1 × 12.5 mm pre-column and a 2.1 × 150 mm Zorbax Rx-Silicon column, both with 5 µm particle size (Agilent Technologies, CA, USA). Two mobile phases were used: A= acetonitrile with formic acid (0.1%) and trifluoroacetic acid (0.1%) and B= ammonium formate (15 mmol/L) with formic acid (0.1%) and trifluorocetic acid (0.2%) in water. The solvent gradient started with 95% A and 5% B the first 2 min, followed by a linear gradient to 55% A and 45% B at 4 min, maintained till the end of the run such that the total analytical time was 9 min. Flow rate was maintained at 0.5 ml/min. The LC column and autosampler were maintained at 25°C and 5°C, respectively. The MS was operated in positive ion multiple reaction monitoring (MRM) mode using the following transitions [m/z (compound)]: 103.9 /59.9 (choline), m/z 113/68.9 (choline-d9), 184.1/125.0 118  (PhosC), 193.1/125.0 (PhosC-d9), 258/124.9 (GPC), 289.0/221.0 (GPC-d9).The analytical inter-assay and intra-assay coefficient of variations (CV) for each of the WSC compounds, based on 5 replicate analyses, were as follows: for PhosC, 6.4% and 5.2%, respectively; for FC, 5.5% and 4.1%, respectively; and for GPC 9.5% and 2.3%, respectively (4). 4.2.2 Statistical Analysis The normality of data distribution was tested using Kolmogorov-Smirnov test, and results are given as mean ± SD, or median and ranges as appropriate. Differences in nutrient content between PM and DM samples were analyzed using independent samples t-test or Mann Whitney U test for normal and skewed data, respectively. Statistical analyses were done using SPSS version 21 and a P-value < 0.05 was considered statistically significant.   4.3 Results  A total of 45 DM and 219 PM samples were collected, from which n=40 DM and n=90 PM samples were randomly selected for WSC analysis. The PM samples were from mothers who have been producing milk for a median (range) of 31 (6-93) days at the time of sample collection. DM samples are prepared in batches by pooling and pasteurizing term milk from at least three different donors. The total lipid, but not protein content in PM and DM samples was skewed (Figure 4.2), with an expected wider variability for lipid among PM (3.02±1.54 g/dL; median: 3.02; range: 1.30-6.79) than DM samples (2.76 ± 0.98 g/dL; median: 2.68; range: 1.10-4.77). PM samples were significantly higher in total lipid (P=0.03) and protein content than DM (3.5 ±0.5; 3.2 ±0.4 g/dL respectively; P<0.001). Assuming 50% of the calories in milk are from lipid (20), estimated energy content in PM would be also higher than in DM samples (63.3±27.8 kcal/dL; median: 54.4; range: 23.6-122 and 49.7±17.6; median: 48.2; range19.2-85.8, respectively P=0.003). 119  Figure 4.2 Box plots of the distribution of total lipid and protein content (g/dL)  in preterm and donor milk samples                               PM, preterm milk; DM, donor milk                              Difference in total lipid analyzed using Mann Whitney U-test (P=0.03)                              Difference in total protein analyzed using Independent student t-test (P<0.001)                     Table 4.2 Content of water-soluble choline compounds in preterm and donor milk samples  120   Mean ± SD Median (range) PhosC PM DM  791 ± 399 719 ± 251  794 (3.4-2071) 701 (50.8-1223) GPC  PM DM  400 ± 282 377 ± 192  344 (33.5-1813) 384 (60.5-772) FC PM DM  211 ± 142 157 ± 88.4  179 (34.2-685) 147 (16.3-296) Value are µmol/L; SD, standard deviation PM, preterm milk (n=90); DM, donor milk (n=40)                     PhosC, phosphocholine; GPC,   glycerophosphocholine; FC, free choline                     Differences between groups analyzed using Mann Whitney U test (P>0.05)  The milks fed at our NICU contained variable amounts of WSC, with no significant differences between PM and DM samples (Table 4.2).  Consistent with the results from the limited number of available studies (Table 4.1), PhosC and GPC, not free choline, were the major forms of choline in milk. Ranges are presented in the table to highlight the wide variability in WSC, particularly for PM PhosC ranging from 3 to ~2000 µmol/L (The closest value after 3 µmol/L was 55.2 µmol/L). Our data renders the question of whether the different forms of WSC within one milk sample are interrelated, as valid (i.e whether samples with low PhosC were high in other WSC forms). No significant correlations were found between the different WSC in PM, with some milk samples either relatively high or low in all their WSC forms, and others inconsistently variable in their WSC forms (Figure 4.3).  This emphasizes that not only do infants consume variable amounts of total WSC from milk (1403± 542; median: 1370; range: 121  544-3865 µmol/L), but the proportion of the different WSC forms also differs, for which the biological implications are not understood. Of note, correlations in DM were not explored, as the milk samples were not from one biological sample (pooled milk).  When mother’s own milk is insufficient in volume to meet infant needs, pasteurized term donor milk is alternatively provided for infants at our NICU. Our finding of no significant differences in WSC between PM and DM, with DM WSC content consistent with previous results in term milk, suggests limited effects of gestational stage (term vs preterm) and/or donor milk handling (i.e, pasteurization) on WSC.   Figure 4.3 Scatter plots of the relationship between the different WSC forms in preterm milk 1 Note the differences in the scales of the different axes in the figure Associations determined by Pearson’s Correlation; n=75  122  4.4 Discussion  The results presented in this study are part of a bigger research project addressing the biological roles of the different forms of WSC in human milk and their potential interaction with lipid metabolism. Here we report two main findings: 1) WSC content and composition are highly variable across milk samples and do not differ between PM and term donor milk and; 2) Individuals WSC components in milk are not associated with one another.  Our findings on variable total and individual WSC contents are consistent with others (185-189), adding to the available literature than the individuals WSC components are not interrelated. The biological importance of the total amount of WSC intake by the infant remains a question to be answered, given the variability in total WSC content in milk. In exploring the composition of individual milk samples (Figure 4.3), we add another complex question: If infants also consume variable proportions of individual WSC components, does the latter vary in biological role, independent of total WSC intake?  Figure 4.3 Percent distribution of the different WSC compounds in individual PM samples  PM, preterm milk; n=75  123   In relation to the brain, our speculations include the presence of an interrelation between the form of choline availability in circulation (to which the dietary WSC contribution is not known) and supply of DHA to the brain. As recently reviewed (245, 280), DHA supply to the brain originates either from plasma albumin-bound non-esterified DHA or delivered by choline-containing plasma carriers: albumin-bound lyso-PC and lipoproteins, mainly chylomicrons and very low density lipoproteins (VLDL). Interestingly, not only does lyso-PC carry DHA in plasma, but the molecule itself was recently found to cross the BBB via a major facilitator transporter Mfsd2a (281). The role of lyso-PC as preferential source of brain DHA compared to non-esterified DHA (248, 282) is more appreciated with recent evidence of reduced brain DHA in Mfsd2a-knock out mice (281). However, we suggest the need to explore the concept of brain efficiency, such that the brain uptakes one molecule (i.e lyso-PC) as source of two fundamental compounds in brain function (DHA and also choline). Moreover, circulation of chylomicrons is often simplistically discussed as transport of TG from the enterocyte to the liver via lymph. However, chylomicrons transport TG-DHA to the brain, with DHA crossing the BBB after cleavage by endothelial lipoprotein lipase (280). PC, a major PL needed for chylomicron formation in the enterocyte, is derived from de novo synthesis in the enterocyte rather than from esterification of lyso-PC, a by-product of PC digestion under normal human physiological conditions (283). The extent to which the individual WSC compounds in milk are uniquely important precursors for chylomicron PC is not known and is challenging to study, as this would require access to chylomicrons in the mesenchymal lymphatic duct and a better understanding of PhosC and GPC digestion and absorption. WSC compounds are absorbed via the portal circulation, but the proportion of PhosC and GPC used for enterocyte PC synthesis is not known (284). 124   Similar to the unknown role of dietary WSC in DHA delivery to the brain, whether dietary milk WSC contribute differently to brain choline metabolism is not known. De novo synthesis of choline in the brain is insufficient to sustain its cholinergic functions, with plasma FC transport across the BBB (carrier-mediated and saturable) being crucial for maintaining brain choline homeostasis related to the neurotransmitter acetylcholine and membrane PC (207). Therefore, the lower plasma FC in formula-fed compared to breast-fed infants despite similar total choline intake (185) suggests individual WSC compounds differ in bioavailability and/or metabolic fate, with potential implications to brain choline metabolism.  From a clinical perspective, human milk banking and the use of pasteurized term donor milk when mother’s own milk is unavailable or sufficient is increasing, with increasing evidence of superiority of DM over formula in protecting against necrotizing enterocolitis and improving feeding tolerance (285, 286). The biological safety and nutrient adequacy of donor milk are main concerns in feeding infants during critical periods of illness and growth restriction. Pasteurization of donor milk impacts milk nutrients at varying degrees, with the biggest decrease found in heat-sensitive water-soluble nutrients including vitamin C, B6 and folate (286). Our data suggests little impact of pasteurization on WSC, however this needs to be confirmed by comparing WSC content in the same milk sample pre- and post-pasteurization. Although PM and donor milk were not found to be different in WSC content, whether they are adequate in meeting infant requirements remains to be determined. Relevant to our work, human milk fortifiers include choline in free form and are supplemented to milks fed at our NICU by volume (1120 mg/l), based on average total choline in term milk. In utero, FC is actively transported across the placenta to the fetus, with newborn plasma FC significantly higher than in maternal plasma (287-289). Intriguingly, plasma FC concentrations of preterm infants at the NICU (n=56) were found 125  to have around half the concentrations that would be found in cord blood of newborns (n=176) at equivalent gestation ages, between 24-42 wks (290). As noted by the authors (290), concern over the implications of the lower plasma FC on the preterm brain arises from evidence that 25% of infants had plasma FC concentrations around the cut-off (14 µmol/L) at which choline efflux occurs from the rat brain back into plasma. Whether postnatal metabolic adaptation, such as the increase in hepatic synthesis of PC-containing VLDL requires a dietary supply of PhosC and GPC as in milk, compared to prenatal FC placental supply needs to be addressed. Of note, the placenta seems to be able to metabolize FC into GPC, with the only evidence being in vitro (291).  At least in the term rat infant, the choline moiety of dietary PhosC, GPC and FC appears in and disappears from the liver at different rates, with faster rate of clearance of choline from PhosC and FC compared to GPC, suggesting different metabolic handling (284).            126  Chapter 5. Conclusion 5.1 Summary of Specific Contributions The overall objective of this dissertation was to gain a better understanding of the unique lipid composition of human milk, particularly MFGM PL and to address the potential role of the MFGM in infant brain development. This PhD work also included an exploratory section that addressed the composition of WSC in milk. The conceptual significance of this research is that lipid structures in human milk contribute to infant development, beyond the roles of ω-3 and ω-6 fatty acids and there needs to be an integration of research areas extending to other milk components, such as WSC when addressing lipid-related functional outcomes. The specific contributions of this work by chapter, are listed below. 