UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Application of metabolomics to identify metabolite patterns showing the importance of dietary carbohydrate… Chetty, Vinodha 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_february_chetty_vinodha.pdf [ 2.63MB ]
Metadata
JSON: 24-1.0223696.json
JSON-LD: 24-1.0223696-ld.json
RDF/XML (Pretty): 24-1.0223696-rdf.xml
RDF/JSON: 24-1.0223696-rdf.json
Turtle: 24-1.0223696-turtle.txt
N-Triples: 24-1.0223696-rdf-ntriples.txt
Original Record: 24-1.0223696-source.json
Full Text
24-1.0223696-fulltext.txt
Citation
24-1.0223696.ris

Full Text

Application of metabolomics to identify metabolite patterns showing the importance of dietary carbohydrate source in neonatal rats  by Vinodha Chetty  B.Tech, Sri Ramaswamy Memorial University, 2012  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Reproductive and Developmental Sciences)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) January 2016  © Vinodha Chetty, 2016 ii  ABSTRACT  Lactose is a glucose and galactose disaccharide and is found exclusively in mammalian milk. The evolution of lactose as a unique component of human milk (with 50% galactose) and its metabolic advantage in infants is poorly understood. There is also a limited understanding of the potential implications of clinical nutrition support using lactose-free nutrition, such as dextrose (glucose) as the sole carbohydrate source in intravenous nutrition and glucose-corn syrup solids (CSS) (100% glucose) in formula for infants. The liver takes up 90% of galactose. In contrast, the liver clears only 30% of glucose. Glucose stimulates release of insulin from the pancreas whereas galactose does not. The goal of my thesis was to use ‘metabolomics’, in a neonatal rat model of gastrostomy feeding, to differentiate metabolic effects in rat pups fed milk formula complete in protein, fat, minerals and vitamins, but with either lactose or CSS as the carbohydrate source. I hypothesize that galactose and specific metabolites of galactose metabolism will be differentiated between lactose-fed and CSS-fed rat pups by target compound analysis; and targeted metabolomics will highlight differences in metabolite patterns in hepatic metabolism from lactose or CSS feeding. Liver, plasma and urine samples were collected at 10 and 14 days after birth. Significant differences in galactose and galactonate levels were observed in the liver at day 10 and 14, with no differences glucose. Metabolites that were different between the groups were: D-Ribose, pyrimidine, glycine and malate in rat pups at day 10 and leucine and isoleucine at day 10 and 14, which were significantly higher in the lactose-fed group. The data obtained show that early dietary lactose has definitive effects on hepatic metabolites iii  that can be mapped to pathways of carbohydrate, protein, and fat metabolism and are different to non-lactose carbohydrate feeding. My findings suggest that utilization of galactose, as a consistent source of glucose could prove beneficial for supply to glucose-dependent organs such as the rapidly developing infant brain. Advancing the use of metabolomics to enhance understanding of the impact of diet will provide an opportunity to improve health outcome for infants. iv  PREFACE This thesis contains unpublished work accomplished by myself, under the supervision of Dr. Sheila M. Innis, and was prepared in accordance to the University of British Columbia and the Faculty of Graduate studies. Principal investigator, Dr. Sheila M. Innis generated the hypothesis and study design for this project. Gastrostomy experiments and sample collection was done by myself, with help from lab members, Roger A. Dyer, Dr. Cyrielle Garcia, Alejandra Wiedeman, Jannette King, Sara Moukarzel, Dr. Jie Yang and Kelly Mulder. I prepared the samples for GC-MS and LC-qTOF-ESI-MS analyses. I performed GC-MS analyses and developed methods for Principal Component analysis, with guidance from Roger A. Dyer. Roger A. Dyer did the initial instrument set-up. Dr. Bernd Keller performed LC-qTOF-ESI-MS analyses. I did sample preparation and western blotting analysis, referring from data collected from in-vitro studies performed by Dr. Jie Yang. I, along with Dr.Sheila M. Innis, interpreted the data from analyses. Animal experiments in this thesis were approved by the University of British Columbia Animal Care Committee (protocol: A13-0257).       v  TABLE OF CONTENTS  ABSTRACT ..................................................................................................................... ii PREFACE ....................................................................................................................... iv TABLE OF CONTENTS .................................................................................................. v LIST OF TABLES ......................................................................................................... viii LIST OF FIGURES ......................................................................................................... ix LIST OF ABBREVIATIONS.......................................................................................... xiii ACKNOWLEDGEMENTS ............................................................................................ xvi DEDICATION ............................................................................................................... xvii CHAPTER 1: INTRODUCTION ....................................................................................... 1 Lactose composition, digestion and absorption ..................................... 5 Differences between galactose and glucose metabolism ...................... 7      Galactose metabolism ........................................................................... 8      Glucose metabolism ............................................................................ 12 Metabolomics workflow ....................................................................... 18 Analytical techniques .................................................................. 19 Ion separation, ion sorting and detection .................................... 20 vi  Target compound analysis and targeted metabolomics ...................... 22      Principal component analysis ..................................................... 24 CHAPTER 2: STUDY DESIGN AND METHODS .......................................................... 31 Rat milk formula composition and  preparation ................................... 33      Gastrostomy experiment ..................................................................... 35      Rearing of rat pups ..................................................................... 35      Tissue and biofluid collection ...................................................... 37      Target compound analysis .................................................................. 39      Preparation of liver, plasma and urine samples .......................... 41      Instruments and analysis ............................................................ 44      Targeted metabolomics ....................................................................... 47      Principal component analysis ..................................................... 47 CHAPTER 3: RESULTS ................................................................................................ 51 58 vii       Plasma ................................................................................................. 71 Urine .................................................................................................... 75 Day 10 liver samples ........................................................................... 80      Day 14 liver samples ........................................................................... 90 CHAPTER 4: SUMMARY .............................................................................................. 95      Target compound analysis ................................................................... 98      Targeted metabolomics ..................................................................... 100      Western blotting ................................................................................. 105 REFERENCES ............................................................................................................ 114 viii  LIST OF TABLES TABLE 1. Macronutrient composition of neonatal rat milk ............................................. 34 TABLE 2. Macronutrient composition of rat milk formula ............................................... 34 TABLE 3.  List of metabolites analyzed for target compound analysis .......................... 40 TABLE 4. Mean body weight (g) and liver weight data (g) of mother-reared and lactose or CSS-fed group of rat pups, at days 5, 10, 14 and 18 after birth. ................... 55 TABLE 5. List of metabolites from loadings view plot of PCA for day 10 liver ............... 82 TABLE 6: List of metabolites from loadings view plot of PCA for day 14 liver ............... 91  ix  LIST OF FIGURES FIGURE 1. Structural representation of lactose ............................................................... 4 FIGURE 2. Leloir pathway ............................................................................................... 9 FIGURE 3. Galactose dehygrogenase pathway and aldose reductase pathways of galactose  metabolism ..................................................................................................... 9 FIGURE 4. Graph representing galactose dehydrogenase activity ................................ 11 FIGURE 5. Glucose metabolism in liver ......................................................................... 13 FIGURE 6. A typical cycle of metabolomics workflow .................................................... 19 FIGURE 7. Rat pup growth curve .................................................................................. 54 FIGURE 8a. GC-MS chromatogram of day 10 liver samples depicting galactose ......... 59 FIGURE 8b. Day 10 liver bar graph (mean area of peak) of galactose .......................... 59 FIGURE 9a. GC-MS chromatogram of day 14 liver depicting galactose ........................ 61 FIGURE 9b. Day 14 liver bar graph (mean area of peak) of galactose .......................... 61 FIGURE 10a. GC-MS chromatogram of day 10 liver depicting galactonate .................. 63 FIGURE 10b. Day 10 liver bar graph (mean area of peak) of galactonate .................... 63 FIGURE 11a. GC-MS chromatogram of day 14 liver depicting galactonate .................. 65 FIGURE 11b. Day 14 liver bar graph (mean area of peak) of galactonate .................... 65 FIGURE 12a. GC-MS chromatogram of day 10 liver depicting glucose ......................... 67 FIGURE 12b. Day 10 liver bar graph (mean area of peak) of glucose ........................... 67 FIGURE 13a. GC-MS chromatogram of day 14 liver depicting glucose ......................... 68 FIGURE 13b. Day 14 liver bar graph (mean area of peak) of glucose ........................... 68 FIGURE 14. Day 10 liver bar graph (mean area of peak) of Glu-6-P ............................. 69 FIGURE 15. Day 14 liver bar graph (mean area of peak) of Glu-6-P ............................. 69 x  FIGURE 16. Day 10 liver bar graph (mean area of peak) of ribose-5-phosphate .......... 70 FIGURE 17. Day 14 liver bar graph (mean area of peak) of ribose-5-phosphate .......... 70 FIGURE 18. Day 10 plasma bar graph (mean area of peak) of glucose ........................ 71 FIGURE 19. Day 14 plasma bar graph (mean area of peak) of glucose ........................ 72  FIGURE 20. Day 10 plasma bar graph (mean area of peak) of Glu-6-P ........................ 73 FIGURE 21. Day 14 plasma bar graph (mean area of peak) of Glu-6-P ........................ 73 FIGURE 22. Day 10 plasma bar graph (mean area of peak) of ribose-5-phosphate ..... 74 FIGURE 23. Day 14 plasma bar graph (mean area of peak) of ribose-5-phosphate ..... 74  FIGURE 24. Day 10 urine bar graph (mean area of peak) of galactonate ..................... 76 FIGURE 25. Day 14 urine bar graph (mean area of peak) of galactonate ..................... 76 FIGURE 26. Day 10 urine bar graph (mean area of peak) of Glu-6-P ........................... 77 FIGURE 27. Day 14 urine bar graph (mean area of peak) of Glu-6-P ........................... 77 FIGURE 28. Day 10 urine bar graph (mean area of peak) of ribose-5-phosphate ......... 78 FIGURE 29. Day 14 urine bar graph (mean area of peak) ribose-5-phosphate ............. 78 FIGURE 30. Day 10 Liver Loadings view plot from Principal component analysis......... 81 FIGURE 31a. GC-MS chromatogram of day 10 liver depicting D-ribose ....................... 84 FIGURE 31b. Day 10 liver bar graph (mean area of peak) of D-ribose ......................... 84 FIGURE 32a. GC-MS chromatogram of day 10 liver depicting pyrimidine ..................... 85 FIGURE 32b. Day 10 liver bar graph (mean area of peak) of pyrimidine ....................... 85 FIGURE 33a. GC-MS chromatogram of day 10 liver depicting glycine .......................... 86 FIGURE 33b. Day 10 liver bar graph (mean area of peak) of glycine ............................ 86 FIGURE 34a. GC-MS chromatogram of day 10 liver depicting malate .......................... 87 FIGURE 34b. Day 10 liver bar graph (mean area of peak) of malate i .......................... 87 xi  FIGURE 35a. GC-MS chromatogram of day 10 liver depicting leucine .......................... 88 FIGURE 35b. Day 10 liver bar graph (mean area of peak) of leucine ............................ 88 FIGURE 36a. GC-MS chromatogram of day 10 liver depicting isoleucine ..................... 89 FIGURE 36b. Day 10 liver bar graph (mean area of peak) of isoleucine ....................... 89 FIGURE 37. Day 14 Liver Loadings view plot from Principal component analysis......... 90 FIGURE 38a. GC-MS chromatogram of day 14 liver depicting leucine .......................... 92 FIGURE 38b. Day 14 liver bar graph (mean area of peak) of leucine ............................ 92 FIGURE 39a. GC-MS chromatogram of day 14 liver depicting isoleucine ..................... 93 FIGURE 39b. Day 14 liver bar graph (mean area of peak) of isoleucine ....................... 93 FIGURE 40. Biochemical reactions depicting D-ribose metabolism ............................. 101 FIGURE 41. Biochemical reactions depicting glycine metabolism ............................... 102 FIGURE 42a. Western blot showing abundance of ChREBP in day 10 rat liver .......... 135 FIGURE 42b. Day 10 liver bar graph (ChREBP/Actin ratio), by western blotting. ........ 135 FIGURE 43a. Western blot showing abundance of SREBP1-c in day 10 rat liver ....... 136 FIGURE 43b. Day 10 liver bar graph (SREBP1-c/Actin ratio), by western blotting. ..... 136 FIGURE 44a. Western blot showing abundance of SCD in day 10 rat liver ................. 137 FIGURE 44b. Day 10 liver bar graph (SCD/Actin ratio), by western Blotting ............... 137 FIGURE 45a. Western blot showing abundance of ChREBP in day 14 rat liver .......... 138 FIGURE 45b. Day 14 liver bar graph (ChREBP/Actin ratio), by western blotting ......... 138 FIGURE 46a. Western blot showing abundance of SREBP1-c in day 14 rat liver ....... 139 FIGURE 46b. Day 14 liver bar graph (SREBP1-c/Actin ratio), by western blotting. ..... 139 FIGURE 47a. Western blot showing abundance of SCD in day 14 rat liver ................. 140 FIGURE 47b. Day 14 liver bar graph (SCD/Actin ratio), by western blotting. .............. 140 xii  FIGURE 48a. Western blot showing abundance of ChREBP in day 18 rat liver .......... 141 FIGURE 48b. Day 18 liver bar graph (ChREBP/Actin ratio) d, by western blotting. ..... 141 FIGURE 49a. Western blot showing abundance of SREBP1-c in day 18 rat liver ....... 142 FIGURE 49b. Day 18 liver bar graph (SREBP1-c/Actin ratio), by western blotting. ..... 142 FIGURE 50a. Western blot showing abundance of SCD in day 18 rat liver ................. 143 FIGURE 50b. Day 18 liver bar graph (SCD/Actin ratio), by western blotting. .............. 143          xiii  LIST OF ABBREVIATIONS  ADP: adenosine diphosphate ATP: adenosine triphosphate BCAA: branched chain amino acids CE: capillary electrophoresis ChREBP: carbohydrate response element-binding protein CSS: corn syrup solids DNA: deoxyribonucleic acid EI: electron impact EIC: extracted ion chromatogram ER: endoplasmic reticulum ESI: electrospray ionization Gal-1-P: galactose -1-phosphate GALT: galactose-1-phosphate uridylyltransferase GC-MS: gas chromatography mass spectrometry GC: gas chromatography GCS: glycine cleavage system Glu-1-P: glucose-1-phosphate Glu-6-P: glucose-6-phosphate GLUT2: glucose/galactose transporter 2 HCFS: high fructose corn syrup solids HPLC: high performance liquid chromatography xiv  IUGR: Intrauterine growth restriction LC-MS: liquid chromatography mass spectrometry LC-qTOF-ESI-MS: liquid Chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry m/z: mass/charge MALDI: matrix assisted laser desorption/ionization MS: Mass spectrometry MSTFA: N-methyl-N- (trimethyl silyl) trifluoroacetamide MW: molecular weight NAD: nicotinamide adenine dinucleotide NADPH: nicotinamide adenine dinucleotide phosphate NMR: nuclear magnetic resonance OAA: oxaloacetate PCA: principal component analysis PPP: pentose phosphate pathway RBC: red blood cell RNA: ribonucleic acid RT: retention time SCD: stearoyl-CoA desaturase SGLUT1: sodium dependent glucose/galactose cotransporter 1 SREBP: sterol regulatory element-binding protein TBA: tributylamine TCA cycle: tricarboxylic acid cycle xv  TFA: trifluoro acetic acid TIC: total ion chromatogram TOF: time of flight UDP-Gal: uridine diphosphate galactose α- LA: α- lactalbumin      xvi  ACKNOWLEDGEMENTS I would like to offer my gratitude to my supervisor, Dr. Sheila M.Innis for providing me this opportunity and instilling her experience and knowledge in me, inspiring me to be a better person.  I am extremely thankful to Dr. Rajavel Elango for co-supervising this project and helping me sincerely when I needed it the most. I specially thank Dr. Dan Rurak and Dr. Angela Devlin for providing their valuable insights to shape this project. I am very grateful to my lab manager Roger A. Dyer, for tirelessly guiding me through every step in this project, and Dr. Bernd O. Keller for efficiently sharing his vast knowledge in analytical chemistry.  My lab members have been a constant source of support for me and I thank them deeply for standing by me at all times. I would like to thank Dr. Jie Yang for guiding me at each step and stimulating my mind with thought-provoking discussions, and to Dr. Cyrielle Garcia, Alejandra Wiedeman, Sara Moukarzel, Lynda Soberanes, Dr. Kelly Mulder, Guilaine Boyce, Julie Matheson, Brian Wu and Jannette King for helping me with my experiments and teaching me how to work in a team.  I would like to offer my enduring gratitude to my family and friends for believing in me, and giving me the strength and will power to prove my worth.  xvii  Dedication  Dedicated to my parents Hema & Lingi, and all those who have supported me through this wonderful journey  1  CHAPTER 1: INTRODUCTION  Human milk is an exclusive, natural source of nutrition that enables the infant to sustain the vulnerable neonatal period. It is a complex fluid composed of carbohydrates (6.9-7.2% by weight), lipids (3-5% by weight), proteins (0.8-0.9% by weight), minerals (0.2% by weight), vitamins, numerous growth factors and immunological components (Jenness 1979). The predominant carbohydrate in mammalian milk is lactose, a disaccharide of glucose and galactose, and it is a major source of energy. For example in human milk, lactose provides 40% of the total calories (Bruno 1983). The high lactose content of human milk meets the high-energy needs of the developing neonate, particularly the glucose needs of organs, such as the brain, where there is rapid growth. (Dobbing & Sands 1973, Kalhan & Kiliç 1999). My thesis focuses on lactose. Lactose occurs only in mammalian milk, which suggests that this disaccharide is an important nutrient during development. Although it has been known for many decades that lactose is present in milk and is digested and absorbed by infants, the purpose of lactose in milk and its metabolism not completely understood. Other carbohydrate sources are used in human milk substitutes and clinical nutrition solutions to support infants who cannot be fed human milk. Enteral products used to support preterm and term infants may contain CSS (pure glucose), and some contain sucrose. Intravenous nutrition uses dextrose (glucose) as the only carbohydrate source (Hall & Carroll 2000, Bhatia et al 2008, Chawla et al 2008).  2   In contrast to glucose, galactose does not stimulate release of insulin from beta-cells and its extraction by the liver from portal circulation is also much greater (Kliegman et al 1984, Coelho et al 2015). The actions of insulin are well known. In the liver, insulin stimulates glycolysis and lipogenesis (fatty acid and triglyceride synthesis), as well as stimulation of fatty acid uptake and glucose oxidation in skeletal muscle, and fatty acid uptake and storage in adipose (Saltiel & Kahn 2001). In the liver, glucose can also function independent of insulin to stimulate fatty acid synthesis. Studies done several years ago reported that 90% of galactose is taken up by the liver, while 70% of glucose is taken up by non-hepatic organs and cells (Arola & Tamm 1994). This suggests that galactose may contribute to hepatic metabolism and have different metabolic implications than glucose. Notably, infants ingest milk frequently and obtain a sizeable amount of lactose (50% galactose), and therefore may significantly impact postnatal development of carbohydrate metabolism.  Neonatal rat model of feeding The purpose of my thesis is to understand the metabolic effects of different dietary carbohydrates, in a neonatal rat model. I used target compound analysis and targeted metabolomics to characterize metabolite patterns from liver, urine, and plasma, to identify and differentiate metabolites in neonatal rat pups that were fed formula containing lactose compared to CSS by gastrostomy.  Rats are similar to humans in their genes and biochemical pathways (Puiman & Stoll 2008), and are easy to breed and cost-effective, particularly useful for dietary studies (Staley et al 1998). However, rats and humans have certain physiological and metabolic differences. Rodents having a greater basal metabolic rate (Kowalski & 3  Bruce 2014), and glycogen storage than humans (Chandrasekera & Pippin 2013). The lactose concentration in human milk is also 5-10 times higher than in rat milk (Keen et al 1981). In the following sections, I will provide an introduction and current literature on lactose (glucose and galactose) metabolism in humans followed by use of non-lactose carbohydrates in infant nutrition support and Metabolomics for infant nutrition studies. I will then discuss the rationale behind this project, the hypotheses and the objectives.   Milk composition Human milk is the ideal diet for newborn infants and primarily satisfies all the requirements of a neonate. Maternal milk is a complex fluid, with a unique and optimal mixture of constituents essential for the developing infant (Pereira 2014). Breast milk provides a wide range of benefits for infants, which include supporting growth and development, acting as an energy source, and supporting immune function. Human milk is mainly composed of carbohydrates, lipids, and proteins, constituting 6.9-7.2%, 3-5%, and 0.8-0.9% by weight, respectively. It also provides minerals, vitamins, and numerous growth factors and immunological components (Jenness 1979). Human breast milk is a major source of energy, providing an average of about 60-75 Kcal per 100 ml of milk. Carbohydrates (mainly lactose) are an important source of energy in milk (40% of total energy from human milk) with 50% of the energy from lactose coming from galactose (Bruno 1983). Lactose is the predominant 4  carbohydrate found exclusively in milk (Urashima et al 2012), and is further discussed below. The concentration of lactose in rat milk increases progressively from 1.1% to 3.3% in the early suckling period (Kuhn 1972).  Lactose Lactose is synthesized in the mammary glands from uridine diphosphate galactose (UDP-Gal) and glucose by a transglycosylation reaction, catalyzed by a complex of β4-galactosyl transferase and ⍺-lactalbumin (⍺-LA), with the latter being the rate controlling factor in lactose synthesis (Brew et al 1968, Lönnerdal & Lien 2003). The appearance of lactose synthesis in the mammary gland is one of the most striking metabolic changes in lactation and is detected from about one day before parturition (Kuhn & White 1975), mainly influenced by the trigger of ⍺-LA and hormonal changes, such as the fall in progesterone and rise in prolactin (Neville et al 2002, Nicholas & Hartmann 1981). The milk of most eutherian mammals contains lactose as the dominant disaccharide with its synthesis being higher than other carbohydrates found in breast milk (Urashima et al 2012).                   Figure 1. Structural representation of lactose (source: Berg et al 2002) 5   Lactose composition, digestion and absorption  Lactose is a disaccharide of glucose and galactose. Lactose digestion involves the activity of the small intestinal brush-border membrane enzyme lactase-phlorizin hydrolase, also known as lactase (Ingram et al 2009). Lactase hydrolyzes lactose to glucose and galactose, which are then readily absorbed by the small intestine. Absorption involves transfer across the brush border membrane, mediated by the sodium dependent glucose/galactose co-transporter (SGLUT1), and transport across the basolateral membrane into the portal circulation, mediated by insulin-independent glucose transporter, GLUT2 (Wright et al 2007).  After intestinal absorption, glucose and galactose enter the portal circulation. The liver is a key metabolic organ important for governing energy metabolism. The liver mainly functions to maintain whole body glucose homeostasis and converts excess energy (carbohydrate) into fatty acids in the liver (Cahill et al 1959, Rui 2014). Insulin is responsible for integration and moderation of several metabolic responses in key organs, including liver, skeletal muscle, and adipose tissue. Insulin promotes uptake of glucose by adipose tissue and muscle, and stimulates storage of glucose in liver in the form of glycogen, preventing its degradation and entry into the circulatory system (Saltiel & Kahn 2001, Boden et al 2002) Insulin stimulates glycolysis in the liver, muscle and adipose tissue (Cherrington AD et al 2007). In the adipose tissue, insulin inhibits lipolysis and increases the activity of lipoprotein lipase, leading to fatty acid uptake and storage in the adipose tissue (Cassis 2000). In the skeletal muscle, insulin stimulates glycogen synthesis and glucose oxidation, and decreases glycogen breakdown in the liver and muscle (Newsholme & Dimitriadis 6  2001). Insulin also increases protein synthesis in liver, muscle, adipose, and other tissues. It ensures conservation of amino acid and protein reserves by increasing protein synthesis and inhibiting release of amino acids in the liver and muscle, also inhibiting catabolism of branched chain amino acids, leucine, isoleucine, and valine in liver (Lacy 1973). Branched chain amino acids are uniquely catabolized in the muscle where they serve as an energy source, unlike other amino acids that are primarily catabolized in the liver (Felig et al 1977, Brosnan & Brosnan 2006). The action of insulin, though well described in adults, may not always be comparable to neonates. A study has demonstrated differences in insulin concentrations and its association with blood glucose levels in infants when compared to older subjects with neonatal pancreatic insulin secretion being less closely linked to blood glucose concentrations than older children. This study was done on healthy children aged 1 month to ten years, appropriate for gestational age (AGA) term infants and preterm infants. The study depicted that basal insulin secretion (at low blood glucose concentrations) was higher in term infants than in older children. Plasma insulin levels were lower in preterm infants, relative to term infants and children, when controlled for blood glucose concentrations using multiple regression analysis, hence demonstrating that higher plasma insulin levels was not because of high glucose concentration but was probably due to gestational immaturity. The variation in plasma insulin concentrations was also higher in infants than in children (Hawdon et al 1993).  7   Differences between galactose and glucose metabolism  The liver is believed to take up 90% of the circulating galactose, whereas only 30% of circulating glucose is taken up by the liver (Goresky et al 1973, Moore et al 2012). Glucose (70%) is predominantly utilized in non-hepatic tissues, such as brain, red blood cells, and kidney (Moore et al 2012). Over 30 years ago, a study reported that galactose utilization is greater in hepatocytes from 7-14 day old suckling rats than glucose (Rogers et al 1983). Galactose is different from glucose not only in high hepatic clearance, but also in stimulating pancreatic insulin secretion. Galactose does not stimulate insulin secretion by the pancreas (Kliegman et al 1984) and functions independently of insulin. In contrast, glucose metabolism is regulated in part by insulin secretion and can also function independent of insulin in the liver (Newsholme & Dimitriadis 2001, Kahn et al 1994).   Metabolic fate of galactose and glucose  Glucose, galactose, fatty acids, and amino acids are transport into the portal circulation after digestion and absorption, and are taken up by the liver (Rui 2014). Glucose and galactose molecules that enter the liver can be metabolized and utilized in different ways. Glucose is mainly utilized for energy production in the glycolytic pathway, producing intermediates and products for lipid, amino acid and nucleotide synthesis, via the pentose phosphate pathway (PPP) and the tri-carboxylic acid (TCA) cycle. These pathways work to completely oxidize glucose to generate ATP (Dashty 2013). In the fed- state, glucose that is not used for energy production can be stored as glycogen and/or converted to acetyl CoA and used for de novo fatty acid 8  synthesis in the liver (Kunst et al 1989). On the other hand, galactose is readily converted to glucose in the endoplasmic reticulum (ER), with predisposal for release of glucose produced in the ER into the bloodstream (Fehr et al 2005, Cornblath et al 1963). This seems to align with the high glucose demands of infants to support rapidly expanding tissue such as the brain, representing 10% of body weight (Dobbing & Sands 1973), and adipose tissue.   Galactose metabolism After absorption, most of the galactose is taken up by the liver and uniquely metabolized. The metabolism of galactose is highly active at birth and decreases with age (Kliegman & Sparks 1985). The hepatic metabolism of galactose and its role in infants as a major component of dietary energy (representing 20% for humans), however, is poorly understood. Galactose is metabolized via three major pathways in the liver. The Leloir pathway mediates conversion of galactose to galactose-1-phosphate (Gal-1-P) using galactokinase (Holden et al 2003). Gal-1-P can be converted by Gal-1-P uridylyltransferase (GALT), with the uridine diphosphate-glucose (UDPG) coenzyme, to glucose -1-phosphate and used for glycogen or glucose-6-phosphate (Glu-6-P) production and subsequent formation of glucose in the ER (Dashty 2013) (Figure 2).       9                      Figure 2. Leloir pathway (source: Holden et al 2003)             Figure 3. Galactose dehygrogenase pathway and aldose reductase pathways of  galactose  metabolism (Adapted from: Cuatrecasas & Segal 1966, Ficicioglu            et al 2010, Magnusson et al 1988) 10   Several studies have supported the entry of galactose to the pathway of glucose production and its contribution to hepatic glucose concentrations. Previous studies also suggest that galactose is preferentially converted to glycogen, when compared to glucose, in 5-7 day old rat pups (Kunst et al 1989). Thus, it is likely that galactose contributes to glucose production, with storage as glycogen to allow for consistent production of glucose and availability to glucose dependent, non-hepatic tissue such as brain, kidney and red blood cells.   Galactose can also be metabolized to galactonate by galactose dehydrogenase, and can either provide intermediates for the PPP, with production of Xylulose-5-phosphate (Cohn & Segal 1973). Xylulose-5-phosphate and ribose-5-phosphate are isomers. Ribose-5-phosphate can be used for fructose-6-phosphate production, which then enters the glycolytic pathway (Magnusson et al 1988). Experiments done in the 1960s by Cuatrecasas & Segal are one of the few studies that have reported changes in galactose dehydrogenase activity in rat liver, with dramatic increase of activity from five to ten days after birth, after which activity progressively declined (Figure 4). Another route for galactose metabolism is via aldose reductase pathway, with production of galactitol (Figure 3) (Cohn & Segal 1973).    11   Figure 4. Graph representing sudden increase in galactose dehydrogenase activity at  5-10 days after birth, in rat liver (source: Cuatrecasas & Segal 1966).         12   Glucose metabolism  Following hepatic uptake, glucose is phosphorylated to Glu-6-P, which is used for oxidation, and prevents its entry back into the blood stream. In the post-prandial state, Glu-6-P can also be shunted towards glycogen synthesis. Glucose is stored as glycogen and may later be used to maintain blood glucose concentrations (Rui 2014). Glu-6-P can also be metabolized to pyruvate in the cytoplasm, via the non-oxidative pathway, glycolysis. Pyruvate can undergo the oxidative decarboxylation reaction, catalyzed by pyruvate dehydrogenase complex, and hence acts as a link between glycolysis and the TCA cycle with production of oxaloacetate (OAA), which can be transported from the cytoplasm to the mitochondria. The TCA cycle occurs in the mitochondria and is a major producer of ATP. It also supplies acetyl-CoA, a precursor for fatty acid synthesis, stimulated by excess carbohydrate (glucose) intake into the hepatocyte. Glu-6-P also acts as a substrate of the PPP another significant pathway for glucose metabolism. Ribose-5-phosphate produced from PPP is used for nucleotide synthesis. This reaction occurs with the production of NADPH, an important reducing equivalent for fatty acid synthesis (Dashty 2013, Postic et al 2001, Rui 2014).  13    Figure 5. Glucose metabolism in liver, depicting glycolysis, TCA cycle and pentose  phosphate pathways. (source: Berg et al 2002) 14  In the liver, glucose has been reported to regulate mRNA expression of transcription factors such as carbohydrate response element-binding protein (ChREBP), sterol regulatory element-binding protein (SREBP), and enzymes such as stearoyl-CoA desaturase (SCD) (Filhoulaud et al 2013). Recent studies have characterized two isomeric forms of ChREBP: ChREBP-α and ChREBP-β (Filhoulaud et al 2013). ChREBP regulates expression of glycolytic, gluconeogenic, and lipogenic genes, independent of insulin, is activated by glucose and plays an important role in hepatic de novo fatty acid synthesis (Iizuka 2013), (McDevitt et al 2001). SREBP-1c is insulin-dependent and is a major regulator of enzymes promoting lipogenesis in the liver (Xu et al 2013). SCD is an important lipogenic enzyme found in the liver and adipocyte (Yee et al 2013). Glucose is utilized extensively in the brain, with neonates utilizing approximately 8 times more glucose than adults, due to larger increase in brain weight in relation to body weight in infants, when compared to adults (Kalhan & Kiliç 1999). However, though advantageous at optimal levels, excess intake of glucose in newborn infants receiving parenteral nutrition shows increased plasma triglyceride levels and net lipogenesis (Jones et al 1993), and increased hyperglycemia (Ben 2008).    Use of glucose for clinical nutrition support in infants Infant formulas have been used for several centuries as an alternative or supplement to breast-feeding (Fomon 2001). Human milk substitutes, available as infant formula/enteral products may contain cow’s milk protein, soy protein or protein 15  hydrosylates (Agostoni et al 2006, Turck 2007) with the carbohydrate as lactose, or in case of lactose-free products, sucrose, corn syrup solids (glucose), and maltodextrins (Bhatia et al 2008). The addition of glucose in formula increases osmolality and is not recommended for use in infant formula (Koletzko et al 2005, Ben 2008). Intravenous nutrition uses dextrose (glucose) as the only carbohydrate source (Chawla et al 2008). Glucose water is sometimes provided to newborn infants that experience hypoglycemia. (Page-Goetrz 2010). Studies have claimed provision of glucose to be unnecessary leading to interference with normal metabolic mechanisms and breast-feeding a preferred option for neonates with hypoglycemia (Wight et al 2014, Martin-Calama et al 1997).  