1. Separation and recovery of milk Pls-PE for fatty acid analysis is achievable using SPE and HPLC. LC-PUFA including DHA are higher in Pls-PE than diacyl-PE, and PC in cow and human milk. Maternal DHA intake (17-354 mg/day; FFQ) is not related to DHA in Pls-PE, diacyl-PE and PC. (Chapter 2)   2. Brain PL and metabolite composition differ between artificially-reared neonatal rats fed formula with or without MFGM, with PL composition in the MFGM-fed group closer to that of the group of mother-reared rats. (Chapter 3) 3. WSC content and composition are highly variable across milk samples fed at BC Children’s Hospital NICU and do not differ between preterm and term donor milk. No association was found between the individual WSC components (FC, GPC and PhosC) in milk. (Chapter 4) 127  5.2 Strengths and Limitations The main strength of this work is the combination of multiple nutrition research study designs to address the complexity in understanding human milk components and infant brain development: analytical method development, two cohort studies in healthy lactating women and infants admitted to the NICU, and one study in animals. Table 5.1 summarizes the strengths and limitations of the developed analytical method, followed by a discussion on the strengths and limitations of other aspects of this research. Table 5.1 Strengths and limitations of the analytical method for the separation and recovery of selected milk PL.   In the cohort study described in Chapter 2, n=4 women provided milk samples on separate days, with data showing DHA in Pls-PE is indeed consistently higher than in diacyl-PE and PC within one woman. Nonetheless, the implications of the variability in DHA proportion in milk PL within and across women on infant total DHA intake and the net PL intake could not be Brief summary: Milk lipids are extracted by a modified Folch method. PL are separated and recovered in methanol by SPE using amino propyl cartridges. Pls-PE, diacyl-PE and PC are separated and recovered by HPLC, connected to two diol columns and an ELSD detector. Strengths Limitations Volume needed per sample (5 ml) is feasible to obtain for mature term human milk by electrical or hand pumping. Method optimization to reduce aliquot volume is needed for colostrum and preterm milk (may be more challenging to pump).  Accurate and reproducible extraction and separation of Pls-PE, diacyl-PE, and PC Loss of PS and PI (>50%), with further work needed to extract them accurately from milk Method allows recovery of Pls-PE, diacyl-PE and PC for fatty acid analysis under non-lipolytic conditions (i.e choice of HPLC mobile phases and solvent drying technique) Method is non-quantitative for milk PL and can be time consuming, which limits high throughput analysis (3.5 hour total per sample)  128  determined with this study design and current methodology. Additionally, the unexpected lack of significant association between DHA intake among lactating women and DHA in total milk lipid may be due to the small range of total milk DHA (0.1-0.8%) and limited variability in DHA intake, with more than 50% of women consuming 125 mg/d DHA or less. Of reference, 2.5 ounces of salmon, equivalent to one serving of fatty fish, contain approximately 1200 mg DHA (ESHA Food Processor, Salem, OR, 2009). It is also possible that using an FFQ, averaging intake over the last month, to assess dietary DHA intake does not reflect the absolute amount of dietary DHA contributing to DHA in the collected milk sample. For example, a woman who may have consumed one serving of salmon the day before milk collection (1200 mg DHA) but no additional dietary source of DHA during the last month and a women who consumed smaller daily amounts of DHA (eg. one omega-3-rich egg with 43 mg DHA) for the duration of one month would have a similar average daily DHA intake (43 mg/d). Therefore, dietary intake assessment methods best suited for accurately measuring dietary DHA which contributes to DHA in milk needs to be explored further.  As discussed in Chapter 3, we used an experimental model of artificially-feeding (i.e, pup-in-the-cup) to feed infant rats formula, with the only difference between the experimental formulas being the presence/absence of the MFGM. The strength of this design is in allowing complete control of the volume of formula fed and the composition of the formula. By using peristaltic pumps instead of bolus gastric injections, we mimicked the infant’s suckling pattern and thus the volume of formula entering the stomach at any particular time. Artificial-rearing was also used as means to bypass the unknown but potential effect of maternal physiology on the composition of the MFGM. For instance, in studies on essential fatty acids and DHA, researchers keep the infant rats with their dams and only manipulate the fat and fatty acid content of the maternal diet 129  to alter milk composition (92, 251-254). However, this approach would not possible for composition and biological determinants between the human and the rat. On the other hand, several limitations associated with the study design need to be addressed when interpreting the data. First, the inclusion of a reference group of mother-reared rats was only exploratory to develop reference data on brain PL composition of the Sprague-Dawley rat during early post-natal develop. Therefore, comparisons between the experimental groups and the mother-reared group should be made with caution, with differences between the groups extending beyond the presence or absence of the MFGM: bioactive milk components not present in the formula, such as growth factors and hormones (256, 257) and lack of maternal touch which may promote brain development (292). Second, it is possible that additional brain components, such as gangliosides, cerebrosides, sialic acid, PS and PI, which we did not measure due to logistic limitations, may have differed between the experimental groups and therefore, would have given additional insight into how the MFGM affects brain development. Third, the rat brain at birth is less developed than that of the human and the age at which different morphological brain changes occur also differ. By starting artificially-feeding at d5 and not birth, we intended to reduce the discrepancy between the two species in terms of brain maturity. Nonetheless, our findings need to be interpreted with acknowledgment of these differences. For example, myelination in the rat mostly occurs between d13-d18, a period when milk is still the only nutrition source. In humans however, myelination begins prenatally, has maximum rate during the first 2 years of life and continues into adulthood. Therefore, whether the potential role of the MFGM in myelination is more significant in the rat compared to the human merits further investigation. Before reporting the WSC content of PM and donor milk samples from the NICU in Chapter 4, knowledge on milk WSC was limited to a few studies, for which analytical methods were 130  inconsistent. This limited our ability to compare findings across studies and consider the reported high variability in milk WSC as conclusive.  Based on our findings, the high variability in milk WSC was confirmed, using LC-MS/MS as the same analytical method for all three WSC compounds. However, the concentrations of WSC reported in Chapter 4 should not be used to estimate daily WSC intake of infants in the NICU because milk samples were collected from separate vials available for nurses to prepare the morning feeds. It is possible that milk from two or more vials were mixed to obtain the required volume for one feed and that WSC in morning feeds do not reflect intake over 24 hours. On another note, the creamatocrit method was used to determine milk lipid content (278), because of its simple and relatively quick procedure, allowing us to clinically evaluate the nutritional quality of the milk within 24 hours of its receipt into the NICU. However, milk lipid content measured using a human milk analyzer would have been determined more accurately.  5.3 General Comments This research has also raised several important implications related to work towards the understanding the role of milk lipids in brain development.  1) The role of dietary lipids in brain development cannot be fully understood without further understanding lipid digestion/absorption and the interaction of milk components in the gastrointestinal (GI) tract.  In Chapter 2, we suggested the need for studies that determine the digestion/absorption process of Pls-PE in humans, particularly in the infant and with Pls-PE in the MFGM. This is one example of several dietary lipids for which limited knowledge in their biological role is attributed partly to insufficient knowledge on GI physiology in humans. These include dietary PS, PI, PC, diacyl-PE and Sph. For instance, whether choline in PC remains lipid-bound after 131  digestion may help understand its metabolic fate and therefore its functionality (eg. part of plasma DHA- carrying lyso-PC compared to precursor of betaine, when in free form). Whether PL digestion/absorption differs when consumed separately or as part of the MFGM also needs to be understood when discussing their potential addition to enteral formulas and functional foods.  2) Researchers using data from one milk sample to associate milk lipid composition with health outcomes and/or usual dietary patterns need to acknowledge the potential for poor content validity.  It is common to use one milk sample per study participant to associate milk lipid composition with maternal dietary intake and/or infant health outcomes, with examples for studies on milk DHA cited here (293-298). However, as shown in Chapter 2, milk lipids vary within a woman in total amount (eg. g/dL) and in fatty acid composition, both in total lipid and selected PL. Therefore using one milk sample may not always be representative of usual milk lipid composition and infant intake, which are more likely to affect health outcomes. Additionally, reporting milk PL and fatty acids as a percentage without reporting milk lipid content (g/dL) provides limited information on actual infant intake. Quantitative measures for infant PL and fatty acid intakes in future studies are encouraged.  3) The potential importance of the nature of lipid molecules to which milk DHA are esterified needs to be addressed when debating whether dietary recommendations for DHA during infancy need to be developed.   Whether DHA deficiency occurs in infancy is beyond the scope of this research. However as part of the ongoing debate of whether DRI’s for DHA for infants needs to be set, our research adds a couple of complex questions that should be addressed: 1) Is infant DHA intake based on the total amount of milk DHA, reflecting DHA in milk TG and/or based on the amount of DHA 132  in selected milk PL (eg. Pls-PE) appropriate as dependent variable in assessing DHA deficiency?  2) The recommended intakes of most nutrients for the infant, mostly as AI, are based on average nutrient composition in human milk. If an AI for DHA during infancy were to be set, how can an average DHA milk content be calculated knowing that milk DHA may vary within a woman? 4) Understanding the role of milk lipids in brain development may require addressing their role in other organs as well. In Chapter 3, we reported findings of altered lipid and metabolite composition in the brain due to feeding MFGM but whether the MFGM indirectly affects the brain through roles in other organs is an interesting area of research that is yet to be explored.  For example, we reported significantly lower brain threonine (essential amino acid) and serine, at d13 and d18 respectively, in rats fed formula without than with the MFGM. These amino acids are not known to have functional roles in the brain and intriguingly can cross the BBB. Therefore, if threonine cannot be synthesized de novo, is it lower in the brain due to its increased use in the brain or other organs? Are changes in brain amino acids associated with changes in plasma? It is not possible to differentiate between free or bound amino acids in our metabolomics study, however it would be interesting to address their characteristics more specifically in the future.  The gut-brain axis, the characterization of the infant gut microbiome under different dietary exposures and the role of the MFGM in brain development are three emerging research topics, but often studied in isolation from one another. Interestingly, changes in the gut microbiome result in changes in brain neurotransmitters and behavior in rodents, although the underlying mechanism (s) by which this happens is not known (299, 300). Our lab is currently investigating whether the MFGM shapes the infant gut microbiome during early development by using a similar pup-in-the-cup model to the one described in Chapter 3. If the MFGM affects the gut 133  microbiome, then it may be biologically plausible the role of the MFGM in brain development may be mediated in part by a microbiota-driven gut-brain axis.  5) Interpreting findings from human studies on milk is complicated by the potential association between nutrients. One limitation of human studies is that foods, not nutrients, are consumed; therefore, it is challenging to interpret which dietary components explain statistically significant associations between dietary and biochemical variables. As highlighted in section 1.3 of the literature review, given that nutrients in human milk are diverse and complex in their structures and arrangements within the milk matrix, it is unlikely that milk lipids alone play a role in development, including that of the brain. In Chapter 3, we reported brain lipids differ in composition between rats fed formula with or without the MFGM. Which MFGM component (s) is (are) responsible for this difference is not known. However for dietary PC and Sph, as an example, it may be difficult to attribute a role for them in development using human studies when choline, one of their constituent molecules is also known be involved in development and is present in other forms in milk: Free and bound in GPC and PhosC. In addition, mammary gland de novo synthesis of PC and Sph is little understood, and whether the amount of dietary PC and Sph are related to the amount of WSC in milk is not known. Therefore, interpreting findings from human studies on milk would be facilitated if animal models are used such that diets are designed to provide different forms of choline, with known absolute amounts added rather than percent composition, which would help determine which form (s) determines a biological role.  5.4 Future Directions The results of this work have raised important considerations for future research.  Our analytical method presented in Chapter 2 for the first time enabled the determination of the fatty 134  acid composition of milk Pls-PE, diacyl-PE and PC. Although it was determined that human and cow milk Pls-PE contains significantly higher PUFA, including DHA than other PL and TG, the method was not quantitative for these milk PL. An important next step in determining the physiological determinants and importance of milk PL and their fatty acids to the infant is to develop quantitative methodology for PL analysis in milk. An accurate and precise method that allows the separation, quantification, and recovery of all milk PL including the acid-labile PS and PI would be ideal. In addition, our exploratory dietary data suggests DHA intake is not a determinant of DHA in milk PL. However, several limitations in our study need to be overcome in future work to determine whether this finding is replicable, by 1) Assessing dietary and milk PL DHA from a group of women with a wider range of dietary DHA intake than the one in our study 2) determining the dietary assessment method that best reflects recent dietary DHA intake (eg. FFQ, 24-hr recall, Food records), as means of reducing risk of a false negative (Type II error). In identifying other modifiable determinants of DHA in milk PL, the relative contribution of de novo DHA synthesis in both liver and mammary gland, potentially modifiable themselves by dietary PUFA intake, needs to be determined. Furthermore, our finding of significantly higher PUFA in Pls-PE of cow colostrum compared to mature milk is intriguing and raises the question of whether human colostrum Pls-PE is enriched in PUFA, for which the biological importance during the first week of life also needs to be determined. When discussing the potential role (s) of milk Pls-PE in infant health, future studies in human physiology need to undermine Pls-PE digestion which remains very little understood, including digestion byproducts, their absorption, post-absorption handling and bioavailability.  Our findings in Chapter 3 on differences in brain lipid and metabolite composition between infant rats fed formula with or without MFGM suggests the MFGM is involved in brain 135  development. Future studies need to address fundamental questions on species-specific differences in milk lipid composition and brain development, to allow accurate extrapolations from the rat to the human. These include: What is the composition of the rat MFGM, and how is the MFGM digested and handled post-absorption compared to the human? Which specific timeframe (s) in human brain development do changes in rat brain lipids between d10 and d18 reflect? Are the changes found in the rat brain (eg. lower PE at d18 due to feeding MFGM) expected to occur in the human brain, or are these changes only a reflection of altered brain metabolism in the rat, with the need to identify other markers of altered brain metabolism in the human? More importantly, future studies are needed to understand whether the changes in brain composition due to feeding MFGM are functionally relevant during infancy and/or other life stages. This can be addressed, for example, by exploring whether feeding MFGM to infant rats alters behavior and/or cognitive function using well-established human-related behavioral tests, such as the touchscreen cognitive testing method (301). Clinically, a next step to better understand the role of the MFGM in human infants is to determine whether differences in brain metabolites between MFGM-fed and non-MFGM-fed pups correlate with altered plasma or urine metabolites that have the potential to be used as markers in human infant studies (eg. plasma PL or amino acids). For example, the use of a cotton ball in the diaper is a simple and promising approach to collect urine for clinical evaluation (302, 303). Finding a urinary marker of altered brain metabolism due to feeding MFGM can then be measured non-invasively in infants, with urine collected using the cotton ball method. The importance of human milk lipids as source of DHA for the developing brain has been an active area of research driving, in part, the addition of DHA to infant formula and prenatal supplements and the development of consensus-based dietary recommendations for increasing 136  DHA intake by lactating mothers (6, 304, 305). However, as suggested in Chapters 2 and 3, milk lipids may be involved in brain development in more complex ways than simply providing DHA. Future studies are required to determine which component (s) of the MFGM are involved in brain development and how. For instance, do PL in the MFGM, including Pls-PE, contribute to brain PL? Does the type of lipid molecule in which DHA is esterified to (i.e Pls-PE, diacyl-PE, TG) dictate the metabolic fate of DHA and thus, its role in brain development? Is the structural arrangement of lipids and proteins within the tri-layered MFGM (eg. lipid rafts, glycolipids on the outside leaflet) important to ensure the bioactivity of the individual MFGM components? These research questions would help address whether the structure of lipids in the milk matrix is important, beyond the potential benefits associated with adding individuals components separately to the infant diet. Future studies also need to address the potential interaction between milk components, specifically milk WSC and the MFGM, in mediating brain development.  To better understand the role and/or interaction between the MFGM and WSC in delivering DHA and choline to the brain, additional research is needed to definitively determine the source (s) of DHA for update into the human brain and the potential role of plasma lyso-PC in delivering both DHA and choline to the brain (proposed concept of brain efficiency). As recently reviewed (245, 280), DHA supply to the brain may originate either from plasma albumin-bound non-esterified DHA or delivered by choline-containing plasma carriers: albumin-bound lyso-PC and lipoproteins, mainly chylomicrons and VLDL. Lyso-PC can also cross the BBB (281) and could therefore also be a source of brain choline. The contribution of the MFGM and WSC to plasma DHA and choline carriers also needs to be determined.  In chapter 4, we reported milk samples fed at our NICU, whether preterm milk or term donor milk, contain variable total amounts and composition of WSC, with no significant associations 137  between the individual WSC components in milk.  Future studies are required to determine the biological importance of the individual WSC in milk and the implications of feeding infants with milk having variable amounts of WSC across feeds.  These studies are clinically relevant, as current human milk fortifiers contain choline in free form and are supplemented to milks fed at our NICU by volume (1120 mg/l), independent of actual WSC content in milk.  In conclusion, MFGM alters lipid and metabolite composition in the infant rat brain. A method for the extraction and recovery of milk Pls-PE, a component of MFGM, has been achieved and our findings suggest Pls-PE may be a dietary source of LC-PUFA including DHA for the infant.  Future work is needed to determine how MFGM affects the human brain composition and function and whether milk Pls-PE is involved. In addition, human milk provides the infant with variable amounts and forms of WSC (FC, GPC and PhosC), for which their role in brain function and development remains unclear. We suggest the need to address the potential metabolic interaction between MFGM and WSC in infant brain development.             138  Bibliography 1. Andreas NJ, Kampmann B, Mehring Le-Doare K. Human breast milk: A review on its composition and bioactivity. Early Hum Dev 2015,91(11):629-35.  2. Capuco AV, Akers RM. The origin and evolution of lactation. J of Biol 2009;8(4):37.  3. Garcia C, Innis S. Structure of the human milk fat globule. Lipid Technology 2013;25(10):223-6.  4. Soberanes Garcia LS. Water–soluble choline compounds in human milk : their variation and impact of storage and diet. University of British Columbia, 2015. 5. Innis SM. Dietary omega 3 fatty acids and the developing brain. Brain research 2008;1237:35-43.  6. McNamara RK, Carlson SE. Role of omega-3 fatty acids in brain development and function: Potential implications for the pathogenesis and prevention of psychopathology. Prostaglandins Leukot Essent Fatty Acids 2006;75(4–5):329-49.  7. Zeisel SH. Nutrition in pregnancy: the argument for including a source of choline.Int J Wom Health 2013;5:193-9. 8. Zeisel SH. Choline: critical role during fetal development and dietary requirements in adults. Annu Rev Nutr 2006;26:229-50.  9. Rustan AC, Drevon CA. Fatty Acids: Structures and Properties. Edtion ed. eLS: John Wiley & Sons, Ltd, 2001. 10. Los DA, Murata N. Membrane fluidity and its roles in the perception of environmental signals. Biochimica et Biophysica Acta (BBA) - Biomembranes 2004;1666(1–2):142-57.  11. Martinez M, Mougan I. Fatty acid composition of human brain phospholipids during normal development. J Neurochem 1998;71(6):2528-33. 12. Kikuchi K, Sakai K, Suzuki T, Takama K. Fatty acid composition of choline and ethanolamine glycerophospholipid subclasses in heart tissue of mammals and migratory and demersal fish. Comparative biochemistry and physiology Part B, Biochemistry & molecular biology 1999;124(1):1-6. 13. Darin-Bennett A, Poulos A, White IG. The fatty acid composition of the major phosphoglycerides of ram and human spermatozoa. Andrologia 1976;8(1):37-45. 14. Goldfine H. The appearance, disappearance and reappearance of plasmalogens in evolution. Prog Lipid Res 2010;49(4):493-8.  15. Jenness R. The composition of human milk. Seminars in perinatology 1979;3(3):225-39. 139  16. Bitman J, Freed LM, Neville MC, Wood DL, Hamosh P, Hamosh M. Lipid composition of prepartum human mammary secretion and postpartum milk. J Pediatr Gastroenterol Nutr 1986;5(4):608-15. 