Glucose-corn syrup solids are widely used in infant formulas for term and pre-term infants, either in combination with lactose or in lactose-free products such as soy protein formulas, with most consumers buying lactose-free products for infants perceived to have lactose intolerance, which may not always be true (Hall & Carroll 2000). CSS is obtained by hydrolysis of cornstarch and are made entirely of glucose residues. The glucose molecules form linear chains, linked by α1, 4 glucoside bonds, with five to 100 residues in each chain. (Lebenthal et al 1983). These corn syrup solids are not high fructose corn syrup solids (HCFS) that are used as sweetening agents in several food items. Alternatively, these glucose-based corn syrups are used extensively in human milk substitutes. An experiment with neonatal pigs fed lactose or CSS showed no differences in body accretion rates of proteins and fats, but a faster growth rate from ten to twenty days of age was observed in the lactose-fed pigs compared to CSS-fed pigs, which suggests that there could be possible metabolic implications at these ages due to carbohydrate source (Oliver et al 16  2002). There however are no studies done that shed light on the metabolic implications of lactose vs CSS feeding.  Understanding the contribution of carbohydrates to metabolism in infants will help formulate and develop recommendations and guidelines for infant formula.   Metabolomics and infant nutrition Appropriate nutrition in early life is indispensable for proper functioning of the metabolic system in the body. Metabolomics has provided invaluable information for better diagnostics in studies concerning infants. Metabolomics is a technology that explains metabolic processes, corresponding to the metabolism in the body. It gives us an understanding about the phenotypic outcome of a cell and aims to provide a comprehensive overview of the physiology of a cell, tissue/bio-fluid, organ or organism, with complete analysis of all metabolites in an organism (metabolome) (Weckwerth & Morgenthal 2005, Zhang et al 2012). Metabolomics is an emerging field focusing on identification of a complete set of metabolites, which are intermediates and by-products of biochemical processes, in a specific biologic system, containing the downstream products of genomic, and proteomic processes. However, due to diversity and vast range of metabolites, it is practically impossible to determine the entire metabolome. Hence, the term ‘Metabolomics’ should be interpreted as an area of study which results in identification of metabolites in a specific sample or specific set of conditions such as environmental changes, genetic modifications etc. (Villas-Bôas et al 2005). Metabolomics sheds light on molecular mechanisms by providing information about potential biomarkers. A biomarker is a 17  measurable characteristic of a biologic process, pathogenic process or pharmacological response to a treatment. Understanding mechanisms of biochemical pathways may involve identification of a single responsible biomarker, or a set of related metabolites combined to form a metabolite pattern for a biologic process (Syggelou et al 2012). In several cases, knowledge about the clinical outcome is more applicable for study and looking at the phenotype of a system, which allows for combination of contributing aspects such as genes, and environmental factors (diet, drugs etc.) may prove to be highly beneficial and more relevant for clinical situations (Zeisel et al 2005). Metabolomics facilitates high-throughput analysis of samples in a timely fashion, and by integrating several constitutive factors, encourages understanding a biological process with an extensive and broader viewpoint. Metabolomics is also proven to be advantageous as it can provide valuable information from extremely small amounts of biological samples. Analysis can also be carried out on bio-fluids such as urine, which is non-invasive and easy to collect, particularly relevant for human infant studies. One such study used urine and serum samples from rhesus monkey and human infants to compare metabolites between the models (O’Sullivan et al 2013).  Nutrimetabolomics is the identification of a cluster of metabolites associated with nutritional variables. Over the years, the consideration of metabolomics as a beneficial approach for infant nutrition studies has helped portray this emerging field in good light. A recent study has looked at the metabolite profile of human milk, with identification and quantification of about sixty-five different metabolites (Smilowitz et al 2013). Metabolomics has been used to study health complications in infants. 18  Infants with nephrouropathies (renal dysplasia, urinary tract infections) were differentiated from healthy children by metabolic profiling of urine (Fanos et al 2013). Metabolomics is usually combined with protein expression studies to obtain a comprehensive understanding of biochemical pathways, underlying mechanisms and the interaction of metabolites and proteins (Cai et al 2010, Barallobre-Barreiro et al 2013, Zhao et al 2015).   Metabolomics workflow Metabolomic analysis usually includes sample preparation and injection into instrument, sample analysis using a favorable analytical technique for separation of compounds in the sample and detection of the fragments, data export and analysis of data based on spectra, statistical analysis, metabolite identification and interpretation of results, utilized for biochemical pathway reconstruction (Figure 6). 19   Figure 6. A typical cycle of metabolomics workflow (source: Mussap et al 2013)   Analytical techniques Metabolomic analysis includes the use of highly complex and well-designed instruments for separation and detection of metabolites, which is associated with software for data analysis, quantification, data storage, and interpretation of results. Mass spectrometry (MS) and nuclear magnetic resonance (NMR) are the two major analytical tools of choice for metabolomics. NMR can be used for analysis of 20  large molecules (>500 Da), sugars, amines etc., but has low sensitivity (Johnson & Gonzalez 2012). MS is highly recommended for metabolomics analysis, owing to its high sensitivity and ability to efficiently detect several compounds in a short period of time. MS, coupled to a separation technique, is essential for effective extraction of metabolites of interest. These typically include MS coupled to gas chromatography (GC), MS with high performance liquid chromatography (HPLC), or less commonly used MS coupled to capillary electrophoresis (CE) (Gowda & Djukovic 2014). Further improvement to mass spectrometry includes coupling of chromatographic methods to mass analyzers, such as Time of Flight (TOF), which provides high spectral quality and extreme accuracy in detection of masses of fragmented ions (Wechwerth & Morgenthal 2005). The basic principle behind MS is the generation of multiple ions from the sample of interest and separation of these constituents based on mass to charge (m/z) ratio. For many analytical techniques, sample preparation is required to facilitate analysis specific to the instrument. Typically, the samples are cells, tissues or urine and plasma.   Ion separation, ion sorting and detection Ionization modes Once a sample is injected into the instrument, the first step to occur is chromatographic separation, which enables separation, identification and quantification of a component in a mixture. The separation is based on solubility of molecules in water and organic solvents, size, and net negative or positive charge. GC works with volatile samples and hence requires conversion of non-volatile 21  compounds prior to analysis, known as derivitization (Ma et al 2011). GC provides higher resolution when compared to other separation methods (Kamel & Prakash 2006). HPLC has the ability to resolve closely related molecules and hence has provides extremely precise peak areas (Bird 1989). The separated components then move through an ion source that transfers analytes to a gaseous phase, imparting them with a charge. The most common ionization methods include Electron ionization or electron impact (EI), Chemical ionization, Electrospray ionization (ESI), and Matrix- Assisted Laser Desorption/Ionization (MALDI). EI is widely used for small (MW<600), neutral, organic molecules that are volatile and stable. EI involves bombarding gas phase molecules with free electrons, hence converting them into ions. The Gas chromatography mass spectrometry (GC-MS) works commonly with EI mode, mainly targeting small, water-soluble molecules (Tsugawa et al 2011) Prior to analysis on the GC-MS, samples undergo silylation or alkylation, that are commonly used derivitization processes, required for analysis of sugars, organic acids, fatty acids, phytohormones, amino acids, hydroxyl acids etc. Silylation gives good volatility and stability of derivatives formed (Ruiz-Matute et al 2011). ESI is a widely used ion source capable of interfacing with LC-MS and works efficiently with polar molecules. ESI, when interfaced with LC-MS, is capable of analyzing a wide range of small and large molecules of various polarities in a biological sample (Ho et al 2003). ESI involves spraying sample solution with a highly charged ESI needle, resulting in formation of charged droplets. Heat gas or dry gas, called desolvation gas, is then applied to cause solvent evaporation, leading to formation of individual gas phase analyte ions. ESI results in little fragmentation of the analyte, as opposed to EI, which 22  results in extensive fragmentation (Pitt 2009). Fragmentation provides structural information and results in reproducible mass spectra. ESI produces singly or multiple-charged ions, depending on size and chemical composition, making it highly selective and sensitive. However, it requires interpretation and mathematical transformation and hence not applicable for non-basic, low polarity compounds (Fenn et al 1989). The ions move to the analyzer from the ion source. The mass analyzer sorts the ions based on their mass-to-charge ratio and resolves the ions into their characteristic mass components. Electromagnetic, quadrupole and ion trap are the most commonly used mass analyzers. The most widespread, tandem mass analyzer, providing high mass accuracy in metabolite identification, is the quadrupole-time of flight (qTOF) instrument (Holcapek et al 2008). The ions are detected by a detector system in the instrument. The detection is based on the abundance of ions emerging from the analyzer and converts this into electrical signals. This results in graphic representation as a mass spectrum, representing relative abundance of ion signals against m/z ratio.   Target compound analysis and targeted metabolomics Typically, metabolomics analysis methods are broadly divided into two groups: untargeted metabolomics and targeted metabolomics. Untargeted metabolomics aims to measure all the analytes in the sample using a certain analytical technique (Roberts et al 2012). Targeted metabolomics measures a defined set of metabolites with characteristic similarities. Targeted metabolomics undertakes a comprehensive view of a vast group of metabolites, which are concentrated to a certain group, based 23  on biochemical and chemical features, and hence can prove to be advantageous for sample preparation and while observing specific biochemical pathways (Dudley et al 2010). Another sub-category of metabolomic analysis is ‘metabolite profiling’. The context in which this is used varies between scientists. Some claim the set of all metabolites or derivative products from a sample to be a ‘metabolite profile’, linking it to untargeted metabolomics (Villas-Bôas et al 2005), whereas other link ‘metabolite profile’ to targeted metabolomics, stating that this is a set of metabolites related to specific biochemical pathways (Dunn & Ellis 2005, Weckwerth & Fiehn 2002, Fiehn 2002). This shift in terminology might be because of transition of metabolomic analysis from a broader, comprehensive approach to a more targeted analysis (Sumner et al 2003). These reviews also discussed another strategy for metabolomic analysis, termed ‘metabolite target analysis’. This results in analysis of one or few known metabolites of interest, which are related to a particular metabolic reaction. In some studies the term ‘target compound analysis’ has also been used synonymous to ‘metabolite target analysis’ (Ligon & May 1984, Wineman & Keto 1994). In this project, my first goal was to identify and differentiate specific, known metabolites of galactose metabolism, including galactonate, galactitol, Gal-1-P, glucose, Glu-6-P, and ribose-5-phosphate between lactose (50% galactose) and CSS (100% glucose) feeding, for which we used the term ‘target compound analysis’. In target compound analysis the sample is run with a standard corresponding to the compound of interest, to get the best estimate of the retention time (RT) of that compound. This is particularly useful for easier identification of a metabolite and faster analysis (Ligon & May 1984, Auray-Blais et al 2014).  24  My next goal was to use ‘targeted metabolomics’ to identify differences in metabolites under the range of small molecular weight, water-soluble compounds between livers collected from lactose-fed and CSS-fed rat pups. This range of metabolites was chosen, as they are relevant to metabolism of the test carbohydrates, which is in turn linked to other metabolic pathways of amino acid and fat metabolism. Identification of differences or variance within metabolites between groups is done with the help of principal component analysis (PCA), which is a statistical tool generally used along with targeted metabolomics (Lai et al 2015, Yamamoto et al 2014, Alexandre-Gouabau et al 2013). The metabolites showing the largest differences can be identified using MS libraries (Schauer et al 2005, Gao et al 2010)   Principal component analysis PCA is a statistical technique that provides the best view of variability in a multi-variate data set. It is a powerful tool for identifying patterns in data and expressing it in such a way as to highlight the similarities and differences. PCA reduces a complex data set into 2D scores plot showing intrinsic patterns in the data set and a 2D loadings plot that highlight the quantities responsible for the patterns (Abdi & Williams 2010). PCA is a type of exploratory data analysis and is used to explore data and formulate hypotheses that could lead to collection of new data (Nyamundanda et al 2010). The principle behind it is to transfer a set of correlated variables into a new set of uncorrelated variables and project it in directions of maximum variability. The first principal component is the direction of greatest 25  variability (covariance) in the data. The second component represents the next highest variation throughout the data. This results in visualizing the groups of a data set in a 2D space in such a way that compounds/ metabolites that do not show much variation between groups will appear together on a PCA scores plot. By displaying groups and metabolites on the same graph, we can understand the contribution of compounds towards differences between groups (Ringńer 2008). A recent metabolomic study of preterm human and formula milk has used PCA in order to depict differences between the groups, based on sample clustering in the scores plot, and identified variables responsible for the clustering using loadings plot (Marincola et al 2015). Recently, PCA has gained popularity to visualize and analyze large amounts of data. PCA has been used for the identification of patients with identical diseases and identified the metabolites responsible for the separation between the groups (Janečková et al 2012). Studies have depicted the importance of PCA in analyzing the effect of diet in metabolism, with identification of metabolites unique to each group. The study investigated the metabolic profile of human breast milk compared to commercial milk (formula milk) and identified metabolites specific to the intestinal contents (Mussap et al 2013). PCA has been efficiently used to extract information from data sets relating to maternal milk evaluation. This resulted in drugs and intermediates contained in human milk (Fanos et al 2012). The distinct gut microbiota of breast-fed and bottle-fed infant rhesus macaques has also been depicted with the help of PCA (Ardeshir et al 2014). PCA hence serves as a beneficial tool and helps in analysis of metabolomics data and efficiently elucidating variations between the groups. In this project, PCA was used to identify distinct 26  biomarker patterns for lactose-fed group compared to the CSS-fed group of rat pups in order to shed light on the implications of feeding lactose or lactose-free carbohydrates to neonates.   Rationale and hypothesis Rationale The World Health Organization recommends breast-feeding from birth to atleast six months of age (WHO 2011). Lactose is the major carbohydrate absorbed and utilized by milk fed infants, accounting for about 40% of dietary energy (Bruno 1983). Lactose provides the neonate with equal amounts of glucose and galactose, and ensures the maintenance of consistent galactose levels during postnatal development. The metabolic effects of galactose in the liver and the pathways of galactose metabolism are less understood. Thus, it is important to specifically address galactose metabolism in the liver and its interaction with other hepatic biochemical reactions.   Currently the available human milk substitutes replace lactose with carbohydrates such as CSS and sucrose for term infants, and dextrose for preterm infants (Bhatia et al 2008, Chawla et al 2008). The implications of feeding neonates with CSS (composed solely of glucose) are not known. Given that galactose and glucose are different in many ways, ranging from their utilization by the liver and their effect on insulin, it is likely that metabolic differences exist while feeding lactose (50% galactose) or CSS (100% glucose) during early infancy.  