17. Pons SM, Bargallo AC, Folgoso CC, Lopez Sabater MC. Triacylglycerol composition in colostrum, transitional and mature human milk. Eur J Clin Nutr 2000;54(12):878-82. 18. Prentice A. Constituents of human milk. Food and nutrition bulletin 1996;17(4):305-11. 19. Kehoe SI, Jayarao BM, Heinrichs AJ. A survey of bovine colostrum composition and colostrum management practices on Pennsylvania dairy farms. J Dairy Sci 2007;90(9):4108-16.  20. Jensen R. Handbook of milk composition. San Diego, California: Academic Press, 1995. 21. Nicholas KR, Hartmann PE. Milk secretion in the rat: progressive changes in milk composition during lactation and weaning and the effect of diet. Comparative biochemistry and physiology A, Comparative physiology 1991;98(3-4):535-42. 22. Innis SM. Polyunsaturated fatty acids in human milk: an essential role in infant development. Adv Exp Med Biol 2004;554:27-43. 23. Lawrence PB. Breast milk. Best source of nutrition for term and preterm infants. Pediatric clinics of North America 1994;41(5):925-41. 24. Innis SM. Perinatal biochemistry and physiology of long-chain polyunsaturated fatty acids. J Pediatr 2003;143(4 Suppl):S1-8. 25. Bitman J, Wood L, Hamosh M, Hamosh P, Mehta NR. Comparison of the lipid composition of breast milk from mothers of term and preterm infants. Am J Clin Nutr 1983;38(2):300-12. 26. Bitman J, Wood DL. Changes in milk fat phospholipids during lactation. J Dairy Sci 1990;73(5):1208-16.  27. Christie W, Han X. Lipid Analysis - Isolation, Separation, Identification and Lipidomic Analysis. Edtion ed.: Oily Press, Bridgwater, U.K., 2010. 28. Innis SM. Human milk: maternal dietary lipids and infant development. Proc Nutr Soc 2007;66(3):397-404.  29. Mather IH, Keenan TW. Origin and secretion of milk lipids. J Mammary Gland Biol Neoplasia 1998;3(3):259-73. 30. Argov N, Lemay DG, German JB. Milk Fat Globule structure & function; nanosciece comes to milk production. Trends food sci tech 2008;19(12). 31. Oftedal OT. The evolution of milk secretion and its ancient origins. Animal 2012;6(3):355-68.  140  32. Lopez C, Menard O. Human milk fat globules: polar lipid composition and in situ structural investigations revealing the heterogeneous distribution of proteins and the lateral segregation of sphingomyelin in the biological membrane. Colloids surf., B 2011;83(1):29-41.  33. Bargmann WK, A. . Morphology of lactation; light & electromicroscopic studies on the mammary glands of rats. Z Zellforsch Mikrosk Anat 1959(49):344–88.  34. Keenan TW, Mather IH. Intracellular Origin of Milk Fat Globules and the Nature of the Milk Fat Globule Membrane. Edtion ed. In: Fox PF, McSweeney PLH, eds. Advanced Dairy Chemistry Volume 2 Lipids: Springer US, 2006:137-71. 35. Patton S, Fowkes FM. The role of the plasma membrane in the secretion of milk fat. J Theor Biol 1967;15(3):274-81.  36. Patton S, Trams EG. The presence of plasma membrane enzymes on the surface of bovine milk fat globules. FEBS Letters 1971;14(4):230-2.  37. Mather IH, Keenan TW. Studies on the structure of milk fat globule membrane. The J Membr Biol 1975;21(1-2):65-85. 38. Pike LJ, Han X, Chung KN, Gross RW. Lipid rafts are enriched in arachidonic acid and plasmenylethanolamine and their composition is independent of caveolin-1 expression: a quantitative electrospray ionization/mass spectrometric analysis. Biochem 2002;41(6):2075-88. 39. Le TT, Van de Wiele T, Do TN, et al. Stability of milk fat globule membrane proteins toward human enzymatic gastrointestinal digestion. J Dairy Sci 2012;95(5):2307-18.  40. Garcia C, Lutz NW, Confort-Gouny S, Cozzone PJ, Armand M, Bernard M. Phospholipid fingerprints of milk from different mammalians determined by 31P NMR: towards specific interest in human health. Food Chem 2012;135(3):1777-83.  41. Garcia C, Millet V, Coste TC, et al. French mothers' milk deficient in DHA contains phospholipid species of potential interest for infant development. J Pediatr Gastroenterol Nutr 2011;53(2):206-12.  42. van Beusekom C, Martini IA, Rutgers HM, Boersma ER, Muskiet FA. A carbohydrate-rich diet not only leads to incorporation of medium-chain fatty acids (6:0-14:0) in milk triglycerides but also in each milk-phospholipid subclass. Am J Clin Nutr 1990;52(2):326-34. 43. Bitman J, Wood DL, Mehta NR, Hamosh P, Hamosh M. Comparison of the phospholipid composition of breast milk from mothers of term and preterm infants during lactation. Am J Clin Nutr 1984;40(5):1103-19. 44. Fuchs B. Analytical methods for (oxidized) plasmalogens: Methodological aspects and applications. Free Radic Res 2015;49(5):599-617.  141  45. Folch J, Lees M, Sloane Stanley GH. A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem 1957;226(1):497-509. 46. Honeyman TW, Strohsnitter W, Scheid CR, Schimmel RJ. Phosphatidic acid and phosphatidylinositol labelling in adipose tissue. Relationship to the metabolic effects of insulin and insulin-like agents. Biochem J 1983;212(2):489-98. 47. Braverman NE, Moser AB. Functions of plasmalogen lipids in health and disease. Biochim Biophys Acta 2012;1822(9):1442-52.. 48. Morrison WR. The distribution of phospholipids in some mammalian milks. Lipids 1968;3(1):101-3.  49. Harzer G, Haug M, Dieterich I, Gentner PR. Changing patterns of human milk lipids in the course of the lactation and during the day. Am J Clin Nutr 1983;37(4):612-21. 50. Bracco U, Hidalgo J, Bohren H. Lipid composition of the fat globule membrane of human and bovine milk. J Dairy Sci 1972;55(2):165-72.  51. Giuffrida F, Cruz-Hernandez C, Fluck B, et al. Quantification of phospholipids classes in human milk. Lipids 2013;48(10):1051-8.  52. Russo M, Cichello F, Ragonese C, et al. Profiling and quantifying polar lipids in milk by hydrophilic interaction liquid chromatography coupled with evaporative light-scattering and mass spectrometry detection. Anal Bioanal Chem 2013;405(13):4617-26.  53. Jensen RG. Lipids in human milk. Lipids 1999;34(12):1243-71. 54. Innis SM, King DJ. trans Fatty acids in human milk are inversely associated with concentrations of essential all-cis n-6 and n-3 fatty acids and determine trans, but not n-6 and n-3, fatty acids in plasma lipids of breast-fed infants. Am J Clin Nutr 1999;70(3):383-90. 55. Innis SM. Human milk and formula fatty acids. J Pediatr 1992;120(4 Pt 2):S56-61. 56. Sanders TA, Reddy S. The influence of a vegetarian diet on the fatty acid composition of human milk and the essential fatty acid status of the infant. J Pediatr 1992;120(4 Pt 2):S71-7. 57. Innis SM, Kuhnlein HV. Long-chain n-3 fatty acids in breast milk of Inuit women consuming traditional foods. Early Hum Dev 1988;18(2-3):185-9. 58. Ruan C, Liu X, Man H, et al. Milk composition in women from five different regions of China: the great diversity of milk fatty acids. J Nutr 1995;125(12):2993-8. 59. Gibson RA, Neumann MA, Makrides M. Effect of increasing breast milk docosahexaenoic acid on plasma and erythrocyte phospholipid fatty acids and neural indices of exclusively breast fed infants. Eur J Clin Nutr 1997;51(9):578-84. 142  60. Smit EN, Koopmann M, Boersma ER, Muskiet FA. Effect of supplementation of arachidonic acid (AA) or a combination of AA plus docosahexaenoic acid on breastmilk fatty acid composition. Prostaglandins Leukot Essent Fatty Acids 2000;62(6):335-40.  61. van Goor SA, Dijck-Brouwer DA, Hadders-Algra M, et al. Human milk arachidonic acid and docosahexaenoic acid contents increase following supplementation during pregnancy and lactation. Prostaglandins Leukot Essent Fatty Acids 2009;80(1):65-9. 62. Sherry CL, Oliver JS, Marriage BJ. Docosahexaenoic acid supplementation in lactating women increases breast milk and plasma docosahexaenoic acid concentrations and alters infant omega 6:3 fatty acid ratio. Prostaglandins Leukot Essent Fatty Acids 2015;95:63-9.  63. Morrison WR, Smith LM. Fatty acid composition of milk phospholipids. II. Sheep, Indian buffalo and human milks. Lipids 1967;2(2):178-82.  64. Cherian G, Sim JS. Changes in the breast milk fatty acids and plasma lipids of nursing mothers following consumption of n-3 polyunsaturated fatty acid enriched eggs. Nutrition 1996;12(1):8-12.  65. Wang L, Shimizu Y, Kaneko S, et al. Comparison of the fatty acid composition of total lipids and phospholipids in breast milk from Japanese women. Pediatrics International 2000;42(1):14-20. 66. Hundrieser KE, Clark RM, Brown PB. Distribution of trans-octadecenoic acid in the major glycerolipids of human milk. J Pediatr Gastroenterol Nutr 1983;2(4):635-9. 67. Conde C, Casado de Frias E, Moro M. Essential fatty acids in phosphoglycerides of human milk. Milk from mothers of term and preterm infants. Acta paediatrica Scandinavica 1983;72(2):255-7. 68. Mather IH. A review and proposed nomenclature for major proteins of the milk-fat globule membrane. J Dairy Sci 2000;83(2):203-47.  69. Cavaletto M, Giuffrida MG, Conti A. Milk fat globule membrane components--a proteomic approach. Advances in experimental medicine and biology 2008;606:129-41.  70. Cavaletto M, Giuffrida MG, Conti A. The proteomic approach to analysis of human milk fat globule membrane. Clinica chimica acta 2004;347(1-2):41-8.  71. Liao Y, Alvarado R, Phinney B, Lonnerdal B. Proteomic characterization of human milk fat globule membrane proteins during a 12 month lactation period. J Proteome Res 2011;10(8):3530-41.  72. Lonnerdal B. Infant formula and infant nutrition: bioactive proteins of human milk and implications for composition of infant formulas. Am J Clin Nutr 2014;99(3):712s-7s.  73. Innis SM. Dietary triacylglycerol structure and its role in infant nutrition. Adv Nutr 2011;2(3):275-83.  143  74. Neville MC, Picciano MF. Regulation of milk lipid secretion and composition. Annu Rev Nutr 1997;17:159-83.  75. Cooper SM, Grigor MR. Fatty acid specificities of microsomal acyltransferases esterifying positions-1 and -2 of acylglycerols in mammary glands from lactating rats. Biochem J 1980;187(2):289-95. 76. Patton S, McCarthy RD, Plantz PE, Lee RF. Phospholipid incorporation into plasma membrane of lactating mammary cell with special reference to sphingomyelin. Nature: New biology 1973;241(112):241-2. 77. Easter DJ, Patton S, McCarthy RD. Metabolism of phospholipid in mammary gland. I. The supply of phospholipid for milk synthesis in the rat and goat. Lipids 1971;6(11):844-9. 78. Gibellini F, Smith TK. The Kennedy pathway--De novo synthesis of phosphatidylethanolamine and phosphatidylcholine. IUBMB life 2010;62(6):414-28.  79. McManaman JL. Lipid transport in the lactating mammary gland. Mammary Gland Biol Neoplasia 2014;19(1):35-42.  80. Kinsella JE. Preferential labeling of phosphatidylcholine during phospholipid synthesis by bovine mammary tissue. Lipids 1973;8(7):393-400. 81. Yang EK, Blusztajn JK, Pomfret EA, Zeisel SH. Rat and human mammary tissue can synthesize choline moiety via the methylation of phosphatidylethanolamine. The Biochemical journal 1988;256(3):821-8. 82. Brites P, Waterham HR, Wanders RJ. Functions and biosynthesis of plasmalogens in health and disease. Biochim Biophys Acta 2004;1636(2-3):219-31.  83. EFSA Panel on Dietetic Products NaaN. Scientific opinion on the tollerable upper intake level of eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA) and docosapentaenoic acid (DPA). Parma, Italy: EFSA J, 2012. 84. ANSES (Agence nationale de sécurité sanitaire de l'alimentation dleedt. Opinion of the French Food Safety Agency on the update of French population reference intakes (ANCs) for fatty acids. 2010. 85. Burr GO, Burr MM. A new deficiency disease produced by the rigid exclusion of fat from the diet. J biol chem 1929;82:345-67. 86. Burr GO, Burr MM. The nature and role of the fatty acids essential in nutrition. J biol chem 1930;86:587-621. 87. Turpeinen O. Further studies on the unsaturated fatty acids essential in nutrition. J nutr  1938;15(4):351-66. 88. Hume EM, Nunn LC, Smedley-Maclean I, Smith HH. Fat-deficiency disease of rats. The relative curative potencies of methyl linoleate and methyl arachidonate with a note on the 144  action of the methyl esters of fatty acids from cod liver oil. Biochem J 1940;34(6):879-83. 89. Hansen AE, Stewart RA, Hughes G, Soderhjelm L. The relation of linoleic acid to infant feeding. Acta paediatrica Supplementum 1962;137:1-41. 90. Collins FD, Sinclair AJ, Royle JP, Coats DA, Maynard AT, Leonard RF. Plasma lipids in human linoleic acid deficiency. Nutri metab 1971;13(3):150-67. 91. Holman RT. The slow discovery of the importance of omega 3 essential fatty acids in human health. J Nutri 1998;128(2 Suppl):427S-33S. 92. Lamptey MS, Walker BL. A possible essential role for dietary linolenic acid in the development of the young rat. J Nutr 1976;106(1):86-93. 93. Anderson RE, Maude MB. Lipids of ocular tissues. 8. The effects of essential fatty acid deficiency on the phospholipids of the photoreceptor membranes of rat retina. Arc biochem biophys 1972;151(1):270-6. 94. Futterman S, Downer JL, Hendrickson A. Effect of essential fatty acid deficiency on the fatty acid composition, morphology, and electroretinographic response of the retina. Invest Ophthalmol 1971;10(2):151-6. 95. Neuringer M, Connor WE, Van Petten C, Barstad L. Dietary omega-3 fatty acid deficiency and visual loss in infant rhesus monkeys. J Clin Invest 1984;73(1):272-6. 96. Neuringer M, Connor WE, Lin DS, Barstad L, Luck S. Biochemical and functional effects of prenatal and postnatal omega 3 fatty acid deficiency on retina and brain in rhesus monkeys. Proceedings of the National Academy of Sciences of the United States of America 1986;83(11):4021-5. 97. Reisbick S, Neuringer M, Gohl E, Wald R, Anderson GJ. Visual attention in infant monkeys: effects of dietary fatty acids and age. Developmental psychology 1997;33(3):387-95. 98. Holman RT, Johnson SB, Hatch TF. A case of human linolenic acid deficiency involving neurological abnormalities. Am J Clin Nutr 1982;35(3):617-23. 99. Burri L, Johnsen L. Krill Products: An Overview of Animal Studies. Nutrients 2015;7(5):3300-21.  100. Rocquelin G, Guenot L, Astorg PO, David M. Phospholipid content and fatty acid composition of human heart. Lipids 1989;24(9):775-80. 101. Diagne A, Fauvel J, Record M, Chap H, Douste-Blazy L. Studies on ether phospholipids: II. Comparative composition of various tissues from human, rat and guinea pig. Biochimica et Biophysica Acta (BBA) - Lip Lip Metab 1984;793(2):221-31.  145  102. Lessig J, Fuchs B. Plasmalogens in biological systems: their role in oxidative processes in biological membranes, their contribution to pathological processes and aging and plasmalogen analysis. Curr Med Chem 2009;16(16):2021-41. 103. Svennerholm L. Distribution and fatty acid composition of phosphoglycerides in normal human brain. J Lipid Res 1968;9(5):570-9. 104. Letondor A, Buaud B, Vaysse C, et al. Erythrocyte DHA level as a biomarker of DHA status in specific brain regions of n-3 long-chain PUFA-supplemented aged rats. Br J Nutr 2014;112(11):1805-18.  105. Benediktsdottir VE, Gudbjarnason S. Reversible alterations in fatty acid composition of heart muscle membrane phospholipids induced by epinephrine in rats fed different fats. J Lipid Res 1988;29(6):765-72. 106. Wymann MP, Schneiter R. Lipid signalling in disease. Nature Rev Molec Cell Biol 2008;9(2):162-76.  107. Yaqoob P. Fatty acids as gatekeepers of immune cell regulation. Trends Immunol 2003;24(12):639-45. 108. Sampath H, Ntambi JM. Polyunsaturated fatty acid regulation of gene expression. Nutrition reviews 2004;62(9):333-9. 109. Hou TY, McMurray DN, Chapkin RS. Omega-3 fatty acids, lipid rafts, and T cell signaling. European journal of pharmacology 2015. 110. Jump DB. N-3 polyunsaturated fatty acid regulation of hepatic gene transcription. Current opinion in lipidology 2008;19(3):242-7.  111. Wietrzych-Schindler M, Szyszka-Niagolov M, Ohta K, et al. Retinoid x receptor gamma is implicated in docosahexaenoic acid modulation of despair behaviors and working memory in mice. Biological psychiatry 2011;69(8):788-94.  112. Cao D, Kevala K, Kim J, et al. Docosahexaenoic acid promotes hippocampal neuronal development and synaptic function. J Neurochem 2009;111(2):510-21.  113. Stiles J, Jernigan TL. The basics of brain development. Neuropsy Review 2010;20(4):327-48.  114. Jurkowitz-Alexander M, Ebata H, Mills JS, Murphy EJ, Horrocks LA. Solubilization, purification and characterization of lysoplasmalogen alkenylhydrolase (lysoplasmalogenase) from rat liver microsomes. Biochim Biophys Acta 1989;1002(2):203-12. 115. Nishimukai M, Wakisaka T, Hara H. Ingestion of plasmalogen markedly increased plasmalogen levels of blood plasma in rats. Lipids 2003;38(12):1227-35. 116. Jurkowitz MS, Horrocks LA, Litsky ML. Identification and characterization of alkenyl hydrolase (lysoplasmalogenase) in microsomes and identification of a plasmalogen-active 146  phospholipase A2 in cytosol of small intestinal epithelium. Biochim Biophys Acta 1999;1437(2):142-56. 117. Nishimukai M, Yamashita M, Watanabe Y, et al. Lymphatic absorption of choline plasmalogen is much higher than that of ethanolamine plasmalogen in rats. Eur J Nutr 2011;50(6):427-36.  118. Wu LC, Pfeiffer DR, Calhoon EA, et al. Purification, identification, and cloning of lysoplasmalogenase, the enzyme that catalyzes hydrolysis of the vinyl ether bond of lysoplasmalogen. J biol chem 2011;286(28):24916-30.  119. Bergstrom S, Blomstrand R. The intestinal absorption and metabolism of chimyl alcohol in the rat. Acta physiologica Scandinavica 1956;38(2):166-72.  120. Hara H, Wakisaka T, Aoyama Y. Lymphatic absorption of plasmalogen in rats. Br J Nutr 2003;90(1):29-32. 121. Paltauf F, Holasek A. Enzymatic synthesis of plasmalogens. Characterization of the 1-O-alkyl-2-acyl-8n-glycero-3-phosphorylethanolamine desaturase from mucosa of hamster small intestine. J Bio Chem 1973;248(5):1609-15. 122. O'Brien JS, Sampson EL. Fatty acid and fatty aldehyde composition of the major brain lipids in normal human gray matter, white matter, and myelin. J Lipid Res 1965;6(4):545-51. 123. O'Brien JS, Sampson EL. Lipid composition of the normal human brain: gray matter, white matter, and myelin. J Lipid Res 1965;6(4):537-44. 124. Levitt P. Structural and functional maturation of the developing primate brain. J Pediatr 2003;143(4 Suppl):S35-45. 125. Tau GZ, Peterson BS. Normal development of brain circuits. Neuropsychopharmacology 2010;35(1):147-68.  126. Smart JL, Dobbing J. Vulnerability of developing brain. II. Effects of early nutritional deprivation on reflex ontogeny and development of behaviour in the rat. Brain Res 1971;28(1):85-95.  127. Kiss P, Vadasz G, Kiss-Illes B, et al. Environmental enrichment decreases asphyxia-induced neurobehavioral developmental delay in neonatal rats. International journal of molecular sciences 2013;14(11):22258-73.  128. Wells MA, Dittmer JC. A comprehensive study of the postnatal changes in the concentration of the lipids of developing rat brain. Biochem 1967;6(10):3169-75. 129. Dekaban AS. Changes in brain weights during the span of human life: relation of brain weights to body heights and body weights. Ann neurol 1978;4(4):345-56.  130. Sinclair AJ, Crawford MA. The accumulation of arachidonate and docosahexaenoate in the developing rat brain. J Neurochem 1972;19(7):1753-8. 147  131. Martinez M. Tissue levels of polyunsaturated fatty acids during early human development. J Pediatr 1992;120(4 Pt 2):S129-38. 132. Green P, Glozman S, Kamensky B, Yavin E. Developmental changes in rat brain membrane lipids and fatty acids. The preferential prenatal accumulation of docosahexaenoic acid. J Lipid Res 1999;40(5):960-6. 133. Yusuf HK, Dickerson JW. The effect of growth and development on the phospholipids of the human brain. J Neurochem 1977;28(4):783-8. 134. Svennerholm L, Vanier MT. The distribution of lipids in the human nervous system. II. Lipid composition of human fetal and infant brain. Brain Res 1972;47(2):457-68. 135. Martinez M, Ballabriga A. A chemical study on the development of the human forebrain and cerebellum during the brain 'growth spurt' period. I. Gangliosides and plasmalogens. Brain research 1978;159(2):351-62. 136. Braverman NE, D'Agostino MD, Maclean GE. Peroxisome biogenesis disorders: Biological, clinical and pathophysiological perspectives. Developmental disabilities research reviews 2013;17(3):187-96. 137. Bakovic M, Fullerton MD, Michel V. Metabolic and molecular aspects of ethanolamine phospholipid biosynthesis: the role of CTP:phosphoethanolamine cytidylyltransferase (Pcyt2). Biochem Cell Biol 2007;85(3):283-300.  138. Nagan N, Zoeller RA. Plasmalogens: biosynthesis and functions. Prog Lipid Res 2001;40(3):199-229. 139. Glaser PE, Gross RW. Plasmenylethanolamine facilitates rapid membrane fusion: a stopped-flow kinetic investigation correlating the propensity of a major plasma membrane constituent to adopt an HII phase with its ability to promote membrane fusion. Biochem 1994;33(19):5805-12. 140. Gremo F, De Medio GE, Trovarelli G, Dessi S, Porru S. Mature and immature synaptosomal membranes have a different lipid composition. Neurochem Res 1985;10(1):133-44. 141. Sollner T, Whiteheart SW, Brunner M, et al. SNAP receptors implicated in vesicle targeting and fusion. Nature 1993;362(6418):318-24.  142. Pollard HB, Rojas E, Burns AL. Synexin (annexin VII) and membrane fusion during the process of exocytotic secretion. Progress brain Res 1992;92:247-55. 143. Farooqui AA, Horrocks LA. Plasmalogens, phospholipase A2, and docosahexaenoic acid turnover in brain tissue. J molec neurosc : MN 2001;16(2-3):263-72; discussion 79-84. 144. Farooqui AA, Horrocks LA. Plasmalogens: workhorse lipids of membranes in normal and injured neurons and glia. Neurosci 2001;7(3):232-45. 148  145. Dorninger F, Brodde A, Braverman NE, et al. Homeostasis of phospholipids - The level of phosphatidylethanolamine tightly adapts to changes in ethanolamine plasmalogens. Biochim Biophys Acta 2015;1851(2):117-28.  146. Martinez M, Mougan I. Fatty acid composition of brain glycerophospholipids in peroxisomal disorders. Lipids 1999;34(7):733-40. 147. Martinez M. Abnormal profiles of polyunsaturated fatty acids in the brain, liver, kidney and retina of patients with peroxisomal disorders. Brain Res 1992;583(1-2):171-82. 148. Mandel H, Sharf R, Berant M, Wanders RJ, Vreken P, Aviram M. Plasmalogen phospholipids are involved in HDL-mediated cholesterol efflux: insights from investigations with plasmalogen-deficient cells. Biochem biophys Res 1998;250(2):369-73.  149. Mankidy R, Ahiahonu PW, Ma H, et al. Membrane plasmalogen composition and cellular cholesterol regulation: a structure activity study. Lipids Health Dis 2010;9:62.  150. Dietschy JM, Turley SD. Cholesterol metabolism in the brain. Current opinion in lipidology 2001;12(2):105-12. 151. Mangold HK, Weber N. Biosynthesis and biotransformation of ether lipids. Lipids 1987;22(11):789-99. 152. Reiss D, Beyer K, Engelmann B. Delayed oxidative degradation of polyunsaturated diacyl phospholipids in the presence of plasmalogen phospholipids in vitro. Biochem J 1997;323 ( Pt 3):807-14. 153. Sindelar PJ, Guan Z, Dallner G, Ernster L. The protective role of plasmalogens in iron-induced lipid peroxidation. Free Radic Bio Med 1999;26(3-4):318-24. 154. Kuczynski B, Reo NV. Evidence that plasmalogen is protective against oxidative stress in the rat brain. Neurochem Res 2006;31(5):639-56. 9061-7. 155. Luoma AM, Kuo F, Cakici O, et al. Plasmalogen phospholipids protect internodal myelin from oxidative damage. Free radical biology & medicine 2015;84:296-310.  156. Zeisel SH. The fetal origins of memory: the role of dietary choline in optimal brain development. J Pediatr 2006;149(5 Suppl):S131-6.  157. Wang B. Sialic acid is an essential nutrient for brain development and cognition. Annu Rev Nutr 2009;29:177-222 158. Palmano K, Rowan A, Guillermo R, Guan J, McJarrow P. The Role of Gangliosides in Neurodevelopment. Nutrients 2015;7(5):3891-913.  159. Guirland C, Zheng JQ. Membrane lipid rafts and their role in axon guidance. Adv exp med biol 2007;621:144-55.  149  160. Xu XZ, Zhu TC. Effect of ganglioside in repairing the neurological function of chinese children with cerebral palsy: Analysis of the curative efficacy in 2230 cases. Chinese J Clin Rehab 2005;9:122-3. 161. Oshida K, Shimizu T, Takase M, Tamura Y, Shimizu T, Yamashiro Y. Effects of dietary sphingomyelin on central nervous system myelination in developing rats. Pediatr Res 2003;53(4):589-93.  162. Jurevics H, Hostettler J, Muse ED, et al. Cerebroside synthesis as a measure of the rate of remyelination following cuprizone-induced demyelination in brain. J Neurochem 2001;77(4):1067-76. 163. Wang B, Yu B, Karim M, et al. Dietary sialic acid supplementation improves learning and memory in piglets. Am J Clin Nutr 2007;85(2):561-9. 164. Carlson SE, House SG. Oral and intraperitoneal administration of N-acetylneuraminic acid: effect on rat cerebral and cerebellar N-acetylneuraminic acid. J Nutr 1986;116(5):881-6. 