27  There have been only a few studies, done several years ago, which address the topic of galactose metabolism (Cuatrecasas & Segal 1966, Donnell et al 1967, Cohn & Segal 1973, Rogers et al 1984, Kliegman & Sparks, 1985). Most of these studies investigated the conversion of galactose to glucose, with methods using intravenous monosaccharide administration and radiolabelling. However, understanding lactose and more specifically galactose metabolism in the developing liver, as it relates to breast-fed infants compared to milk-substitute feeding is important. In addition, the interaction of galactose and glucose with metabolic pathways of carbohydrate, fat and amino acid metabolism is essential to highlight implications of feeding non-lactose carbohydrates in the neonatal period. Metabolomics is a promising new technology that helps identify the consensus of metabolites in a biological sample that depicts the measure of the biologic state. MS techniques when combined with PCA enable an efficient statistical tool to understand the variability/altered metabolic pathways in a multi-variate data set (Marincola et al 2015, Janečková et al 2012, Nyamundanda et al 2010). In this study, I used an infant rat model of gastrostomy feeding using Sprague Dawley rats, a well-characterized model for infant dietary studies (Perez-Cano et al 2012, Patel et al 1993). Although human metabolism is not identical to rats, I used this model mainly because of the similarity of rats in terms of biochemical pathways to humans (Puiman & Stoll 2008), and ease of breeding and performance of gastrostomy experiments (Staley et al 1998,). My thesis focuses on development and application of metabolomic methods for high throughput analysis of metabolites and to characterize metabolite patterns 28  obtained from feeding lactose or non-lactose (solely glucose) carbohydrates during the neonatal period. This is the first study to characterize the metabolites from using an animal (rat) model of infant feeding, and will guide the implications of feeding different carbohydrate sources and nutritional intake in neonates.   Hypotheses I hypothesize that:  Galactose and metabolites specific to galactose metabolism, which mainly include galactonate, galactitol and Gal-1-P, and other metabolites that maybe linked to galactose metabolism such as glucose, Glu-6-P, and ribose-5-phosphate will be differentiated in liver, plasma and urine of lactose-fed rat pups when compared to liver, plasma and urine of CSS-fed rat pups, by target compound analysis.   Small molecular weight, water-soluble metabolites that differentiate liver from pups fed lactose compared to liver from pups fed CSS will be identified using targeted metabolomics, with the help of PCA and MS libraries, and these metabolites will be mapped to biochemical pathways of carbohydrate, amino acid and fat metabolism.      29   Objective and specific aims Objective 1.        To use a neonatal rat model to show metabolomics as a beneficial tool, and to                    highlight and differentiate metabolic effects of feeding lactose compared to              glucose-corn syrup solids (CSS) to rat pups.  2.       To identify metabolite patterns to show that lactose and CSS feeding    differ in their impact on hepatic metabolism.  Specific aims 1.       To feed rat pups by gastrostomy from 5 days of age with identical milk formula,              which differ only in the carbohydrate source: lactose or CSS, and to determine              the optimal length of feeding (day 10, 14, vs 18) to identify differences             between the groups.  2. To use target compound analysis by gas chromatography mass spectrometry (GC-MS) and negative-mode liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry (LC-qTOF-ESI-MS) of liver, plasma and urine samples to show differences in specific, targeted metabolites of galactose metabolism due to feeding lactose (which consists of 50% galactose) compared to CSS (which is 100% glucose) as the carbohydrate source in rat pups. The metabolites analyzed include galactose and those that are directly involved in galactose metabolism, such as galactonate, galactitol 30  and Gal-1-P, and few other metabolites that maybe linked to galactose metabolism such as glucose, Glu-6-P and ribose-5-phosphate.  3.        To use targeted metabolomics by GC-MS analysis of liver samples with              PCA to identify metabolites in the range of small molecular weight, water-             soluble molecules that differentiate rat pups  fed  lactose compared to CSS.             The RT values of metabolites that cause variance between the groups will be             identified by MS libraries and these will be mapped to pathways of glucose,              galactose, amino acid and fatty acid metabolism.  4. To use western blotting to determine if there is lower expression of key  enzymes and transcription factors of liver lipogenesis and glycolysis, including carbohydrate response element binding-protein (ChREBP), sterol regulatory element-binding protein (SREBP1-c) and stearoyl CoA desaturase (SCD) in lactose fed rat pups compared to CSS fed rat pups.     31  CHAPTER 2: STUDY DESIGN AND METHODS  Experimental design    *Note: Day 18 samples were collected and tested for protein expression by western blotting analysis, which showed no difference. Also, these animals are closer in terms of growth and maturity to a rat that is gradually progressing towards consumption of solid food from exclusive milk feeding. Thus, we did not continue with metabolomic analysis.  Neonatal rats Day 5: initiate feeding Intervention group: Carbohydrate source 32  This project used an infant rat model of feeding with two groups, lactose-fed group and CSS-fed group. The rat pups were fed with rat milk formula that had the same constituents other than the carbohydrate source, which was either lactose or CSS. Feeding was initiated five days after birth, represented as day 5. The different time points at which samples have been collected is represented as day 10, which implies samples collected from 10 day old rats fed for a time period of 5 days, day 14, which implies samples collected from 14 day old rats fed for a time period of 9 days, and day 18, which implies samples collected from 18 day old rats fed for a time period of 13 days.  Metabolomic analysis was performed on the samples collected. This includes two kinds of approaches. Target compound analysis was done on liver, urine and plasma samples collected at day 10 and day 14, using GC-MS and LC-qTOF-ESI-MS to identify differences in galactose and specific, known metabolites that are relevant to galactose metabolism, between the lactose-fed and CSS-fed group of rat pups. The metabolites that we tested include galactose, galactonate, galactitol and Gal-1-P and other metabolites such as glucose, Glu-6-P, and ribose-5-phosphate. Targeted metabolomics was performed on liver samples collected from day 10 and day 14 to identify metabolites in the range of small molecular weight, water-soluble compounds that differentiate the lactose-fed rat pups to the CSS-fed rat pups. This range was selected because we wanted to focus on metabolites that are relevant to galactose and glucose metabolism. These metabolites were identified with the help of PCA and 33  MS libraries and then mapped to biochemical pathways of galactose, glucose, amino acid and fat metabolism.   Western Blotting was performed on liver samples collected at day10, day 14 and day 18 to quantify protein expression of key enzymes and transcription factors of liver lipogenesis and glycolysis, ChREBP, SREBP1-c and SCD.     Neonatal rat model of feeding   Rat milk formula composition and  preparation This study uses lactose or CSS as the carbohydrate source in the formula. The composition of energy and all other components is identical and is comparable to published neonatal rat milk composition (Auestad et al 1989). Macronutrient composition of neonatal rat milk composition is depicted in Table 1 and macronutrient composition of rat milk formula prepared is depicted in Table 2.         34  Source % Energy content Kcal/g g/L Kcal/L Carbohydrates 4 4 11.3 45.2 Protein 24 4 69 276 Lipids 72 9 93 837  Table 1. Macronutrient composition of neonatal rat milk. (values are approximate)  Source % Energy content Kcal/g g/L Kcal/L Carbohydrates 12.1 4 60 240 Protein 23.7 4 117 468 Lipids 64 9 140 1260  Table 2. Macronutrient composition of rat milk formula. (values are approximate)  The pre-milk was prepared using casein, whey, basic buffer and water. The mixture was stirred using a magnetic stirrer for 2 hours at 40°C. After stirring, the formula mixture was tested for consistency by passing through the tubing of the pumps that are used for feeding rat pups. It was ensured that the formula formed no clumps and the fluidity of the formula was consistent. It is important to perform this step as it enables ease in the production of final formula, which is free of clumps and homogenized. After this, the mineral components were added to the mixture. To 35  improve homogeneity of the formula mixture, a mineral mix (#AIN93G Teklad) and vitamin mix (#40060 Teklad) were used, instead of adding mineral and vitamin components individually. The amino acid mixture consisting of arginine, glycine, taurine and picoline was then added to the formula and stirred. The formula mixture was separated into two equal parts by volume. 60 grams of lactose was dissolved in 240ml of boiling water and added for each liter of formula mixture. To the other half, the same amount of CSS was added instead of lactose. After mixing thoroughly, a fat mix was prepared and added to the formula. The entire mixture was stirred using a manual stirrer to increase homogeneity of the mixture. 140 grams of fat mix was added to 860 grams of milk formula to make a liter. The fat mix composition was optimized so as to improve fluidity and solubility of fat in milk formula. The milk formula with the fat mix was stirred well for 1 hour; 120ml was aliquoted into plastic cups and frozen at -20 °C for further use. The rat milk formula was thawed at room temperature and stirred well prior to use. The nutrient composition of the milk formula and the amount of each component added is depicted in Appendix 1.   Gastrostomy experiment  Rearing of rat pups Sprague Dawley rats were used for this project. In each experiment, rat pups from 3-4 litters of rats (each litter size about 7-9) were randomly divided into two groups of carbohydrate-lactose or CSS, with both groups having equal number of rat pups from each litter (sex of the rat pups was not determined). Each experiment typically started with approximately 20-25 rat pups at day 5 after birth, with a mortality 36  rate of 30-40% during the course of the gastrostomy experiment.  For gastrostomy feeding, rat pups were fed by gastric tube and kept in a temperature-controlled water bath, with a floating container for each animal. The rat pups were cannulated on day 5 after birth, using methods established in the Innis lab and as previously described (Moore et al 1990, Patel et al 1993). They were anaesthetized for a short duration using halothane as the necessity of anaesthetizing for giving the gastric catheter. Further details follow. A 7cm, PE-50 silica tubing, with a wire inside it, was used to enable the cannulation for feeding. The wire, along with the tubing, was inserted into the mouth, moved through the esophagus, and then pulled out of the stomach. To avoid potential esophagus scratching with wire, the silica based PE-50 tubing was covered in Muko©, a mucous-like substance often used as a lubricant. Once a small hole was made to enable external exposure to the stomach, the 7cm PE-50 tubing was removed from the mouth and esophagus, with the PE-50 replaced with PE-10 silica tubing, with a plastic circle (diameter greater than the PE-10 tubing) at its end. The plastic circle, attached to the PE-10 tubing, remained fitted in the stomach, then used for feeding via gastric tubes attached to a pump. Increased risk of infection due to “surgery” was carefully avoided and checked for. The rat pups were monitored every hour and stimulated every 3 hours. They were weighed every time they were stimulated (every 3 hours). The tubings were flushed with warm water and filled with formula, thrice a day to remove any clumped or precipitated milk formula and to allow easy flow of formula to the infant rat. They were checked for signs of stomach bloating, redness at the catheter exit, skin color 37  and dehydration. Rat pups were injected with Pedialyte through their cannulas, three times a day to prevent dehydration and allow maintenance of electrolyte balance. An extra amount was provided if signs of dehydration, such as wrinkly and dry skin, were observed. If dehydration occurred, 0.1% of NaCl was provided subcutaneously and monitored for response. Pups facing issues, such as excessive stomach bloating, gasping due to lack of oxygen, reduction in body temperature, cannulas pulled out, were monitored closely and euthanized if necessary. The flow rate (µl/min) and volume of milk (µl) was adjusted every day in the morning and at night, with a flow rate of 25% average body weight, increasing as the pup grew. The calibration of pumps was checked every day and re-calibrated if any error in volume greater than 5% occurred. The pumps were adjusted to provide one cycle of feeding for approximately ten mins, followed by non-feeding for 20 mins throughout the day.     Tissue and biofluid collection The tissue and biofluids were collected from rat pups at 10, 14 and 18 days after birth with isoflurane as the anesthetic. The pups were weighed, and the animal and litter number were recorded.  For blood, samples were collected via cardiac puncture using a 25G needle and 1ml syringe. The blood was centrifuged at 10000 rpm, 4°C for 15 mins, to separate plasma and red blood cells (RBCs). Ten microlitre and 75µl aliquots of plasma were collected then immediately placed in liquid nitrogen before transfer to the -80°C freezer. Urine was collected into a 2 ml cryovial, using a 30G needle and 38  1ml syringe, then placed in liquid nitrogen, and stored in a -80°C freezer. For liver, the method for collection was slightly modified to enable Western Blotting in addition to metabolomics in this study. The liver tissue was collected by slicing 2 pieces, weighing 100 mg each, and placed in separate cryovials. These were frozen immediately and used for protein expression analysis using Western Blot and metabolomic analysis using GC-MS and LC-qTOF-ESI-MS. The method of liver collection considered the complexity of the liver and heterogeneity of the liver lobes, with non-hepatocyte cells constituting 40% of the total liver cells. Due to limited knowledge about different liver lobes and their development in rat pups, consistent liver samples were collected based on information from published work (Malarkey et al 2005, Faa et al 1994). The left lobe of the liver was identified (Kogure et al 1999) and divided along the vertical axis into two equal parts. The left half was weighed and used for western blotting and the right half was used for metabolomic analysis. The samples were placed in cryovials, quickly frozen using liquid nitrogen then transferred to -80°C. The rest of the liver (right and median lobes) was placed into another cryovial and frozen.   Several other organs were collected, although not utilized for this project. These included brain, and intestinal ileum, jejenum, ileum, colon and cecum. For brain, it was quickly collected, the cerebellum discarded, and the left and right sides of the cortex weighed, placed in separate cryovials, then frozen at -80°C. Intestine: The colon and the cecum were obtained, feces removed from the colon and collected into a cryovial. Sections of the jejunum and ileum were placed in separate cryovials and frozen in liquid nitrogen, before moving to -80°C.  39    Metabolomic analysis Target compound analysis and targeted metabolomics was used to analyze collected samples at day 10 and day 14.   Target compound analysis Target compound analysis was performed using GC-MS and LC-qTOF-ESI-MS for liver, plasma and urine samples.  Metabolites that were selected for target compound analysis include galactose and those directly linked to galactose metabolism, galactonate, galactitol and Gal-1-P. Other metabolites relevant to galactose metabolism were also selected for comparison. These metabolites are mentioned in Table 3.  The samples were run along with corresponding standards specific to the metabolite tested, to get the best estimate of RT and enable accurate identification of the mass spectra fragmentation pattern corresponding to these metabolites. Liver, plasma and urine samples were run on the GC-MS and the LC-qTOF-ESI-MS. Metabolites that could not be analyzed using the GC-MS were tested on the LC-qTOF-ESI-MS. The LC-qTOF-ESI-MS was used in negative mode, mainly to measure and compare carbohydrate phosphates and other acidic compounds. GC-MS analysis requires derivitization of samples to ensure that the samples are volatile.      40  Metabolite Retention time in mins (approx. values)* m/z value of target ion GC-MS Galactose Galactonate Galactitol Glucose LC-qTOF-ESI-MS Galactose1-phosphate Glucose-6-phosphate Ribose-5-phosphate  19.30 &19.68+ 20.63 20.02 19.41 & 19.50+  12.2 10.9 12.3  - - - -  259.0224 259.0224 229.0119  Table 3.  List of metabolites analyzed for target compound analysis. * Retention time values are approximate and differ within a window of  ~0.25 mins     with time of injection into the instrument.   +Two retention times representing different elution times of syn and anti     isomers (discussed in results section 3.2).             41   Preparation of liver, plasma and urine samples Gas chromatography mass spectrometry (GC-MS)  For GC-MS, liver samples were prepared using 30-40 mg of tissue with 9 volumes of distilled water. This was sonicated at low pressure to homogenize. A 100 µl aliquot of the homogenized tissue was taken, and then 100 µl of distilled water and 800 µl of methanol added, with the sample kept on ice for 60 mins. The samples were centrifuged at 14,000 rpm in a refrigerated Eppendorf centrifuge at 4°C for 15 mins, and the supernatant removed to a clean 1.7ml Eppendorf tube. The supernatant was dried down using a Speedvac, typically taking about 2 hours. Precautions were taken to avoid sample over-drying as this may lead to temperature sensitive compound loss. To derivitize, 100 µl methoxylamine HCl in pyridine 20 mg/ml was used and the samples were kept at 40°C for 2 hours. Methoxylamine derivatises ketone groups, forming an oxime (CH3ON) (oximation). Short sonication was done using low power for approx. 2-3 sec to help dissolve all material.  Speedvac was then used to completely remove the pyridine, avoiding incomplete drying, which impairs GC-MS chromatography. A second derivitization used 75 µl of N-Methyl-N- (Trimethyl silyl) trifluoroacetamide (MSTFA) for one hour at 40°C to derivitize for OH and NH groups, substituting it with a trimethylsilyl group  (-O-Si (CH3)3) (silylation). The samples were centrifuged at 4°C for 15 mins, the supernatant removed to a 250 µl insert and GC autovial, 75 µl of hexane added; 1 µl of the sample was then injected into the GC-MS.   42  Plasma samples were prepared for GC-MS by aliquoting 100 µl of plasma into a 1.7ml Eppendorf tube, then adding 100 µl of distilled water and 800 µl of methanol. This was left on ice for 60 mins, and then centrifuged at 14,000 rpm in a refrigerated Eppendorf centrifuge at 4°C for 15 mins. The supernatant was removed to a clean 1.7ml Eppendorf tube, and then dried down using a Speedvac for about 2 hours, as above, avoiding potential loss of temperature sensitive compounds. The first step of derivitization used 100 µl methoxylamine HCl in pyridine 20 mg/ml with the samples kept at 40°C for 2 hours. Methoxylamine derivatises ketone groups forming an oxime (CH3ON). A short sonication was done using low power for approx. 2-3 sec to help dissolve all material.  Speedvac was then used to completely remove the pyridine. The second derivitization step used 75 µl of N-Methyl-N- (Trimethyl silyl) trifluoroacetamide (MSTFA) for one hour at 40°C, to derivitize for OH and NH groups. The samples were centrifuged at 4°C for 15 mins, the supernatant removed into 250 µl insert and GC autovial, 75 µl of hexane added and 1 µl sample injected into the GC-MS.  Due to small amount of urine collected, we ran the urine samples and tested them for all the metabolites only on the LC-qTOF-ESI-MS.  Samples were tested for galactose, galactonate, galactitol and glucose on the GC-MS. Galactose, galactonic acid (corresponding to galactonate), dulcitol (corresponding to galactitol) and glucose standards were derivitized and similarly injected into the GC-MS for relative quantitative comparison. 43  Standard for Gal-1-P was first run on the GC-MS. We were not able to obtain a clear chromatogram to identify the RT and hence we tried identifying Gal-1-P on the LC-qTOF-ESI-MS.   Because of variation in RT at different injections, the standards were run along with samples in each set of sample analysis to ensure accuracy in comparison of RT and mass spectra.  Liquid Chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry (LC-qTOF-ESI-MS)  For LC-qTOF-ESI-MS, liver samples were prepared using 50 mg of tissue and 3 volumes of water. The homogenate was centrifuged at 14,000 rpm at 4°C for 30 mins, using Amicon® ULTRA, 0.5ml centrifugal filters to concentrate the liver homogenate and filter out impurities. 25 µl of the filtered homogenate was added to an LC autovial and 5 µl was injected into the LC-qTOF.  For plasma samples, 50 µl of plasma was placed in a 1.7ml Eppendorf tube and to this equal volume of acetonitrile was added for protein precipitation. This tube was then centrifuged at 14,000 rpm at 4°C for 15 mins. 25 µl of the supernatant was added to an LC autovial and 5 µl was injected into the LC-qTOF-ESI-MS. For urine samples, 50 µl of urine was placed in a 1.7ml Eppendorf tube and centrifuged at 14,000 rpm at 4°C for 15 mins, to remove impurities. 25 µl of the supernatant was added to an LC autovial and 5 µl was injected into the LC-qTOF-ESI-MS. 44   Samples were tested for metabolites (refer table 3) on the LC-qTOF-ESI-MS. Standards for these compounds were run along with the samples.   Instruments and analysis Gas chromatography mass spectrometry (GC-MS) A Quattro micro™-GC, Agilent 6890N Network GC system was used for GC-MS analysis. A column of length 30m (Agilent DB5MS) was used along with Helium carrier gas. The instrument scanned from 75 to 600 MW and identified compounds based on mass spectra and RTs. Data was collected via Electron Impact mode (EI), which generates electrons via a filament and causes fragmentation of compounds in the sample. The mass spectrum is generated based on this fragmentation (fragmentation pattern) and gives intensity vs m/z plot, representing distribution of ions by mass. The spectrometry data is also represented as a chromatogram with intensity vs RT. The RT is the interval between the injection and detection of the component and varies with the identity of the compound. Prior to acquisition of data, the system needs to be set up. A sample list was created depicting information about the samples available for analysis, including sample file name, bottle number for automated analysis and inject volume, kept at a default volume of 1 µl. MassLynx v4.1 software was used for viewing and processing chromatographic and spectrometric data. Background subtraction and smoothing can be used to reduce background noise by increasing signal-to-noise ratio and hence improves presentation and aids in interpretation of the chromatogram. The chromatograms are 45  then processed for highlighting areas of interest by integration of peaks. This results in location of peaks, positioning baselines and calculating both heights and areas of peaks above the baselines. This is necessary for quantitative analysis and comparison between groups. For target compound analysis, the RT values from the chromatogram and mass spectrum fragmentation patterns were compared with standards to identify the metabolite of interest and the peaks were integrated for comparison of groups and quantification.   Liquid Chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry (LC-qTOF-ESI-MS) For this project, LC-qTOF-ESI-MS analysis was based on negative mode electrospray ionization mass spectrometry employing a hybrid qTOF mass spectrometer after separation by high performance liquid chromatography (HPLC) using an ion-pairing reagent (tributylamine, TBA) in the mobile phase. The LC method has been adopted and modified from previous studies (Luo et al 2007, Lu et al 2010). The method was used in negative mode to measure and do semi-quantitative comparisons of carbohydrate phosphates and other acidic compounds (negatively charged ions) relevant to energy metabolism in a variety of biological samples, including cell extracts, tissue extracts (e.g. liver), plasma, urine and other. An Agilent HP1290™ Infinity HPLC (Palo Alto, CA, USA) was connected to Agilent 6530™ accurate mass q-TOF LC/MS (Palo, Alto, CA, USA) and was used for mass spectrometric analysis. An analytical reversed-phase column, Agilent Zorbax Eclipse AAA (3.5µm particle size), with an internal diameter of 3.0mm and length of 150mm, 46  was utilized. A binary mobile phase was used for gradient generation, with Solvent A- 10mM TBA and12.5mM acetic acid in de-ionized water with 1% acetonitrile and Solvent B- 100% acetonitrile. For gradient elution Solvent A was kept at 99% for 10 mins, followed by a linear increase of Solvent B to 30% over the next 15 mins, then Solvent B was increased to 99% over the next five mins and kept at 99% for another 5 mins, then decreased to the original 1% within the next 3 mins and kept at 1% for the remaining 7 mins. The total run time was 45 mins. The flow rate was kept at 0.45 ml/min and column temperature was maintained at 30 °C and data acquisition was done only in MS mode, which measures all ions at a given time for the chosen m/z range. The mass spectra acquired were in the mass range of 65 m/z and 850 m/z. The q-TOF is a high-resolution instrument and provides accurate mass values, up to 4 decimal places for compounds less than 1000 Da. The instrument generates a Total Ion chromatogram (TIC) and with the use of software we obtained the Extracted Ion Chromatogram (EIC). The extracted chromatogram is based on the pseudomolecular ion mass, corresponding to highly accurate mass values. The number of compounds specific to this mass is lower and hence it is easier to identify individual compounds based on their masses. The Tributylammonium acetate in the mobile phase enables the separation of these compounds as TBA forms ion pairs with the compounds and the ion-pair complexes separate on the reversed-phase HPLC-column. Samples from different groups were analyzed randomly mixed at the same day to avoid any artifacts by potential instrument changes. Peak areas of eluting compounds were used for relative semi-quantitative comparison between groups. 47    Targeted metabolomics Targeted metabolomics was performed using GC-MS for liver samples. The same samples prepared for target compound analysis were used for targeted metabolomics, excluding the standards. (Refer section 2.3.1.1). Targeted metabolomics was used to identify metabolites in the range of small molecular weight, water-soluble molecules that differentiate rat pups fed lactose compared to CSS. Chromatograms from the samples run on the GC-MS were obtained and sent to MarkerLynx v4.1 software for PCA (discussed in detail in following section). PCA resulted in identification of RTs of metabolites that contribute to variance between the groups. These values were noted and corresponding metabolites were identified using MS libraries, and then compared with standards to ensure accurate identification. The peaks were then integrated to obtain quantitative area values for the peaks to enable comparison between groups, and then were tested for statistical significance using Independent samples t-test.   Principal component analysis MarkerLynx™ software was used to perform post acquisition processing of samples using specified parameters. It identifies mass-retention pairs, or markers, present in at least two of the samples, through a combination of spectral deconvolution, peak integration and sample alignment. It determines the abundance of each marker in terms of height and area, across all samples and then submits the 48  marker abundance matrix to PCA. PCA is a statistical technique that provides the best view of variability in a multivariate data set. PCA identifies inherent group clustering and highlights markers responsible for this clustering. It helps identify principal components and highlights variation among groups considered.  MarkerLynx™ extracted data from the raw file, obtained from GC-MS analysis. The software requires a method to process the sample list. The processing parameters for MarkerLynx™ were specified. These included Mass tolerance kept at 0.20 Da, which specifies the mass accuracy and the peak width at 5% height at 1.00 second. The peak-to peak baseline noise was kept at 5000.00, which gives an estimate of the baseline noise between the peaks on a typical extracted ion chromatogram. Other parameters included Mass window =0.30, the value in which spectral peaks are considered to be the same marker, RT window =0.25, the RT in which spectral peaks are considered to be the same marker and Noise elimination level was kept at 2.00, specifying the degree to which background noise is eliminated.  MarkerLynx™ detects the chromatographic peaks, collects potential markers and performs PCA. The PCA scores plot allows us to see natural clustering and relationships among samples. The loadings view plot helps determine the markers that are responsible for clustering and allows us to understand markers that may contribute to variability between the groups. Using the PCA loading view plot, the mass values of the metabolites of interest were identified. The metabolites furthest from the origin of the plot are the ones that contribute most to the variance. These metabolites were selected and their RT values were noted from the PCA list of metabolites. Chromatographic peaks of these markers were looked at individually. 49  The spectrum for each peak was obtained using MassLynx™ V4.1 software and the metabolite was identified using MS libraries (NIST MS Search 2.0). The areas of chromatographic peaks of the metabolites were obtained for all the samples from each group by integrating the peaks and Independent samples t-test was applied to show statistical difference.   Western blotting Liver was lysed using lysis buffer (50mM Tris pH 7-8, 150mM NaCl, 0.1% SDS, 0.5% sodium deoxycholate, 0.1% Triton X 100) with phenylmethylsulfonyl fluoride (PMSF) and NaF as a protease inhibitors. The liver samples were quantified for protein using BSA protein assay kit and 40 µg of total protein was run on the gel. The samples were resolved using denaturing SDS-PAGE at 70V for 30 mins and 100V for 90 mins. The gel was then transferred to an Immuno-Blot® PVDF membrane at 100 V for 60 mins, and blocked using milk. The membrane was incubated overnight with a primary antibody specific to the target (diluted with 5% BSA to 1:100), which include goat polyclonal antibody (sc21189, Santa Cruz Biotechnology) for ChREBP, mouse polyclonal antibody (sc13551, Santa Cruz Biotechnology) for SREBP1-c, and rabbit polyclonal antibody (sc30081, Santa Cruz Biotechnology) for SCD. This is followed by incubation with alkaline phosphatase conjugated secondary antibody, which include donkey anti-goat IgG (sc2022, Santa Cruz Biotechnology) for ChREBP, goat anti-mouse IgG (sc2008, Santa Cruz Biotechnology) for SREBP1-c, and goat anti-rabbit IgG (sc2007, Santa Cruz Biotechnology) for SCD. Relative 50  intensity of the bands was visualized by chemiluminescence using Amersham™ CDP-Star™ detection reagent. β-actin was used as internal control   Statistical analysis Normality of data distribution was assessed using the Kolmogorov-Smirnov test. The areas of chromatographic peaks of samples from lactose-fed and CSS-fed group were analyzed using descriptive statistics. Values for all analytical data are expressed as mean  standard deviation. Independent samples t-test was used for comparison between the two groups. All statistical analyses were performed using IBM SPSS statistics software (IBM SPSS Statistics for Windows, Version 22.0, Chicago, IL, USA: SPSS Inc.), using a two-sided model and p values < 0.05 as statistically significant.  51  CHAPTER 3: RESULTS    Methodology development Aim 1: To feed rat pups from 5 days of age with identical milk formula, which differ only in the carbohydrate source: lactose or CSS; and to determine the optimal length of feeding (day 10, 14, vs 18) to identify differences between the groups.  The initial phase of this study was setting up of the gastrostomy experiment for feeding 5-day old rat pups with milk formula differing only in the carbohydrate source. Due to the novelty of this project, a bulk of time was spent on designing optimum conditions for rearing of rat pups, as well as optimizing and developing suitable rat milk formula for feeding. During the course of the study, the protocol for performing cannulation on rat pups and conditions for rearing them were improved, hence resulting in an increased survival rate. Coating the wire used for cannulation with Muko© enabled ease of insertion of tubing into the stomach and reduced rate of loss of animals during “surgery”. Further, to ensure healthy growth of rat pups, the conditions for rearing were refined, with extensive and detailed care provided throughout the course of the experiment. The rat pups were kept under the bedding, at the bottom of the cups, in order to keep them warmer, and were weighed every time they were stimulated and fed formula via syringes if they had lost weight, the pump calibration was checked twice a day and recorded in a notebook, the rat pups were rotated around the water bath so that they get fed from different tubing each time they are removed for stimulating or weighing, and the pump tubing was checked 52  for problems with formula flow and disposed if tubing appeared to be damaged or crushed.  An experiment where rat pups are fed milk formula, kept similar in all other components other than carbohydrate source (lactose or CSS), has never been done before. Hence, it was important to carefully consider the composition of individual components, including carbohydrate, protein, fat, mineral and vitamin content, in the rat milk formula mixture. Thus far, there have been no published works that indicate the use of CSS as the sole carbohydrate in rat milk substitutes used for gastrostomy experiments. In this study, we prepared a rat milk formula with only CSS as the carbohydrate source. This study is novel for the successful use of CSS as the sole carbohydrate in feeding infant without any major issues. The carbohydrate content was increased from 35g/l in previous formula used, to 60g/l, which is 6.9% energy content to 12.1% energy content, fed every 10 mins each day, which is different from human babies fed only every 3 hours. Hence, though it is difficult to compare the feeding nature of rat pups and human infants, the carbohydrate content was increased to slightly match human physiological levels, but also keeping energy content similar to rat’s milk (4% energy content from carbohydrate source).   Studies in our first experiments, particularly at 18 days of age, encountered problems during the course of the gastrostomy due to heavy formula clumping. The formula preparation included addition of mineral components individually as NaCl, CuSO4, ZnSO4, MgSO4, CaCO3, CaGluconate, FeCl3, KI, NaF, AlSO4, and MnSO4. Modification to a homogenized mix of minerals with better-optimized individual components was developed to prevent agglutination of formula and enable healthy 53  growth of rat pups. The survival rate of the rat pups during the course of the gastrostomy experiment was 60-70%. The growth rate of rat pups has been depicted in Figure 7. The figure depicts mean weight of rat pups, with 11 rat pups used to calculate average body weight of mother-reared pups, at all days. For lactose and CSS-fed pups, average body weight was recorded at different days using weights of the rat pups from which samples were collected, which is day 5 (n=36), day 6 (n=36), day 7 (n=36), day 8 (n=36), day 9 (n=36), day 10 (n=36), day 11 (n=24), day 12 (n=24), day 13 (n=24), day 14 (n=24), day 15 (n=12), day 16 (n=12), day 17 (n=12), day 18 (n=12). The growth of the rat pups that were artificially reared was comparable to mother-reared pups, and growth rate of rat pups with different carbohydrate feeding by gastrostomy (lactose or CSS) was also comparable. 54    Figure 7. Rat pup growth curve Comparison of average body growth rate of rat pups from 5-18 days after  birth, between mother-reared (filled circle, straight line) and artificially reared  pups (lactose feeding (filled squares, dashed line) and CSS feeding; glucose   corn syrup solids (empty triangle, dotted line)). Data presented as mean ± SD   05101520253035405 6 7 8 9 10 11 12 13 14 15 16 17 18Average body weight (g)Age (days postnatal)Rat pup growth curvemotherrearedlactoseCSS55  Details on the average body weight and liver weight for mother-reared, and lactose or CSS-feeding, at different time points is given in Table 4. The growth of the liver was proportional to the body weight, and was similar between mother-reared and artificially reared rat pups, with carbohydrate source provided by gastrostomy (lactose or CSS) not affecting liver growth rate.  Age of rat pups after birth Group Day 5  Day 10  Day 14  Day 18  Mother-reared (n=11) Body weight (g) (Mean ± SD) 10.9±1.1 19.8±1.2 27.7±1.9 35.9±2.4 Liver weight (g) (Mean ± SD) 0.34±0.08 0.69±0.95 1.25±0.15 1.45±0.09 Lactose  (n=13, day 5) (n=6, day10,14,18) Body weight (g) (Mean ± SD) 12.4±0.09 18.4±0.9 24.5±1.2 33.5±2.2 Liver weight (g) (Mean ± SD) - 0.58±0.06 1.1±0.12 1.41±0.13 CSS (n=13, day 5) (n=6, day10,14,18) Body weight (g) (Mean ± SD) 12.8±1.03 18.6±0.81 25.5±1.24 36.1±1.9 Liver weight (g) (Mean ± SD) - 0.66±0.10 1.23±0.12 1.47±0.12 Table 4. Mean body weight (g) and liver weight data (g) of mother-reared and lactose or CSS-fed group of rat pups, at days 5, 10, 14 and 18 after birth.    