165. Park EJ, Suh M, Ramanujam K, Steiner K, Begg D, Clandinin MT. Diet-induced changes in membrane gangliosides in rat intestinal mucosa, plasma and brain. J Pediatr Gastroenterol Nutr 2005;40(4):487-95. 166. Wainwright PE, Lomanowska AM, McCutcheon D, Park EJ, Clandinin MT, Ramanujam KS. Postnatal dietary supplementation with either gangliosides or choline: effects on spatial short-term memory in artificially-reared rats. Nutri Neurosci 2007;10(1-2):67-77.  167. Meck WH, Smith RA, Williams CL. Organizational changes in cholinergic activity and enhanced visuospatial memory as a function of choline administered prenatally or postnatally or both. Behav Neurosci 1989;103(6):1234-41. 168. Carlson SE. N-acetylneuraminic acid concentrations in human milk oligosaccharides and glycoproteins during lactation. Am J Clin Nutr 1985;41(4):720-6. 169. Wang B, Brand-Miller J, McVeagh P, Petocz P. Concentration and distribution of sialic acid in human milk and infant formulas. Am J Clin Nutr 2001;74(4):510-5. 170. Meck WH, Williams CL. Characterization of the facilitative effects of perinatal choline supplementation on timing and temporal memory. Neuroreport 1997;8(13):2831-5. 171. Tees RC, Mohammadi E. The effects of neonatal choline dietary supplementation on adult spatial and configural learning and memory in rats. Developmental psychobiology 1999;35(3):226-40. 172. Burling H, Graverholt G. Milk – A new source for bioactive phospholipids for use in food formulations. Lipid Tech 2008;20(10):229-31. 173. Liu H, Radlowski EC, Conrad MS, Li Y, Dilger RN, Johnson RW. Early supplementation of phospholipids and gangliosides affects brain and cognitive development in neonatal piglets. J Nutr 2014;144(12):1903-9.  150  174. Vickers MH, Guan J, Gustavsson M, et al. Supplementation with a mixture of complex lipids derived from milk to growing rats results in improvements in parameters related to growth and cognition. Nutrition research (New York, NY) 2009;29(6):426-35.  175. Guillermo RB, Yang P, Vickers MH, McJarrow P, Guan J. Supplementation with complex milk lipids during brain development promotes neuroplasticity without altering myelination or vascular density. Food Nutr Res 2015;59:25765 176. Gurnida DA, Rowan AM, Idjradinata P, Muchtadi D, Sekarwana N. Association of complex lipids containing gangliosides with cognitive development of 6-month-old infants. Early Hum Dev 2012;88(8):595-601.  177. Billeaud C, Puccio G, Saliba E, et al. Safety and tolerance evaluation of milk fat globule membrane-enriched infant formulas: a randomized controlled multicenter non-inferiority trial in healthy term infants. Clinical medicine insights Pediatrics 2014;8:51-60.  178. Timby N, Domellof E, Hernell O, Lonnerdal B, Domellof M. Neurodevelopment, nutrition, and growth until 12 mo of age in infants fed a low-energy, low-protein formula supplemented with bovine milk fat globule membranes: a randomized controlled trial. Am J Clin Nutr 2014;99(4):860-8.  179. Timby N, Lonnerdal B, Hernell O, Domellof M. Cardiovascular risk markers until 12 mo of age in infants fed a formula supplemented with bovine milk fat globule membranes. Pediatr Res 2014;76(4):394-400.  180. Timby N, Hernell O, Vaarala O, Melin M, Lonnerdal B, Domellof M. Infections in infants fed formula supplemented with bovine milk fat globule membranes. J Pediatr Gastroenterol Nutr 2015;60(3):384-9.  181. Zavaleta N, Kvistgaard AS, Graverholt G, et al. Efficacy of an MFGM-enriched complementary food in diarrhea, anemia, and micronutrient status in infants. J Pediatr Gastroenterol Nutr 2011;53(5):561-8.  182. Veereman-Wauters G, Staelens S, Rombaut R, et al. Milk fat globule membrane (INPULSE) enriched formula milk decreases febrile episodes and may improve behavioral regulation in young children. Nutrition 2012;28(7-8):749-52. 183. Sheard NF, Zeisel SH. Choline: an essential dietary nutrient? Nutrition 1989;5(1):1-5. 184. Zeisel SH, Da Costa KA, Franklin PD, et al. Choline, an essential nutrient for humans. FASEB 1991;5(7):2093-8. 185. Ilcol YO, Ozbek R, Hamurtekin E, Ulus IH. Choline status in newborns, infants, children, breast-feeding women, breast-fed infants and human breast milk. J Nutr Biochem 2005;16(8):489-99.  186. Fischer LM, da Costa KA, Galanko J, et al. Choline intake and genetic polymorphisms influence choline metabolite concentrations in human breast milk and plasma. Am J Clin Nutr 2010;92(2):336-46.  151  187. Zeisel SH, Char D, Sheard NF. Choline, phosphatidylcholine and sphingomyelin in human and bovine milk and infant formulas. J Nutr 1986;116(1):50-8. 188. Holmes-McNary MQ, Cheng WL, Mar MH, Fussell S, Zeisel SH. Choline and choline esters in human and rat milk and in infant formulas. Am J Clin Nutr 1996;64(4):572-6. 189. Holmes HC, Snodgrass GJ, Iles RA. Changes in the choline content of human breast milk in the first 3 weeks after birth. Eur J Pediatr  2000;159(3):198-204. 190. Institute of Medicine (US) Standing Committee on the Scientific Evaluation of Dietary Reference Intakes and its Panel on Folate OBV, and Choline. Dietary Reference Intakes for Thiamin, Riboflavin, Niacin, Vitamin B6, Folate, Vitamin B12, Pantothenic Acid, Biotin and Choline. Washington, DC., 1998. 191. Zeisel SH, Blusztajn JK. Choline and human nutrition. Ann Res Nutr 1994;14:269-96.  192. Ueland P. Choline and betaine in health and disease. J Inherit Metab Dis 2011;34(1):3-15.  193. Zeisel SH. Diet-gene interactions underlie metabolic individuality and influence brain development: implications for clinical practice derived from studies on choline metabolism. Ann Nutr Metab 2012;60 Suppl 3:19-25.  194. Fisher MC, Zeisel SH, Mar MH, Sadler TW. Perturbations in choline metabolism cause neural tube defects in mouse embryos in vitro. FASEB 2002;16(6):619-21. 195. Fisher MC, Zeisel SH, Mar MH, Sadler TW. Inhibitors of choline uptake and metabolism cause developmental abnormalities in neurulating mouse embryos. Teratology 2001;64(2):114-22.  196. Mehedint MG, Craciunescu CN, Zeisel SH. Maternal dietary choline deficiency alters angiogenesis in fetal mouse hippocampus. Proceedings of the National Academy of Sciences of the United States of America 2010;107(29):12834-9.  197. Craciunescu CN, Albright CD, Mar MH, Song J, Zeisel SH. Choline availability during embryonic development alters progenitor cell mitosis in developing mouse hippocampus. J Nutr 2003;133(11):3614-8. 198. Albright CD, Tsai AY, Friedrich CB, Mar MH, Zeisel SH. Choline availability alters embryonic development of the hippocampus and septum in the rat. Brain Res Develop brain Res 1999;113(1-2):13-20. 199. Albright CD, Friedrich CB, Brown EC, Mar MH, Zeisel SH. Maternal dietary choline availability alters mitosis, apoptosis and the localization of TOAD-64 protein in the developing fetal rat septum. Brain research Developmental brain research 1999;115(2):123-9. 200. Meck WH, Smith RA, Williams CL. Pre- and postnatal choline supplementation produces long-term facilitation of spatial memory. Develop Psychobiol 1988;21(4):339-53.  152  201. Meck WH, Williams CL. Choline supplementation during prenatal development reduces proactive interference in spatial memory. Develop Brain Res 1999;118(1-2):51-9. 202. Friesen RW, Novak EM, Hasman D, Innis SM. Relationship of dimethylglycine, choline, and betaine with oxoproline in plasma of pregnant women and their newborn infants. J Nutr 2007;137(12):2641-6. 203. Wu BT, Dyer RA, King DJ, Richardson KJ, Innis SM. Early second trimester maternal plasma choline and betaine are related to measures of early cognitive development in term infants. PloS one 2012;7(8):e43448. 204. Meck WH, Williams CL, Cermak JM, Blusztajn JK. Developmental periods of choline sensitivity provide an ontogenetic mechanism for regulating memory capacity and age-related dementia. Front Integr Neurosci 2007;1:7 205. Lauder JM, Schambra UB. Morphogenetic roles of acetylcholine. Environ Health Persp 1999;107(Suppl 1):65-9. 206. Hasselmo ME. The role of acetylcholine in learning and memory. Curr Opin Neurobiol 2006;16(6):710-5.  207. Lockman PR, Allen DD. The transport of choline. Drug Dev Ind Pharm 2002;28(7):749-71.  208. Wecker L. The synthesis and release of acetylcholine by depolarized hippocampal slices is increased by increased choline available in vitro prior to stimulation. J Neurochem 1991;57(4):1119-27. 209. Ross BM, Moszczynska A, Blusztajn JK, Sherwin A, Lozano A, Kish SJ. Phospholipid biosynthetic enzymes in human brain. Lipids 1997;32(4):351-8. 210. Numata S, Ye T, Hyde TM, et al. DNA methylation signatures in development and aging of the human prefrontal cortex. Am J Hum Genet 2012;90(2):260-72.  211. Kennedy EP, Weiss SB. The function of cytidine coenzymes in the biosynthesis of phospholipides. J Biol Chem 1956;222(1):193-214. 212. Vance JE. Phospholipid synthesis in a membrane fraction associated with mitochondria. J Biol Chem 1990;265(13):7248-56. 213. DeLong CJ, Shen YJ, Thomas MJ, Cui Z. Molecular distinction of phosphatidylcholine synthesis between the CDP-choline pathway and phosphatidylethanolamine methylation pathway. J Biol Chem 1999;274(42):29683-8. 214. Sundler R, Akesson B. Regulation of phospholipid biosynthesis in isolated rat hepatocytes. Effect of different substrates. J Biol Chem 1975;250(9):3359-67. 215. Yao ZM, Vance DE. The active synthesis of phosphatidylcholine is required for very low density lipoprotein secretion from rat hepatocytes. J Biol Chem 1988;263(6):2998-3004. 153  216. Buchman AL, Moukarzel A, Jenden DJ, Roch M, Rice K, Ament ME. Low plasma free choline is prevalent in patients receiving long term parenteral nutrition and is associated with hepatic aminotransferase abnormalities. Clinic Nutri 1993;12(1):33-7. 217. Buchman AL, Dubin MD, Moukarzel AA, et al. Choline deficiency: a cause of hepatic steatosis during parenteral nutrition that can be reversed with intravenous choline supplementation. Hepatology 1995;22(5):1399-403. 218. Buchman AL, Dubin M, Jenden D, et al. Lecithin increases plasma free choline and decreases hepatic steatosis in long-term total parenteral nutrition patients. Gastroenterology 1992;102(4 Pt 1):1363-70. 219. Buchman AL, Ament ME, Sohel M, et al. Choline deficiency causes reversible hepatic abnormalities in patients receiving parenteral nutrition: proof of a human choline requirement: a placebo-controlled trial. JPEN 2001;25(5):260-8. 220. Vance JE. Secretion of VLDL, but not HDL, by rat hepatocytes is inhibited by the ethanolamine analogue N-monomethylethanolamine. J lipid res 1991;32(12):1971-82. 221. Couture P, Hulbert AJ. Membrane fatty acid composition of tissues is related to body mass of mammals. J Membr Biol 1995;148(1):27-39. 222. Hulbert AJ, Rana T, Couture P. The acyl composition of mammalian phospholipids: an allometric analysis. Comparative biochemistry and physiology Part B, Biochemistry & molecular biology 2002;132(3):515-27. 223. Mansson HL. Fatty acids in bovine milk fat. J Food Nutr Res 2008;52.  224. Sokol E, Ulven T, Færgeman NJ, Ejsing CS. Comprehensive and quantitative profiling of lipid species in human milk, cow milk and a phospholipid-enriched milk formula by GC and MS/MS(ALL). Eur J Lipid Sci Tech 2015;117(6):751-9.  225. Innis SM. Omega-3 Fatty acids and neural development to 2 years of age: do we know enough for dietary recommendations? J Pediatr Gastroenterol Nutr 2009;48 Suppl 1:S16-24.  226. Yuhas R, Pramuk K, Lien EL. Human milk fatty acid composition from nine countries varies most in DHA. Lipids 2006;41(9):851-8. 227. Fidler N, Sauerwald T, Pohl A, Demmelmair H, Koletzko B. Docosahexaenoic acid transfer into human milk after dietary supplementation: a randomized clinical trial. J Lipid Res 2000;41(9):1376-83. 228. Gottrand F. Long-chain polyunsaturated fatty acids influence the immune system of infants. J Nutr 2008;138(9):1807s-12s. 229. Antonny B, Vanni S, Shindou H, Ferreira T. From zero to six double bonds: phospholipid unsaturation and organelle function. Trends Cell Biol 2015;25(7):427-36.  154  230. Crawford MA, Casperd NM, Sinclair AJ. The long chain metabolites of linoleic and linolenic acids in liver and brain in herbivores and carnivores. Comparative Biochemistry and Physiology Part B: Comparative Biochemistry 1976;54(3):395-401.  231. Crawford MA, Sinclair AJ. Nutritional influences in the evolution of mammalian brain. In: lipids, malnutrition & the developing brain. Ciba Foundation symposium 1971:267-92. 232. Lecomte M, Paget C, Ruggiero D, Wiernsperger N, Lagarde M. Docosahexaenoic acid is a major n-3 polyunsaturated fatty acid in bovine retinal microvessels. J Neurochem 1996;66(5):2160-7. 