Data presented as mean ± SD. CSS, glucose-corn syrup solids; n=sample size.    56   Metabolomic analysis:  The liver samples gathered from the following experiments at day 10 and day 14 were from rat pups fed optimized formula with homogenized minerals (Mineral Teklad mix). Day 18 samples were not analyzed for target compound analysis and metabolomics based on data from Western Blotting (see section below). At day 10 and day 14, analysis showed that the key metabolite of liver galactose metabolism, galactonate was significantly different between the two groups (p < 0.001). At day 10, targeted metabolomics of liver revealed that metabolites including D-ribose, pyrimidine, malate, glycine, leucine and isoleucine were significantly different between the groups. But at day14, only leucine and isoleucine were significantly different between the groups. (Data presented in section 3.2 and 3.3) Western Blotting:  Western blot analysis of liver samples for ChREBP, SREBP1-c and SCD showed no differences between the lactose compared to the CSS-fed group at day 10, 14 and 18 (Data presented in Appendix B). Because there were no differences at day 18, I hypothesized that this was because these animals were closer in terms of growth and maturity to a rat that is gradually progressing towards consumption of solid food from exclusive milk feeding, and decided to only look at day 10 and 14 for all other analysis. The samples next tested were collected at day 10. Western blotting showed no difference in expression of ChREBP, SREBP1-c and SCD in day 10 liver samples. However, I continued with metabolomic analysis to understand and validate 57  work done in previous studies (Cuatrecasas & Segal 1966), which depicts the sudden increase in galactose dehydrogenase enzyme at day 5-10. In addition, the collection of liver samples from day 18 and day 10 did not consider the different regions of the liver, and how liver cells differ within a section (Malarkey et al 2000, Faa et al 1994). Following this, the method for collection of liver samples was optimized and consistency was maintained while collecting liver samples at day 14. I was hence not able to quantitatively compare compounds at different time points.    Target compound analysis Aim 2: To use target compound analysis by GC-MS and LC-qTOF-ESI-MS of liver, plasma and urine samples to show differences in specific, targeted metabolites of galactose metabolism due to feeding lactose (which consists of 50% galactose) compared to CSS (which is 100% glucose) as the carbohydrate source in rat pups. The metabolites analyzed include galactose and those that are directly involved in galactose metabolism, such as galactonate, galactitol and Gal-1-P, and few other metabolites that maybe linked to galactose metabolism such as glucose, Glu-6-P and ribose-5-phosphate.      58   Liver The means of area of chromatographic peaks of galactose were significantly different between lactose and CSS group, at day 10 (p<0.05). Derivitisation of samples by oximation and silylation leads to the formation of two chromatographic peaks for galactose corresponding to anti and syn isomers (Ruiz-Matute et al 2011, Sanz et al 2003). The two peaks for galactose are obtained at RT ≈19.28 and 19.68 mins. In this case, the peaks were combined and integrated for relative quantitative comparison. The galactose levels were higher in the lactose-fed group when compared to the CSS-fed group at day 10. An illustration of the comparison of area of chromatographic peaks from two infant rat liver samples, one from each group is presented in Figure 8a, with lactose-fed group expressing a high mean area of peak value for galactonate when compared to CSS-fed group at day 10 (Figure 8b).            59   Figure 8a. GC-MS chromatogram of day 10 liver samples from rat pups depicting higher area of peak of galactose at RT≈19.28 &19.69 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using target compound analysis, by running a standard for galactonate to get best estimate of RT. Illustration depicts an example of one sample from each group.     Figure 8b. Day 10 liver bar graph (mean area of peak) of galactose in lactose-fed group (11594572) compared to CSS-fed group of rat pups (6911275), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p<0.05).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; RT, retention time; SD, standard deviation  115945726911275020406080100120140160180200Mean Areaof Peak x 105*Lactose CSSRT ≈19.28 & 19.69 Lactose CSS 60  At day 14, means of area of chromatographic peaks of galactose were also significantly different between lactose and CSS group (p<0.05). Galactose elutes at RT≈19.26 and 19.68 mins. An illustration of the peak differences from two infant rat liver samples, one from each group is presented in Figure 9a. Significantly higher mean area of peak values for galactose was observed in the lactose group when compared to the CSS group at day 14 (Figure 9b).                  61   Figure 9a. GC-MS chromatogram of day 14 liver samples from rat pups depicting higher area of peak of galactose at RT≈19.26 &19.70 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using target compound analysis, by running a standard for galactonate to get best estimate of RT. Illustration depicts an example of one sample from each group.    Figure 9b. Day 14 liver bar graph (mean area of peak) of galactose in lactose-fed group (4597757) compared to CSS-fed group of rat pups (4705148), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p<0.05).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; RT, retention time; SD, standard deviation  45977574705148020406080100120140160180200Mean Areaof Peak x 105*Lactose CSSLactose CSS RT ≈19.26 & 19.70 62     The means of area of chromatographic peaks of galactonate were significantly different between lactose and CSS group, at day 10 (p<0.001). RT of galactonate≈20.66 mins. Galactonate levels were higher in the lactose-fed group when compared to the CSS-fed group at day 10. An illustration of the comparison of area of chromatographic peaks from two infant rat liver samples, one from each group is presented in Figure 10a, with lactose-fed group expressing a high mean area of peak value for galactonate when compared to CSS-fed group at day 10 (Figure 10b).           63     Figure 10a. GC-MS chromatogram of day 10 liver samples from rat pups depicting higher area of peak of galactonate at RT≈20.66 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using target compound analysis, by running a standard for galactonate to get best estimate of RT. Illustration depicts an example of one sample from each group.       Figure 10b. Day 10 liver bar graph (mean area of peak) of galactonate in lactose-fed group (46990) compared to CSS-fed group of rat pups (4166), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p<0.001).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; RT, retention time; SD, standard deviation  4699041660102030405060Mean Area of Peak  x 103Lactos CSS*RT≈ 20.66 Lactose CSS 64   Interestingly at day 14, means of area of chromatographic peaks of galactonate were also significantly different between lactose and CSS group (p<0.001). RT of galactonate≈20.60 mins. An illustration of the peak differences from two infant rat liver samples, one from each group is presented in Figure 11a. Significantly higher mean area of peak values for galactonate was observed in the lactose group when compared to the CSS group at day 14 (Figure 11b).   The chromatographic peaks of galactitol were too small to be detected by the GC-MS, implying negligible amounts of galactitol in liver samples of both lactose-fed and CSS-fed groups, at day 10 and 14.        65   Figure 11a. GC-MS chromatogram of day 14 liver samples from rat pups depicting higher area of peak of galactonate at RT≈20.60 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using target compound analysis, by running a standard for galactonate to get best estimate of RT. Illustration depicts an example of one sample from each group.                Figure 11b. Day 14 liver bar graph (mean area of peak) of galactonate in lactose-fed group (17431) compared to CSS-fed group of rat pups (3218), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p<0.001). CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; RT, retention time; SD, standard deviation  1743132180510152025Mean Areaof Peak x 103*RT≈ 20.60 Lactose   CSS Lactose    CSS  66   Similar to galactose, two chromatographic peaks for glucose corresponding to anti and syn isomers were obtained, at RTs of ≈19.46 and 19.52 mins at day 10. An illustration of the peak differences from two infant rat liver samples, one from each group is presented in Figure 12a. No difference in mean area of peak values for glucose was observed in the lactose-fed group when compared to the CSS-fed group at day 10 (Figure 12b).  Similarly, at day 14 two chromatographic peaks for glucose corresponding to anti and syn isomers were obtained, at RTs of ≈19.40 and 19.49 mins. An illustration of the peak differences from two infant rat liver samples, one from each group is presented in Figure 13a. No difference in mean area of peak values for glucose was observed in the lactose-fed group when compared to the CSS-fed group at day 14 (Figure 13b).    In the liver, at both day 10 and day 14, no differences for Glu-6-P (Figure 14 and Figure 15) and ribose-5-phosphate (Figure 16 and 17) were observed (as presented in Table 3).      67   Figure 12a. GC-MS chromatogram of day 10 liver samples from rat pups depicting no difference in area of peak of glucose at RTs≈19.46 &19.52 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using target compound analysis, by running a standard for galactonate to get best estimate of RT. Illustration depicts an example of one sample from each group.   Figure 12b. Day 10 liver bar graph (mean area of peak) of glucose in lactose-fed group (16216279) compared to CSS-fed group of rat pups (14281478), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.1961).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; RT, retention time; SD, standard deviation 1621627914281478020406080100120140160180200Mean Areaof Peak x 105Lactose CSS RT≈ 19.46 & 19.52    Lactose     CSS 68   Figure 13a. GC-MS chromatogram of day 14 liver samples from rat pups depicting no difference in area of peak of glucose at RTs≈19.40 &19.49 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using target compound analysis, by running a standard for galactonate to get best estimate of RT. Illustration depicts an example of one sample from each group.   Figure 13b. Day 14 liver bar graph (mean area of peak) of glucose in lactose-fed group (4597757) compared to CSS-fed group of rat pups (4705148), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.6688).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; RT, retention time; SD, standard deviation 459775747051480102030405060Mean Areaof Peak x 105Lactose CSSLactose CSS RT≈ 19.40 & 19.49 69    Figure 14. Day 10 liver bar graph (mean area of peak) of Glu-6-P in lactose-fed group (8279895) compared to CSS-fed group of rat pups (8241561), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.8983).  CSS, glucose-corn syrup solids; Glu-6-P, glucose-6-phosphate; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation              Figure 15. Day 14 liver bar graph (mean area of peak) of Glu-6-P in lactose-fed group (9303431) compared to CSS-fed group of rat pups (10760959), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.4780).  CSS, glucose-corn syrup solids; Glu-6-P, glucose-6-phosphate; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation 8279895 82415610102030405060708090100Mean Areaof Peak x 105Lactose CSS930343110760959020406080100120140160Mean Areaof Peak x 105Lactose CSS70                   Figure 16. Day 10 liver bar graph (mean area of peak) of ribose-5-phosphate in lactose-fed group (5017808) compared to CSS-fed group of rat pups (527624), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.1766).  CSS, glucose-corn syrup solids; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation                Figure 17. Day 14 liver bar graph (mean area of peak) of ribose-5-phosphate in lactose-fed group (5624232) compared to CSS-fed group of rat pups (6721895), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.2313).  CSS, glucose-corn syrup solids; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation  50178085276240102030405060Mean Areaof Peak x 105Lactose CSS562423267218950102030405060708090100Mean Areaof Peak x 105Lactose CSS71   Plasma  Galactose, galactonate and galactitol were present in negligible amounts, in both lactose and CSS groups for day 10 and day 14 plasma samples, with small/undetectable peaks. No differences were seen for glucose in the plasma at day 10 and day 14 (Figure 18 and 19).    Figure 18. Day 10 plasma bar graph (mean area of peak) of glucose in lactose-fed group (71723423) compared to CSS-fed group of rat pups (72845347), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.8718).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; SD, standard deviation   71723423 728453470102030405060708090Mean Areaof Peak x 106Lactose CSS72   Figure 19. Day 14 plasma bar graph (mean area of peak) of glucose in lactose-fed group (10849635) compared to CSS-fed group of rat pups (12384523), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.5317).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; SD, standard deviation    No differences for Glu-6-P (Figure 20 and 21) and ribose-5-phosphate (Figure 22 and 23) were observed in day 10 and day 14 plasma samples (as presented in Table 3)  108496351238452305101520Mean Areaof Peak x 106Lactose CSS73   Figure 20. Day 10 plasma bar graph (mean area of peak) of Glu-6-P in lactose-fed group (1289262) compared to CSS-fed group of rat pups (1057011), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.292).  CSS, glucose-corn syrup solids; Glu-6-P, glucose-6-phosphate; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation                 Figure 21. Day 14 plasma bar graph (mean area of peak) of Glu-6-P in lactose-fed group (783940) compared to CSS-fed group of rat pups (768315), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.865).  CSS, glucose-corn syrup solids; Glu-6-P, glucose-6-phosphate; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation  12892621057011024681012141618Mean Areaof Peak x 106Lactose CSS783940768315024681012Mean Areaof Peak x 106Lactose CSS74   Figure 22. Day 10 plasma bar graph (mean area of peak) of ribose-5-phosphate in lactose-fed group (272463) compared to CSS-fed group of rat pups (252476), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.1701).  CSS, glucose-corn syrup solids; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation  Figure 23. Day 14 plasma bar graph (mean area of peak) of ribose-5-phosphate in lactose-fed group (235185) compared to CSS-fed group of rat pups (166960), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.3823).  CSS, glucose-corn syrup solids; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation  272463252476051015202530Mean Areaof Peak x 104Lactose CSS235185166960051015202530354045Mean Areaof Peak x 104Lactose CSS75    Urine No difference was seen in chromatographic peaks of galactonate for day 10 and day 14 urine samples (Figure 24 and 25). The chromatographic peaks of galactitol were not detected, possibly due to negligible amounts of galactitol in urine samples of both lactose-fed and CSS-fed groups, at day 10 and 14. I was not able to detect galactose and glucose in the urine at day 10 and day 14 because the LC-qTOF-ESI-MS was run in negative ion mode and hence could detect only negatively charged ions.  No differences for Glu-6-P (Figure 26 and 27) and ribose-5-phosphate (Figure 28 and 29) (as presented in Table 3) were observed in day 10 and day 14 urine samples                76   Figure 24. Day 10 urine bar graph (mean area of peak) of galactonate in lactose-fed group (45810745) compared to CSS-fed group of rat pups (56318228), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.5275).  CSS, glucose-corn syrup solids; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation       Figure 25. Day 14 urine bar graph (mean area of peak) of galactonate in lactose-fed group (118477636) compared to CSS-fed group of rat pups (104523609), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.5315).  CSS, glucose-corn syrup solids; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation  45810745563182280102030405060708090Mean Areaof Peak x 106Lactose CSS118477636104523609020406080100120140160Mean Areaof Peak x 106Lactose CSS77   Figure 26. Day 10 urine bar graph (mean area of peak) of Glu-6-P in lactose-fed group (267813782) compared to CSS-fed group of rat pups (331123342), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.1070).  CSS, glucose-corn syrup solids; Glu-6-P, glucose-6-phosphate; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation  Figure 27. Day 14 urine bar graph (mean area of peak) of Glu-6-P in lactose-fed group (311315766) compared to CSS-fed group of rat pups (366395445), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.4791).  CSS, glucose-corn syrup solids; Glu-6-P, glucose-6-phosphate; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation  2678137823311233420510152025303540Mean Areaof Peak x 107Lactose CSS31131576636639544505101520253035404550Mean Areaof Peak x 107Lactose CSS78   Figure 28. Day 10 urine bar graph (mean area of peak) of ribose-5-phosphate in lactose-fed group (630114) compared to CSS-fed group of rat pups (732701), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.6893).  CSS, glucose-corn syrup solids; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation  Figure 29. Day 14 urine bar graph (mean area of peak) ribose-5-phosphate in lactose-fed group (1914946) compared to CSS-fed group of rat pups (1846026), by LC-qTOF-ESI-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p=0.8946).  CSS, glucose-corn syrup solids; LC-qTOF-ESI-MS, liquid chromatography quadrupole-time-of-flight electrospray ionization mass spectrometry; SD, standard deviation  630114732701024681012Mean Areaof Peak x 105Lactose CSS1914946 1846026051015202530Mean Areaof Peak x 105Lactose CSS79  Gal-1-P could not be observed both on the GC-MS and LC-qTOF-ESI-MS in liver, urine and plasma samples. I could not obtain a clear chromatogram and mass spectrum for Gal-1-P standard on the GC-MS and hence could not identify and compare Gal-1-P RT and chromatographic peaks. On the LC-qTOF-ESI-MS, Gal-1-P could not be detected because it co-elutes with glucose-1-phosphate (Glu-1-P), fructose-6-phosphate and fructose-1-phosphate. I was not able to separate the peak for Gal-1-P from all these compounds as they have the same average MW and m/z of target ion, and also similar RT values.   Targeted metabolomics Aim 3: To use targeted metabolomics by GC-MS analysis of liver samples with   PCA to identify metabolites in the range of small molecular weight, water-soluble molecules that differentiate rat pups  fed  lactose compared to CSS. The RT values of metabolites that cause variance between the groups will be identified by MS libraries and these will be mapped to pathways of glucose, galactose, amino acid and fatty acid metabolism.  Targeted metabolomics using GC-MS, along with PCA was used to identify metabolites that are of small molecular weight and water-soluble at day 10 and 14. Using the PCA loadings view plot, the mass values of the metabolites of interest were identified; the metabolites furthest from the origin of the plot are the ones that contribute most to the variance. The RT values of these metabolites were noted from the PCA list of metabolites, and identified using MS libraries. The areas of 80  chromatographic peaks of the metabolite were noted from each group and independent samples t-test was used to detect statistical difference.    Day 10 liver samples The plot (Figure 30) shows the metabolites that contribute to variance solely based on the type of carbohydrate source in the fed formula (lactose or CSS) at day 10. Metabolites selected from loadings view plot, for relative quantitative comparison using integration of peaks and tested for statistical significance, are listed in Table 5. The metabolites are listed with their respective approx. values of RTs and were identified from MS libraries based on their mass spectrum fragmentation patterns.              81               Figure 30. Day 10 Liver Sample Loadings view plot from principal component analysis.  Principal component 1 depicted on x-axis and Principal component 2 depicted     on y-axis. Each ‘x’ is the m/z value for a metabolite.  The metabolites that are furthest from the origin of the plot are our points of interest and contribute to most of the variance between the two groups. The RT values of these metabolites are noted and the metabolites are identified using MS libraries. Relative quantitative comparison was then done between the groups by integrating the peaks and obtaining area of chromatographic peaks, and tested for statistical significance    82   Metabolite Retention time in mins (approx. values) Alanine Valine Leucine Isoleucine Proline Glycine Pyrimidine Serine Threonine Malate Aspartic acid D-ribose Purine Lysine Tyrosine Myo-inositol Inositol 7.07 9.00 9.94 10.35 10.39 10.58 11.08 11.47 11.90 13.52 14.01 16.45 18.01 18.69 18.98 21.72 25.20  Table 5. List of metabolites selected from loadings view plot of PCA for day 10 liver samples, by targeted metabolomics for relative quantitative comparison by integration of chromatographic peaks and tested for statistical significance. 83   Metabolites that were significantly higher in the lactose-fed group when compared to the CSS-fed group are: D-ribose (Figures 31a & 31b); Pyrimidine (Figures 32a & 32b); Glycine (Figures 33a & 33b); Malate (Figures 34a & 34b); Leucine (Figures 35a & 35b); Isoleucine (Figures 36a & 36b), and are presented in the following pages.    84   D-ribose    Figure 31a. GC-MS chromatogram of day 10 liver samples from rat pups depicting higher area of peak of D-ribose at RT≈16.44 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using targeted metabolomics, RT of metabolites leading to variance between groups was determined using PCA, and the metabolite was identified by comparing to MS libraries. Illustration depicts an example of one sample from each group.                         Figure 31b. Day 10 liver bar graph (mean area of peak) of D-ribose in lactose-fed group (1715532) compared to CSS-fed group of rat pups (362080), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p<0.05).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; PCA, principal component analysis; RT, retention time; SD, standard deviation  1715532362080020406080100120140160180200Mean Area of Peak x 104 *Lactose CSS     Lactose  CSS RT≈ 16.44 85   Pyrimidine      Figure 32a. GC-MS chromatogram of day 10 liver samples from rat pups depicting higher area of peak of pyrimidine at RT≈11.08 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using targeted metabolomics, RT of metabolites leading to variance between groups was determined using PCA, and the metabolite was identified by comparing to MS libraries. Illustration depicts an example of one sample from each group.                     Figure 32b. Day 10 liver bar graph (mean area of peak) of pyrimidine in lactose-fed group (899282) compared to CSS-fed group of rat pups (372232), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p<0.05).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; PCA, principal component analysis; RT, retention time; SD, standard deviation  8992823722320102030405060708090100Mean Area of Peak x 104*RT≈ 11.08  Lactose CSS Lactose CSS 86   Glycine       Figure 33a. GC-MS chromatogram of day 10 liver samples from rat pups depicting higher area of peak of glycine at RT≈10.58 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using targeted metabolomics, RT of metabolites leading to variance between groups was determined using PCA, and the metabolite was identified by comparing to MS libraries. Illustration depicts an example of one sample from each group.                  Figure 33b. Day 10 liver bar graph (mean area of peak) of glycine in lactose-fed group (37614437) compared to CSS-fed group of rat pups (2386247), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p<0.05).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; PCA, principal component analysis; RT, retention time; SD, standard deviation  3761443723862470510152025303540Mean Area of Peak x 105*RT≈ 10.58 Lactose     CSS       Lactose      CSS  87   Malate   Figure 34a. GC-MS chromatogram of day 10 liver samples from rat pups depicting higher area of peak of malate at RT≈13.52 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using targeted metabolomics, RT of metabolites leading to variance between groups was determined using PCA, and the metabolite was identified by comparing to MS libraries. Illustration depicts an example of one sample from each group.               Figure 34b. Day 10 liver bar graph (mean area of peak) of malate in lactose-fed group (2177872) compared to CSS-fed group of rat pups (1201672), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p<0.05).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; PCA, principal component analysis; RT, retention time; SD, standard deviation 217787212016720510152025Mean Area of Peak x 105 *    Lactose  CSS RT≈ 13.52 Lactose CSS 88   Leucine  Figure 35a. GC-MS chromatogram of day 10 liver samples from rat pups depicting higher area of peak of leucine at RT≈9.94 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using targeted metabolomics, RT of metabolites leading to variance between groups was determined using PCA, and the metabolite was identified by comparing to MS libraries. Illustration depicts an example of one sample from each group.                                   Figure 35b. Day 10 liver bar graph (mean area of peak) of leucine in lactose-fed group (469748) compared to CSS-fed group of rat pups (353803), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p<0.05).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; PCA, principal component analysis; RT, retention time; SD, standard deviation 4697483538030102030405060Mean Area of Peak x 105Lactose CSS*RT≈ 9.94 Lactose CSS 89   Isoleucine  Figure 36a. GC-MS chromatogram of day 10 liver samples from rat pups depicting higher area of peak of isoleucine at RT≈10.35 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using targeted metabolomics, RT of metabolites leading to variance between groups was determined using PCA, and the metabolite was identified by comparing to MS libraries. Illustration depicts an example of one sample from each group.                     Figure 36b. Day 10 liver bar graph (mean area of peak) of isoleucine in lactose-fed group (295654) compared to CSS-fed group of rat pups (226545), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p<0.05). CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; PCA, principal component analysis; RT, retention time; SD, standard deviation  29565422654505101520253035Mean Area of Peak x 104 *Lactose RT≈ 10.35 Lactose  CSS CSS 90   Day 14 liver samples   Figure 37. Day 14 Liver Sample Loadings view plot from Principal Component analysis.  Principal component 1 depicted on x-axis and Principal component 2 depicted     on y-axis. Each ‘x’ is the m/z value for a metabolite.  The metabolites that are furthest from the origin of the plot are our points of interest and contribute to most of the variance between the two groups. The RT values of these metabolites are noted and the metabolites are identified using MS libraries. Relative quantitative comparison was then done between the groups by integrating the peaks and obtaining area of chromatographic peaks, and tested for statistical significance  Metabolites selected from loadings view plot of day 14 samples, for relative quantitative comparison using integration of peaks and tested for statistical significance, are listed in table 6. The metabolites are listed with their respective approx. values of RTs.  Loadings: Component 1 - Component 2Component 10.000 0.020 0.040 0.060 0.080 0.100 0.120 0.140 0.160 0.180Component 2-0.200-0.150-0.100-0.050-0.0000.0500.1000.1500.20091  Metabolite Retention time in mins (approx. values) Leucine Isoleucine Glycine Pyrimidine Serine Threonine Malate D-ribose Purine Myo-inositol 9.92 10.31 10.53 11.04 11.44 11.88 13.49 16.43 17.98 21.68  Table 6: List of metabolites selected from loadings view plot of PCA for day 14 liver samples, by targeted metabolomics for relative quantitative comparison by integration of chromatographic peaks and tested for statistical significance.    Metabolites that were significantly higher in the lactose-fed group when compared to the CSS-fed group were only for leucine (Figures 38a & 38b) and isoleucine (Figures 39a & 39b). 92   Leucine Figure 38a. GC-MS chromatogram of day 14 liver samples from rat pups depicting higher area of peak of leucine at RT≈9.92 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using targeted metabolomics, RT of metabolites leading to variance between groups was determined using PCA, and the metabolite was identified by comparing to MS libraries. Illustration depicts an example of one sample from each group.                  Figure 38b. Day 14 liver bar graph (mean area of peak) of leucine in lactose-fed group (1526629) compared to CSS-fed group of rat pups (878765), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p<0.05).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; PCA, principal component analysis; RT, retention time; SD, standard deviation  1526629878765050100150200250Mean Area of Peak x 104Lactose CSS*Lactose CSS RT≈ 9.92 93   Isoleucine  Figure 39a. GC-MS chromatogram of day 14 liver samples from rat pups depicting higher area of peak of isoleucine at RT≈10.31 mins in lactose-fed group compared to CSS-fed group.   Samples were analyzed using targeted metabolomics, RT of metabolites leading to variance between groups was determined using PCA, and the metabolite was identified by comparing to MS libraries. Illustration depicts an example of one sample from each group.                    Figure 39b. Day 14 liver bar graph (mean area of peak) of isoleucine in lactose-fed group (1306810) compared to CSS-fed group of rat pups (636223), by GC-MS analysis. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) *Significant difference between groups (p<0.05).  CSS, glucose-corn syrup solids; GC-MS, gas chromatography mass spectrometry; PCA, principal component analysis; RT, retention time; SD, standard deviation 1306810636223020406080100120140160180200Mean Area of Peak x104Lactose CSS*Lactose CSS RT≈ 10.31 94   Western blotting  Aim 4: To use western blotting to determine if there is lower expression of key enzymes and transcription factors of liver lipogenesis and glycolysis, including ChREBP, SREBP1-c and SCD in lactose compared to CSS fed rat pups.  There were no differences at Day 10, 14 and 18 liver samples for expression of ChREBP, SREBP1-c and SCD due to lactose or CSS feeding (Data presented in Appendix B).              95  CHAPTER 4: SUMMARY  Discussion The purpose of this thesis was to understand the metabolic effects of feeding lactose vs non-lactose carbohydrates such as CSS to infants, using metabolomics as a tool and a neonatal rat model. This project has provided novel data to show that early dietary lactose may have effects on key hepatic pathways that are not the same as non-lactose carbohydrate feeding. Lactose is an important component of human milk providing 20% energy as galactose, which is cleared primarily by the liver and does not stimulate insulin release. Understanding lactose, and more specifically, galactose in development is important; particularly to better understand the effects of feeding non-lactose carbohydrates to infants. This is the first study to highlight biomarker patterns depicting the benefits of lactose, particularly galactose, at the metabolic level and the implications of feeding non-lactose carbohydrates in clinical nutrition support and human milk substitutes.  The first step in this project was the development of procedures for rearing of rat pups. Rat formula was optimized to use a homogenized mix of minerals, which prevented clumping that occurred in earlier formula preparations that used individual mineral components. To increase survival rate of rat pups post-surgery, gastrostomy experiments included the use of a mucous-like substance called Muko©, as a lubricant to avoid irritation caused to esophagus while inserting the cannulas into the infant rat stomach. While optimizing steps for animal care with each experiment, the feeding cycle was kept at 10 mins, with a break of 20 mins between each cycle and flow rate of 25% body weight. The development of these 96  procedures led to an increase in survival of rat pups by about 2-5% after surgery, during the course of the gastrostomy experiment compared to our previous experiments.  The growth rate of rat pups reared artificially was comparable to mother-reared rat pups. The carbohydrate source also did not affect growth rate of the pups.  Liver collected at day 10 after birth of rat pups showed the greatest difference in metabolites, followed by day 14. Collection of liver samples was optimized, taking into consideration the heterogeneity of liver sections and consistency was maintained. The liver growth was proportional to the body growth and was comparable between groups fed by gastrostomy (lactose or CSS). There were no prominent changes observed in liver growth due to artificial rearing or the different carbohydrate source (lactose or CSS).  Use of target compound analysis showed differences in galactose levels in the liver, significantly higher in lactose-fed rats compared to CSS-fed rats, at day 10 and day 14. Galactonate, an important metabolite of galactose metabolism was found to be significantly higher in lactose-fed rats compared to CSS-fed rats, at day 10 and day 14. This is in accordance to a study done in 1966 by Cuatrecasas & Segal, which depicted the sudden rise of galactose dehydrogenase enzyme at days 5 to 10 in rat liver and used assays with radioactive preparations. Galactitol levels were undetectable in liver and plasma samples of both groups at all time points considered. There was no difference between groups in the glucose levels in the liver, suggesting possible conversion of galactose to glucose in the liver. 97   Targeted metabolomics used in this project proved as a novel and beneficial approach to study biomarker patterns. Till now there has been no study that focuses on identifying metabolic patterns of lactose, particularly galactose metabolism. This project highlights the application of an emerging field such as metabolomics in the study of variables such as diet, the development of which, allows for identification of crucial metabolites responsible for marking profound differences between lactose-fed rats and CSS-fed rats. The significantly higher amounts of several important metabolites such as D-ribose, malate, pyrimidine, glycine, leucine and isoleucine in lactose-fed rats when compared to CSS-fed rats, suggests the potential utilization of galactose in a beneficial manner at the whole body level. Also, there was no difference in the glucose levels between the two groups. The common factor between most metabolites that differed between the groups was their ability to be involved in glucose synthesis. These results suggest the role of galactose as a slow and consistent source of glucose, for supply to glucose-dependent organs such as the brain, kidney and red blood cells. This project also provides evidence for utilization of galactose in the PPP, suggesting a possible role in fatty acid oxidation in infants. Use of PCA, along with targeted metabolomics allowed for high-throughput analysis of samples, with the identification of over 500 mass values for metabolites of carbohydrate, protein and fat metabolism in the liver. This process seems to be fast and efficient, and can prove to be a beneficial application for future clinical studies.  98   Target compound analysis  Galactose Significantly higher levels of galactose in the lactose-fed rats compared to the CSS-fed rats at day 10 and day 14 in the liver was expected, as lactose is 50% galactose. But its presence in liver and absence in the plasma suggests that galactose is preferably utilized and metabolized in the liver. This is in support to earlier studies which show that 90% of galactose is taken up by the liver (Arola & Tamm 1994).  Galactonate The significantly high expression of Galactonate peaks in the lactose group, when compared to the CSS group, gives us an understanding about the route of metabolism of galactose in the infant liver. Studies done in the 1960s by Cuatrecasas and Segal report changes in galactose dehydrogenase activity in rat liver. The activity of galactose dehydrogenase enzyme increased dramatically from 5 to 10 days of age. Our data corresponds to these early studies and suggests that the galactose dehydrogenase pathway may be an important route for galactose metabolism and regulation in newborn infants. The existing knowledge on biochemical pathways suggests that the galactose dehydrogenase pathway may lead to the PPP, via D-xylulose, and with the production of NADH (Cuatrecasas & Segal 1966). The PPP can either provide a substrate for nucleotide synthesis, which is important for DNA synthesis in a developing infant, or produce fructose - 6 - phosphate for entry into the Glycolytic cycle. 