233. Bisogno T, Delton-Vandenbroucke I, Milone A, Lagarde M, Di Marzo V. Biosynthesis and inactivation of N-arachidonoylethanolamine (anandamide) and N-docosahexaenoylethanolamine in bovine retina. Arc Biochem Biophys 1999;370(2):300-7.  234. Friesen RW, Innis SM. Linoleic acid is associated with lower long-chain n-6 and n-3 fatty acids in red blood cell lipids of Canadian pregnant women. Am J Clin Nutr 2010;91(1):23-31.  235. Novak EM, Innis SM. Dietary long chain n-3 fatty acids are more closely associated with protein than energy intakes from fat. Prostaglandins Leukot Essent Fatty Acids 2012;86(3):107-12.  236. Novak EM, Innis SM. Impact of maternal dietary n-3 and n-6 fatty acids on milk medium-chain fatty acids and the implications for neonatal liver metabolism. Am J Physiol Endocrinol Metab 2011;301(5):E807-17.  237. Kaluzny MA, Duncan LA, Merritt MV, Epps DE. Rapid separation of lipid classes in high yield and purity using bonded phase columns. J Lipid Res 1985;26(1):135-40. 238. Mawatari S, Okuma Y, Fujino T. Separation of intact plasmalogens and all other phospholipids by a single run of high-performance liquid chromatography. Anal Biochem 2007;370(1):54-9.  239. Conquer JA, Tierney MC, Zecevic J, Bettger WJ, Fisher RH. Fatty acid analysis of blood plasma of patients with Alzheimer's disease, other types of dementia, and cognitive impairment. Lipids 2000;35(12):1305-12. 240. Innis SM, Davidson AG. Cystic fibrosis and nutrition: linking phospholipids and essential fatty acids with thiol metabolism. Annu Rev Nutr 2008;28:55-72.  241. Dougherty RM, Galli C, Ferro-Luzzi A, Iacono JM. Lipid and phospholipid fatty acid composition of plasma, red blood cells, and platelets and how they are affected by dietary lipids: a study of normal subjects from Italy, Finland, and the USA. Am J Clin Nutr 1987;45(2):443-55. 242. Makrides M, Neumann MA, Gibson RA. Effect of maternal docosahexaenoic acid (DHA) supplementation on breast milk composition. Eur J Clin Nutr 1996;50(6):352-7. 155  243. Brenna JT, Varamini B, Jensen RG, Diersen-Schade DA, Boettcher JA, Arterburn LM. Docosahexaenoic and arachidonic acid concentrations in human breast milk worldwide. Am J Clin Nutr 2007;85(6):1457-64. 244. Innis SM. Essential fatty acid metabolism during early development. Edtion ed. In: Burrn DG, ed. Biology of Metabolism in Growing Animals. Amsterdam: Elsevier Science, B.V., 2005:235–74. 245. Liu JJ, Green P, John Mann J, Rapoport SI, Sublette ME. Pathways of polyunsaturated fatty acid utilization: Implications for brain function in neuropsychiatric health and disease. Brain research 2015;1597:220-46.  246. Holčapek M, Cífková E, Červená B, Lísa M, Vostálová J, Galuszka J. Determination of nonpolar and polar lipid classes in human plasma, erythrocytes and plasma lipoprotein fractions using ultrahigh-performance liquid chromatography-mass spectrometry. J Chromatogr A 2015;1377:85-91.  247. Small DM. The effects of glyceride structure on absorption and metabolism. Annu Rev Nutr 1991;11:413-34. 248. Thies F, Pillon C, Moliere P, Lagarde M, Lecerf J. Preferential incorporation of sn-2 lysoPC DHA over unesterified DHA in the young rat brain. Am J Phys 1994;267(5 Pt 2):R1273-9. 249. Liu L, Bartke N, Van Daele H, et al. Higher efficacy of dietary DHA provided as a phospholipid than as a triglyceride for brain DHA accretion in neonatal piglets. J Lipid Res 2014;55(3):531-9.  250. Wijendran V, Huang MC, Diau GY, Boehm G, Nathanielsz PW, Brenna JT. Efficacy of dietary arachidonic acid provided as triglyceride or phospholipid as substrates for brain arachidonic acid accretion in baboon neonates. Pediatr Res 2002;51(3):265-72.  251. Garcia-Calatayud S, Redondo C, Martin E, Ruiz JI, Garcia-Fuentes M, Sanjurjo P. Brain Docosahexaenoic Acid Status and Learning in Young Rats Submitted to Dietary Long-Chain Polyunsaturated Fatty Acid Deficiency and Supplementation Limited to Lactation. Pediatr Res 2005;57(5 Part 1):719-23. 252. Haubner LY, Stockard JE, Saste MD, et al. Maternal dietary docosahexanoic acid content affects the rat pup auditory system. Brain Res Bulletin 2002;58(1):1-5. 253. Saste MD, Carver JD, Stockard JE, Benford VJ, Chen LT, Phelps CP. Maternal diet fatty acid composition affects neurodevelopment in rat pups. J Nutr 1998;128(4):740-3. 254. Chen H-F, Su H-M. Exposure to a maternal n-3 fatty acid-deficient diet during brain development provokes excessive hypothalamic–pituitary–adrenal axis responses to stress and behavioral indices of depression and anxiety in male rat offspring later in life. J Nutri Biochem 2013;24(1):70-80.  156  255. Lomanowska AM, Melo AI. Deconstructing the function of maternal stimulation in offspring development: Insights from the artificial rearing model in rats. Hormones Behavior 2015.  256. Koldovsky O, Bedrick A, Thornburg W. Processing of hormones and hormone-like substances from milk in the gastrointestinal tract of suckling rats. Endocrinologia experimentalis 1986;20(2-3):119-30. 257. Thulesen J, Raaberg L, Nexo E, Madsen EL, Poulsen SS. Epidermal growth factor in mammary glands and milk from rats: the influence of insulin. Diab Res Clin Pract 1993;21(1):11-8. 258. Auestad N, Korsak RA, Bergstrom JD, Edmond J. Milk-substitutes comparable to rat's milk; their preparation, composition and impact on development and metabolism in the artificially reared rat. Br J Nutr 1989;61(3):495-518. 259. Innis SM, Nelson CM, Rioux MF, King DJ. Development of visual acuity in relation to plasma and erythrocyte omega-6 and omega-3 fatty acids in healthy term gestation infants. Am J Clin Nutr 1994;60(3):347-52. 260. Rapoport SI, Rao JS, Igarashi M. Brain metabolism of nutritionally essential polyunsaturated fatty acids depends on both the diet and the liver. Prostaglandins Leukot Essent Fatty Acids 2007;77(5-6):251-61.  261. DeMar JC, Jr., Ma K, Bell JM, Rapoport SI. Half-lives of docosahexaenoic acid in rat brain phospholipids are prolonged by 15 weeks of nutritional deprivation of n-3 polyunsaturated fatty acids. J Neurochem 2004;91(5):1125-37.  262. Litman BJ, Niu SL, Polozova A, Mitchell DC. The role of docosahexaenoic acid containing phospholipids in modulating G protein-coupled signaling pathways: visual transduction. J Molec Neurosci : MN 2001;16(2-3):237-42; discussion 79-84.  263. de la Presa Owens S, Innis SM. Docosahexaenoic and arachidonic acid prevent a decrease in dopaminergic and serotoninergic neurotransmitters in frontal cortex caused by a linoleic and alpha-linolenic acid deficient diet in formula-fed piglets. J Nutr 1999;129(11):2088-93. 264. Zimmer L, Vancassel S, Cantagrel S, et al. The dopamine mesocorticolimbic pathway is affected by deficiency in n-3 polyunsaturated fatty acids. Am J Clin Nutr 2002;75(4):662-7. 265. Lopez C, Cauty C, Guyomarc'h F. Organization of lipids in milks, infant milk formulas and various dairy products: role of technological processes and potential impacts. Dairy Sci Tech 2015;95(6):863-93.  266. Kitson AP, Stark KD, Duncan RE. Enzymes in brain phospholipid docosahexaenoic acid accretion: a PL-ethora of potential PL-ayers. Prostaglandins Leukot Essent Fatty Acids 2012;87(1):1-10. 157  267. Chih C-P, Roberts EL, Jr. Energy Substrates for Neurons During Neural Activity: A Critical Review of the Astrocyte-Neuron Lactate Shuttle Hypothesis. J Cereb Blood Flow Metab 2003;23(11):1263-81. 268. Kurbat MN, Lelevich VV. Metabolism of amino acids in the brain. Neurochem 2009;3(1):23-8.  269. Daikhin Y, Yudkoff M. Compartmentation of brain glutamate metabolism in neurons and glia. J Nutr 2000;130(4S Suppl):1026s-31s. 270. Mark LP, Prost RW, Ulmer JL, et al. Pictorial review of glutamate excitotoxicity: fundamental concepts for neuroimaging. Am J Neurorad 2001;22(10):1813-24. 271. Pardo B, Contreras L, Satrústegui J. De novo Synthesis of Glial Glutamate and Glutamine in Young Mice Requires Aspartate Provided by the Neuronal Mitochondrial Aspartate-Glutamate Carrier Aralar/AGC1. Front  Endocrinol 2013;4:149.  272. Bielarczyk H, Lysiak W, Szutowicz A. Synthesis of glutamate and aspartate in rat brain synaptosomes. Acta biochimica Polonica 1986;33(4):239-51. 273. Hawkins RA. The blood-brain barrier and glutamate. Am J Clin Nutr 2009;90(3):867S-74S. 274. Smith QR. Transport of glutamate and other amino acids at the blood-brain barrier. J Nutr 2000;130(4S Suppl):1016s-22s. 275. Nishizuka Y. Protein kinase C and lipid signaling for sustained cellular responses. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 1995;9(7):484-96. 276. Berridge MJ. Inositol trisphosphate and calcium signalling mechanisms. Biochim Biophys Acta 2009;1793(6):933-40.  277. Berridge MJ. Neuronal calcium signaling. Neuron 1998;21(1):13-26. 278. Lucas A, Gibbs JA, Lyster RL, Baum JD. Creamatocrit: simple clinical technique for estimating fat concentration and energy value of human milk. BMJ 1978;1(6119):1018-20. 279. Hammond JB, Kruger NJ. The bradford method for protein quantitation. Methods in molecular biology (Clifton, NJ) 1988;3:25-32.  280. Wang H, Eckel RH. Lipoprotein lipase in the brain and nervous system. Annu Rev Nutr 2012;32:147-60.  281. Nguyen LN, Ma D, Shui G, et al. Mfsd2a is a transporter for the essential omega-3 fatty acid docosahexaenoic acid. Nature 2014;509(7501):503-6.  158  282. Lagarde M, Bernoud N, Brossard N, et al. Lysophosphatidylcholine as a preferred carrier form of docosahexaenoic acid to the brain. J Mol Neurosci 2001;16(2-3):201-4; discussion 15-21.  283. Mansbach CM, 2nd. The origin of chylomicron phosphatidylcholine in the rat. J Clin Invest 1977;60(2):411-20.  284. Cheng W-L, Holmes-McNary MQ, Mar M-H, Lien EL, Zeisel SH. Bioavailability of choline and choline esters from milk in rat pups. J. Nutr. Biochem 1996;7(8):457-64.  285. Landers S, Hartmann BT. Donor human milk banking and the emergence of milk sharing. Pediatr Clin North Am 2013;60(1):247-60.  286. O'Connor DL, Ewaschuk JB, Unger S. Human milk pasteurization: benefits and risks. Current opinion in clinical nutrition and metabolic care 2015;18(3):269-75. 287. McMahon KE, Farrell PM. Measurement of free choline concentrations in maternal and neonatal blood by micropyrolysis gas chromatography. Clinica chimica acta 985;149(1):1-12. 288. Sweiry JH, Page KR, Dacke CG, Abramovich DR, Yudilevich DL. Evidence of saturable uptake mechanisms at maternal and fetal sides of the perfused human placenta by rapid paired-tracer dilution: studies with calcium and choline. J Dev Physiol 1986;8(6):435-45. 289. Molloy AM, Mills JL, Cox C, et al. Choline and homocysteine interrelations in umbilical cord and maternal plasma at delivery. Am J Clin Nutr 2005;82(4):836-42. 290. Bernhard W, Raith M, Kunze R, et al. Choline concentrations are lower in postnatal plasma of preterm infants than in cord plasma. Eur J Nutr 2015;54(5):733-41.  291. Garner SC, Chou SC, Mar MH, Coleman RA, Zeisel SH. Characterization of choline metabolism and secretion by human placental trophoblasts in culture. Biochim Biophys Acta 1993;1168(3):358-64. 292. Ardiel EL, Rankin CH. The importance of touch in development. J Paediatr Child Health 2010;15(3):153-6. 293. Jorgensen MH, Hernell O, Hughes E, Michaelsen KF. Is there a relation between docosahexaenoic acid concentration in mothers' milk and visual development in term infants? J Pediatr Gastroenterol Nutr 2001;32(3):293-6. 294. Iranpour R, Kelishadi R, Babaie S, Khosravi-Darani K, Farajian S. Comparison of long chain polyunsaturated fatty acid content in human milk in preterm and term deliveries and its correlation with mothers' diet. J Res Med Sci 2013;18(1):1-5. 295. Lassek WD, Gaulin SJ. Linoleic and docosahexaenoic acids in human milk have opposite relationships with cognitive test performance in a sample of 28 countries. Prostaglandins Leukot Essent Fatty Acids 2014;91(5):195-201.  159  296. Pedersen L, Lauritzen L, Brasholt M, Buhl T, Bisgaard H. Polyunsaturated fatty acid content of mother/'s milk is associated with childhood body composition. Pediatr Res 2012;72(6):631-6. 297. Kelishadi R, Hadi B, Iranpour R, et al. A study on lipid content and fatty acid of breast milk and its association with mother's diet composition. J Res Med Sci 2012;17(9):824-7. 298. Much D, Brunner S, Vollhardt C, et al. Breast milk fatty acid profile in relation to infant growth and body composition: results from the INFAT study. Pediatr Res 2013;74(2):230-7.  299. Bienenstock J, Kunze W, Forsythe P. Microbiota and the gut-brain axis. Nutri Res 2015;73 Suppl 1:28-31.  300. Holzer P, Farzi A. Neuropeptides and the microbiota-gut-brain axis. Advances Exp Med Biol 2014;817:195-219.  301. Bussey TJ, Padain TL, Skillings EA, Winters BD, Morton AJ, Saksida LM. The touchscreen cognitive testing method for rodents: how to get the best out of your rat. Learning & memory (Cold Spring Harbor, NY) 2008;15(7):516-23. 302. Kennedy MJ, Griffin A, Su R, Merchant M, Klein J. Urine collected from diapers can be used for 2-D PAGE in infants and young children. Proteomics Clinical applications 2009;3(8):989-99.  303. Burke N. Alternative methods for newborn urine sample collection. Pediatr Nurs 1995;21(6):546-9. 304. Fleith M, Clandinin MT. Dietary PUFA for preterm and term infants: review of clinical studies. Crit Rev Food Sci Nutr 2005;45(3):205-29. 305. Koletzko B, Lien E, Agostoni C, et al. The roles of long-chain polyunsaturated fatty acids in pregnancy, lactation and infancy: review of current knowledge and consensus recommendations. Journal Perinat Med 2008;36(1):5-14.  306. Vallejo M, Garcia A, Tunon J, et al. Plasma fingerprinting with GC-MS in acute coronary syndrome. Anal Bioanal Chem 2009;394(6):1517-24.   160  Appendices Appendix A. Supplemental analytical methods   A.1 Procedure for Lipid Extraction  This procedure is modified from Folch et al., 1957 (45)  1. To extract lipids from milk, add 1 ml of milk, 1.25 mL saline, 3 ml methanol, and 6 ml chloroform. To extract lipids from frontal lobe, homogenize 50 mg frontal lobe in 2.25 ml saline using a mortar and pestle, then add 3 ml methanol and 6 ml chloroform to the homogenate.  2. Vortex briefly.  3. Centrifuge tubes at 2000 revolutions per minute (rpm) for 5 minutes at 4ºC to allow separation of the organic and aqueous layers. 4. Remove the lower organic layer.  5. Add 6 ml chloroform to the remaining aqueous layer and vortex briefly. 6. Centrifuge at 2000 rpm for 5 minutes at 4°C. Remove the organic layer and combine with the first organic layer collection.  7. Dry down the organic solvent under nitrogen using a nitrogen evaporator at room temperature.  A.2 Procedure for Methylating Fatty Acids of Different Lipid Fractions  1. Add heptadecanoic acid (17:0).as an internal standard. 2. Add reagents to samples as per Table A2.1, vortex and cap tubes tightly. Incubate tubes at 110ºC for the amount of time provided in Table A2.1.   Table A2.1 Reagents and methylation times for fatty acid methylation.    161  Fraction  BF3 (mL)  Methanol  Benzene  Methylation Time  (minutes) Total Phospholipid (TPL)  1.0  --  --  15  Free Fatty Acids (FFA)  0.5  --  --  15  Pls-PE, PE, PC, PS  1.0  --  --  15  PI  1.0   --  --  20  Sphingomyelin  1.0  --  --  90  Triglycerides (TG)  0.25  0.55  0.20  30  Cholesterol Esters (CE)  0.35  0.35  0.30  45    3. Cool samples at 4ºC for 10 minutes. 4. Add 3 mL saline and 6 mL hexane to the test tube to extract the fatty acid methyl esters and dimethyl acetals. Vortex for 15 seconds.  5. Vortex for 15 seconds then place the test tube on the bench for 2-3 minutes to allow separation of the organic and aqueous layers. 6. Remove the upper hexane layer and transfer to another test tube.   7. Add another 6 mL of hexane to the original test tube. Repeat steps 5 and 6, combining the new hexane layer with the first hexane extraction.   8. Dry down the hexane under nitrogen using a nitrogen evaporator at room temperature.   9. Once completely dry, rinse the walls of the test tubes with 200-300 µL hexane to concentrate the fatty acid methyl esters at the bottom of the tube. Dry gently under nitrogen.  10.  Re-suspend the fatty acid methyl esters and dimethyl acetals in an appropriate volume of hexane (e.g 12 µl for Pls-PE and 25 µl for diacyl-PE and PC). Transfer each sample to a GLC autosample vial.  162  A.3 Procedure for Methylating Total Fatty acids in Milk   1. To a 50 µl milk sample, add 250 µg of each of the following internal standards: nonanoic acid (9:0), tridecanoic acid (13:0), and heptadecanoic acid (17:0) to a 16 ×100 mm in a screw-capped tube. 2. Add 2 mL of benzene/methanol (1/4, v/v). Vortex for 15 seconds.   3. Add 200 µL of acetyl chloride, 50 µL at a time.   4. Cap the tubes tightly. Vortex briefly. Incubate at 110ºC for 60 minutes, vortexing briefly every 15 minutes.  5. Cool samples at 4ºC for 10 minutes.   6. Slowly add 5 mL of freshly-made 6% potassium carbonate solution.    7. Cap the tubes tightly and shake for 3-5 minutes.  8. Carefully open the tubes and add 4 mL of hexane.  Vortex for 15 seconds.  9. Centrifuge at 2000 rpm for 10 minutes at 4ºC.  10. Remove the upper hexane layer, and transfer to another test tube.    11. Add another 4 mL of hexane to the original test tube, and vortex for 15 seconds. Repeat steps 9 and 10, combining the new upper layer with the first hexane extraction.  12. Dry down the hexane under nitrogen using a nitrogen evaporator at room temperature. 13. Once completely dry, rinse the walls of the test tubes with 200-300 µL hexane to concentrate the fatty acid methyl esters and dimethyl acetals at the bottom of the tube. Dry gently under nitrogen. Care should be taken to remove the samples off the evaporator directly after complete drying to prevent the evaporation of the volatile short and medium chain fatty acid methyl esters.  163  14.  Re-suspend the fatty acid methyl esters and dimethyl acetals in 500 µL of hexane and transfer the sample to a GLC autosample vial.  A.4 Procedure for Determining Protein Content in Brain Samples  This procedure was modified from Bradford, 1976 (279). Sample aliquots for protein determination are prepared concurrently during lipid extraction, by aliquoting 10 µl of the aqueous phase in a 1.5 ml Epperdorf tube, in duplicate for each sample.  1. To develop a standard curve, prepare 8 dilutions of Bovine serum albumin (BSA; 1 mg/ml) with distilled water, as per Table A4.1. It is important to keep all preparations on ice to prevent protein breakdown. Table A4.1. Volumes of BSA standard and distilled water for BSA dilutions preparation BSA Dilution (mg/ml) BSA standard (µl) Distilled Water (µl) 0.5  50 50 0.4 40 60 0.35 35 65 0.3 30 70 0.2 20 80 0.15 15 85 0.1 10 90 0 0 100  2. Prepare a 1:10 dilution of the brain sample aliquots, by adding 90 µl distilled water into each Epperdorf tube. Vortex briefly. 3. Prepare a 1:5 dilution of the protein assay dye reagent concentrate (Biorad) with distilled water.  164  4. Filter the diluted reagent using a filter paper (Whatman #1) to ensure the reagent is free of any physical contaminants. 5. Add 10 µl of each of the standard BSA preparations and samples in a separate well of a 96-well plate. 6. Add 200 µl of the protein assay reagent to each well, by reverse multichannel pipetting to ensure a consistent and accurate addition of the reagent. 7. Shake the plate gently and keep the plate at room temperature for 15 minutes. 8. Read the absorbance at 595 nm.  A.5 Procedure for Preparing Brain Samples for GC-MS analysis.  This method was modified from Vallejo et al, 2009 (306). Sample aliquots for derivatizing metabolites are prepared concurrently during lipid extraction, by collecting and drying the aqueous phase in 1.5 ml Eppendorf tubes, using a Speedvac centrifugal concentrator (Speedvac). 1. Add 100 µl methoxylamine HCL in pyridine (20 mg/ml) to each tube. Vortex briefly. 2. Incubate the tubes in a water bath at 40 ⁰C for two hours, sonicating the samples at low power for 5 minutes, every 15 minutes.  3. Dry the pyridine completely from the samples using the Speedvac. This typically takes one hour. 4. Add 75 µl of N-methyl-n-trimethylsilyl trifluoroacetamide to each tube. Vortex briefly. 5. Incubate the tubes in a water bath at 40 ⁰C for one hour, sonicating the samples at low power for 5 minutes, every 15 minutes. 6. Centrifuge the samples at 35000 rpm for 15 minutes at 4⁰C. Transfer 50 µl of the supernatant to a 250 µl insert and a GLC vial.  165  7. Add an additional 50 µl hexane into the insert. Vortex briefly.   A.6 Procedure for Cannula Preparation  1. Cut plastic disks (0.75-1.00 cm diameter) from commercial sealable plastic bags using a single hole punch.  2. Pierce the center of the disk with a 30 Gauge needle to create a small opening for the PE-10 silica tubing. Insert one end of a 25 cm PE-10 tubing into the center of the disk by approximately 1 cm. 3. Flame that end of the PE-10 tubing over a candle. This will result in a slight enlargement of the tubing tip, which will prevent the disk from detaching from the tubing. 4. Insert a 30 Gauge needle into the flamed tip of the PE-10 tubing. Briefly heat the needle with a soldering iron. This will result in minimal melting and therefore further anchoring of the PE-10 tubing to the disk.  5. Remove the needle carefully and check for any leakages in the tubing using a saline solution. 166  Appendix B. Supplemental List of Total Milk Fatty Acids for Individual Participants on 3 Separate Days              Participant #1  Participant #2  Participant #3  Participant #4   d1 d2 d3  d1 d2 d3  d1 d2 d3  d1 d2 d3 8:0 0.02 0.07 0.02  0.21 0.05 0.02  0.16 0.04 0.05  0.08 0.09 0.08 10:0 0.78 0.93 0.87  1.10 0.90 0.35  1.25 1.01 1.18  0.89 0.82 0.82 12:0 4.37 4.23 4.21  3.81 4.37 2.92  5.01 4.79 5.29  3.63 3.64 3.20 14:0 5.15 5.55 4.86  4.73 7.13 4.06  5.99 6.53 6.34  4.96 3.96 5.13 14:1 0.08 0.29 0.28  0.02 0.44 0.24  0.27 0.38 0.31  0.28 0.16 0.29 16:0 17.1 19.4 16.7  18.6 24.9 16.6  19.5 22.2 20.1  21.4 19.7 18.7 16:1 ω-9 2.15 2.90 0.75  2.21 3.20 2.12  1.93 2.91 2.34  2.26 1.89 1.88 18:0 6.16 6.16 5.22  6.76 7.99 6.12  8.02 9.01 6.84  6.16 6.92 5.66 18:1 ω-9 41.6 39.3 46.5  44.2 33.8 46.1  42.2 37.4 36.5  40.2 43.2 41.6 18:1 ω-7 1.68 1.58 1.62  0.19 2.01 1.67  0.18 1.59 1.55  1.59 1.43 1.40 18:2 ω-6 16.7 15.2 15.4  14.2 11.2 15.8  10.7 9.7 14.1  14.9 14.3 16.6 18:3 ω-6 0.07 0.12 0.07  0.04 0.15 0.16  0.09 0.04 0.12  0.05 0.07 0.08 20:0 0.20 0.18 0.06  0.18 0.20 0.17  0.36 0.30 0.23  0.22 0.23 0.26 18:3 ω-3 1.53 1.46 1.40  1.62 1.24 1.17  0.84 1.21 1.69  1.02 1.26 1.84 20:1 0.51 0.39 0.51  0.06 0.45 0.38  0.83 0.54 0.47  0.46 0.56 0.57 18:4 ω-3 0.02 0.02 0.01  0.05 0.00 0.03  0.02 0.00 0.04  0.00 0.01 0.01 20:2 ω-6 0.21 0.17 0.21  0.20 0.24 0.22  0.21 0.14 0.19  0.25 0.29 0.24 20:3 ω-9 0.03 0.09 0.12  0.02 0.07 0.03  0.04 0.11 0.03  0.08 0.09 0.03 20:3 ω-6 0.23 0.29 0.24  0.32 0.38 0.34  0.32 0.25 0.32  0.23 0.20 0.25 20:4 ω-6 0.43 0.45 0.50  0.43 0.38 0.42  0.48 0.46 0.57  0.43 0.45 0.45 22:0 0.07 0.07 0.07  0.06 0.08 0.06  0.23 0.11 0.15  0.15 0.13 0.19 22:1 0.15 0.14 0.17  0.13 0.16 0.13  0.13 0.10 0.10  0.14 0.09 0.10 20:5 ω-3 0.19 0.22 0.37  0.12 0.08 0.15  0.17 0.15 0.26  0.05 0.06 0.05 24:0 0.06 0.04 0.07  0.04 0.06 0.05  0.19 0.08 0.10  0.12 0.07 0.11 22:4 ω-6 0.05 0.03 0.08  0.05 0.08 0.05  0.09 0.06 0.09  0.07 0.08 0.08 22:5 ω-6 0.04 0.04 0.06  0.03 0.06 0.03  0.02 0.06 0.06  0.05 0.04 0.05 22:5 ω-3 0.16 0.17 0.26  0.21 0.13 0.14  0.22 0.26 0.35  0.13 0.12 0.13 22:6 ω-3 0.34 0.37 0.60  0.43 0.28 0.32  0.44 0.54 0.65  0.17 0.13 0.14 167  Appendix C. Fatty Acid Composition of Experimental Formulas and MFGM phospholipids   Total fatty acids1  MFGM-PE MFGM-PC MFGM-PS MFGM-PI 8:0  24.6  <0.01 <0.01 <0.01 <0.01 10:0  13.4  <0.01 <0.01 <0.01 <0.01 12:0  7.58  0.18 0.30 0.15 <0.01 14:0  2.60  0.89 8.50 0.19 1.37 14:1 0.03  0.24 <0.01 0.07 0.46 16:0  7.37  10.5 39.3 3.94 19.7 16:1  0.10  1.65 14.3 0.97 1.11 18:0  2.30  13.8 5.80 38.6 45.1 18:1  14.1  52.2 21.8 37.7 20.8 18:2 ω-6  25.8  11.6 6.70 9.19 3.06 18:3 ω-6 <0.01  0.08 0.10 0.16 <0.01 20:0  <0.01  0.13 <0.01 1.80 <0.01 18:3 ω-3  1.86  1.09 0.74 0.75 0.16 20:1 0.18  0.75 0.33 0.61 0.29 18:4 ω-3 <0.01  0.04 0.01 <0.01 <0.01 20:2 ω-6 <0.01  0.13 0.08 0.10 0.12 20:3 ω-9 <0.01  <0.01 <0.01 <0.01 <0.01 20:3 ω-6 0.04  1.21 0.43 2.03 <0.01 20:4 ω-6 <0.01  1.38 0.32 0.16 0.84 20:3 ω-3 <0.01  <0.01 <0.01 <0.01 <0.01 22:0 <0.01  <0.01 0.02 0.33 0.08 20:5 ω-3 <0.01  0.33 0.13 <0.01 0.14 24:0 0.01  0.16 0.35 0.35 0.09 22:4 ω-6  0.10  0.25 0.05 0.63 0.13 24:1  <0.01  0.03 0.05 0.08 0.07 22:5 ω-6 <0.01  0.04 0.15 0.10 1.65 22:5 ω-3 <0.01  0.68 0.15 1.87 0.06 22: 6 ω-3 <0.01  0.12 0.05 0.21 <0.01 Values are g/100 g fatty acid. 1 Both formulas contain similar total fatty acid compositions. MFGM: milk fat globule membrane (Lacprodan® MFGM-10, Arla Foods Ingredients, Denmark). PE: phosphatidylethanolamine; PC: phosphatidylcholine; PS: phosphatidylserine; PI: phosphatidylinosito 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0228326/manifest

Comment

Related Items