99  The oxidative part of the PPP leads to the conversion of Glu-6-P to a pentose phosphate, resulting in the production of NADPH, an electron donor for substrate reduction during fatty acid synthesis. Galactose entry into the PPP via galactonate bypasses production of NADPH and hence may avoid the production of fatty acid synthesis. This proves favorable for infants as they are already fed a high-fat diet and excessive fatty acid accumulation in the liver is not what is desired for healthy growth and development.  Galactitol The negligible amounts of galactitol suggest that the aldose reductase pathway may be a less utilized route for galactose metabolism, when provided with physiological levels of galactose in milk to infants.  Glucose No differences were observed in D-glucose amounts in lactose-fed rat pups compared to CSS-fed rat pups. This is in alignment with previous studies that emphasized on the conversion of galactose to glucose, most probably via the Leloir pathway (Kohn et al 1962, Kunst et al 1989). These early experiments employed radiolabelling techniques to observe conversion of galactose to glucose. Metabolomic analysis replicated results from previous studies and helped in relative quantitative comparison between the groups. This was a relatively simple process with use of very small amount of sample and no use of harmful radioactive elements. Galactose conversion to glucose may prove to be extremely essential to support the infant’s high-energy demands, especially for those organs that cannot utilize energy from 100  fatty acids. Moreover, galactose is non-insulinogenic and hence may prove to be essential for infants to meet high glucose needs during early life.,   Targeted metabolomics  D-ribose and pyrimidine D-ribose is an important component of mammalian metabolism. It can be used for de-novo nucleotide synthesis or metabolized to glycolytic intermediates (Segal & Foley 1958). The lower area of chromatographic peaks of D-ribose and pyrimidine in CSS-fed rat pups when compared to the lactose-fed rat pups, could suggest that nucleotide synthesis was affected. Nucleotide production, leading to RNA and DNA synthesis is highly essential for a developing infant. D-ribose and pyrimidine can be mapped to the PPP. Ribose can be converted to ribose-5-phosphate, which can either go towards nucleotide production or enter the glycolytic cycle via production of fructose-6-phosphate (Segal & Foley 1958) (Figure 40). Earlier experiments on rat liver slices clearly indicated the conversion of free ribose to glucose (Katz et al 1955) and in-vivo studies with mice provided strong evidence for D-ribose as an efficient precursor to liver glycogen (Hiatt 1957). Conversion of ribose to glucose also requires less energy and reducing power than pyruvate (Clark et al 2014). Maintenance of glycogen stores may be particularly useful to support slow and consistent flow of glucose, to ensure constant supplies to glucose-dependent organs such as brain, red blood cells, and kidney, without increasing insulin levels. This is relevant to infants, where there is rapid growth and a large brain/body weight of 10% compared to 1-2% of body weight in adults.  101   Figure 40. Biochemical reactions depicting D-ribose metabolism towards nucleotide  production and glycolytic cycle (source: Segal & Foley 1958)   Glycine Glycine is a critical player in body metabolism and development, and is necessary for survival. It is an important substrate for protein synthesis and contributes to about 20% of amino acid nitrogen in body proteins (Wu 2010). It is also an anti-inflammatory agent (Stoffels et al 2011) and acts as a neurotransmitter, mainly processing motor and sensory information (López-Corcuera et al 2001). It is majorly metabolized via the Glycine Cleavage System (GCS), an enzyme complex present in 102  the inner mitochondrial membrane (Kikuchi et al 2008), which is reported to show the highest activity in the liver (Dasarathy et al 2009). Glycine can be utilized in multiple ways to generate glutathione, purines (DNA and RNA), creatine, heme and serine (Hall 1998), depicted in Figure 41. The higher amounts of glycine in lactose-fed rat pups when compared to the CSS-fed rat pups may suggest the contribution of lactose towards whole body homeostasis and efficient metabolism, especially during the critical period of infancy.  Figure 41. Biochemical reactions depicting glycine metabolism (Adapted from: Kretzschmar 1996, Hall 1998, Wu 2010) 103  In addition, glycine is a significant component of bile acids, necessary for digestion of dietary fat and absorption of long chain fatty acids (Wang et al 2013). Glutathione, a critical anti-oxidant compound is shown to be a regulator of mitochondrial fatty acid oxidation (Kretzschmar 1996, Nguyen et al 2013). In infants, fatty acid oxidation is essential for energy production and prevents accumulation of fats, which may lead to fatty liver. Glycine can also be directed back to pyruvate via serine, to generate glucose, and hence help in glucose homeostasis.   Malate Malate is an important component of the TCA cycle. It keeps oxaloacetate levels constant, which is critical for conservation of the TCA cycle. The TCA cycle generates energy by oxidation of carbohydrates, fats and proteins and chemical energy as ATP (Bechmann et al 2012). Given that malate is found in the oxidative part of the TCA cycle, the lactose-fed rats are more likely to be inclined towards fatty acid oxidation, hence fulfilling the energy requirements of a neonate. An added benefit is the avoidance of fat accumulation in the liver, leading to fatty liver and predisposal to disease later in life. The TCA cycle is shown to donate gluconeogenic material to the cytosol primarily as Malate (Heath & Threlfall 1968). Malate is important for cytosolic OAA production, a possible precursor to generation of glucose from pyruvate (Dashty 2013).    104   Leucine and Isoleucine Leucine and Isoleucine are branched chain amino acids (BCAA). BCAAs play diverse metabolic and physiological roles. BCAAs are unique from other amino acids because they tend to escape first pass metabolism in the liver and are primarily catabolized in the muscle (Freund & Hanani 2002). BCAAs regulate the mTOR signaling pathway, in the liver and muscle and hence regulate protein synthesis, and are also known to enhance glycogen synthesis (Monirujjaman & Ferdouse 2014). Glycogen production may be useful for infants for glucose homeostasis in the liver. BCAAs are known to prevent accumulation of tissue triglycerides and enhance expression of peroxisome proliferator-activated-receptor (PPAR-α), which is involved in stimulating fatty acid oxidation (Holecek 2013, Kersten et al 1999). Catabolism of BCAAs leads to the production of intermediates that can be diverted for use in fatty acid or cholesterol synthesis. Leucine and Isoleucine catabolism leads to production of Acetyl-CoA, an important precursor for fatty acid synthesis (Cole et al 2010). Our results show a reduced expression of BCAAs in the liver in the CSS-fed group, when compared to the lactose-fed group, which directs possibility of increased fatty acid production in the CSS-fed group and reduced energy production via the TCA cycle. This also implies reduced availability of BCAAs to the muscle. BCAAs promote protein synthesis and produce energy in the muscle. They also constitute a major nitrogen source, which could also be a significant component of DNA (Weinert 2009), and reduced availability of BCAA for a growing infant could be unfavorable.  105   Western blotting No differences were observed in the expression of ChREBP, SREBP1-c and SCD in liver samples at all time points. We selected these proteins because they are involved in glycolysis and lipogenesis in the liver, with these proteins regulating fatty acid synthesis from carbohydrates in the liver. Studies however show that fatty acid synthesis in the liver occurs only if the carbohydrate is present in excess (Hillgartner et al 1995). This is supported by studies from rat pups which suggests that activities of lipogenic enzymes are reduced in suckling rats, as milk is high in fat but low in carbohydrate, and this activity only increases when changed to a high carbohydrate diet at weaning (Girard et al 1997). In this project we used lactose compared to CSS as the carbohydrate source in equal amounts and at physiological levels, comparable to milk received by a suckling rat, it is likely that carbohydrate from both groups of formula were used towards satisfying the high energy demands rather than conversion to fat. We might be able to see differences between groups only if higher amounts of carbohydrate are used.  Some studies also suggest that SREBP1-c is only partially involved in fatty acid synthesis from carbohydrates (Iizuka 2004). Hence, future studies could involve looking at other enzymes closely related to glucose, galactose and fatty acid metabolism. In addition, assessing the expression of transcription factors by western blotting does not indicate the activity of these proteins, which is more indicative of their function. In my project, even though there were no differences in the expression of ChREBP and SREBP 1-c between the lactose-fed and CSS-fed group, there is a possibility of differences in activity, which could affect metabolism, and if found can 106  help support data from metabolomics study. Activity of transcription factors can be analyzed using techniques such as chromatin immunoprecipitation combined with massively parallel sequencing (ChIP-seq) (Jeong et al 2011).    Limitations   The rat milk formula prepared used 60 grams of carbohydrate per liter of formula, accounting for approximately 12% of total energy content. This is an increased amount when compared to neonatal rat milk composition, which contains approximately 11.3 grams per liter, accounting for 4% energy. The higher carbohydrate composition was done in order to match human milk carbohydrate composition of approximately 40% energy, and the effect of this on the infant rat’s metabolism is unknown.  The neonatal rat is closer to a preterm infant in maturity at birth (Quinn 2005). Humans and rats also have differences in intestinal morphology, with humans having greater absorptive surface than rats (DeSesso & Jacobson 2001) and distinct physiological differences. These include absence of gall bladder in rats unlike humans (Malarkey et al 2005), brown adipose tissue contributing to regulation of energy and glucose metabolism in rats, the role of which is less known in humans, rodents having a greater basal metabolic rate than humans (Kowalski & Bruce 2014), and higher glycogen storage in liver of rodents than humans (Chandrasekera & Pippin 2013). The pig might be a better non-primate experimental model because of similarity of physiology of digestion and associated metabolic process with humans (García & Díaz-Castro 2013). Rats 107  also have different energy and food intake when compared to humans. The rat pups were fed every 10 mins with a 20-min break, each day, in our experiments. This is different from a human infant that gets fed only every 3 hours. How this affects metabolism is difficult to interpret.  While interpreting these results and trying to extrapolate to humans, it is important to consider the stress induced by gastrostomy and the lack of maternal care in this model. Studies done in rats have shown that daily handling of rat pups leads to increased glucocorticoid concentrations, especially in the first 21 days of life (Meaney & Aitken 1985). Glucocorticoids have been shown to promote lipogenesis within hepatocytes and stimulate gluconeogenesis, which could lead to stress-induced metabolic deficits (Maniam et al 2014).  Gal-1-P is an important metabolite of galactose metabolism. However, I was not able to measure its concentration and hence was not able to quantitatively compare it between groups. Gal-1-P signal could not be detected on the GC-MS and on the LC-qTOF-ESI-MS. Gal-1-P co-elutes with Glu-1-P, fructose-6-phosphate and fructose-1-phosphate and therefore cannot be used for comparison.   Liver collected from experiments done at day 10 did not consider the heterogeneity between liver lobes and there was inconsistency in collection of samples. Further experiments at day 14 considered the complexity of the liver and consistency during sample collection was ensured.  I was not able to compare chromatograms and mass spectra obtained from analysis at different time points due to inconsistency in collection of liver samples 108  between the time points. Since the samples from day 10 and day 14 experiments were run on different days, it was also possible that there was a shift in RT, and hence the time points were not compared in the current thesis.  GC-MS analysis requires derivitization of samples, which may result in sample degradation, and also makes re-use of samples difficult. GC-MS is suitable only for analysis of volatile and low molecular weight metabolites.  Analysis of samples on the LC qTOF-ESI-MS was based only on negative ion mode, which is relevant for comparison of carbohydrate phosphates and acidic components only. The use of positive ion mode, for analysis, was beyond the scope of this project. The changing of instrument set-up from negative to positive mode also requires change of the ion-pairing agent in the mobile phase, which is a rather tedious process.  PCA for finding differences between groups is exploratory in nature. PCA results are hard to interpret and only directions with largest variance are assumed to be of most interest.   Future directions  Metabolomic analysis on the LC-qTOF-ESI-MS was done only in negative ion mode. A lot of other metabolites can be observed using positive ion mode. These include neutral lipids, phosphotidylcholines, sphingomyelins, acylcarnitines etc. (Dettmer et al 2007). More work needs to be done on identifying the proper ion pairing to be used in mobile phase for positively charged metabolites in ESI. Till 109  now volatile perfluorinated acids such as trifluoroacetic acid (TFA) have been used as an ion-pairing agent for separation of peptides (Lu et al 2008).  In this project, I set up methods for metabolomic analysis of liver, plasma and urine samples. It is important to perform untargeted metabolomics on urine samples to obtain metabolite patterns corresponding to lactose or CSS feeding. This can then be translated to human studies. Urine samples, which are relatively easy to obtain, can then be used to perform targeted and untargeted metabolomic analysis urine samples from infants fed formula containing only CSS, to understand metabolic implications which can lead to better dietary recommendations.  From the results obtained in this project, I interpret that galactose might have a predisposition of forming glucose and glycogen. Further understanding is needed on the route of galactose to glucose/glycogen conversion and the rate of conversion after consumption of lactose in infants.  Further work needs to be done on understanding differences in lipid metabolism in the liver and adipose tissues due to lactose vs. CSS feeding. The next step would be to look at triglyceride content in liver and plasma and compare the lactose-fed and CSS-fed rat pups.   Given that no differences in expression of ChREBP, SREBP1-c and SCD were found by western blotting, I could target other enzymes relevant to carbohydrate and fat metabolism. This can include enzymes such as PPAR-α, which stimulates oxidation of fatty acids, and prevents its accumulation in the liver (Kersten et al 110  1999). It is also important to assess the activity of these proteins, which is more indicative of their function, using techniques such as ChIP-seq.  It will be interesting to investigate utilization of galactose in the brain and implications of feeding glucose-based formula to infants on brain function. Other than the liver, the brain is known to have the capacity to metabolize galactose forming certain amino acids (Roser et al 2009). Studying galactose metabolites in brain and obtaining metabolite patterns using metabolomic analysis can pave the way for study of neurological disorders.  Experiments done in this project can be repeated using pigs or rhesus monkey as models with increased sample size, which might be better models for observing metabolic changes when compared to rat models.   Conclusions It is known that human breast milk is the ideal source of nutrition for a neonate. Lactose (50% galactose) is not only a major component of milk but is also unique to milk, and is consumed exclusively by breast-fed infants. The importance of galactose in infants is poorly understood. The poor understanding of galactose metabolism, with very few studies done several decades ago, and the use of non-lactose carbohydrates in clinical support for infants, formed the central basis of this project.  With CSS, which is composed of glucose chains, being extensively used in infant formula, and lactose differing from glucose in several aspects, I hypothesized that lactose would differ from CSS in their metabolic effects in a neonatal rat feeding model. Metabolomics, a relatively new area of ‘omics’ was used to identify biomarkers 111  relevant to galactose and glucose metabolism. Considering the novelty of this project at several levels, a majority of time was spent on optimizing conditions and developing methodology. CSS has never been fed to rat pups before and conditions needed to be optimized for rearing of rat pups. Along with this, setting up of mass spectrometric instrument and sample preparation for metabolomic analyses, along with obtaining meaningful results from PCA was particularly challenging. Metabolomic analysis provided several interesting results. Galactose and galactonate was found to be different between the groups at both time points, but only in the liver and not in plasma and urine suggesting unique metabolism and utilization of galactose in the liver. Identifying Gal-1-P (a major metabolite of galactose metabolism) will give a clear understanding of the preferred route of galactose metabolism. Differences were observed in metabolites- D-ribose, pyrimidine, glycine, malate, leucine and isoleucine, with these being significantly higher in the lactose-fed group compared to the CSS-fed group. These metabolites are important for a developing and growing infant, particularly ribose and pyrimidine contributing to DNA synthesis, malate contributing to energy production from fatty acids, glycine contributing to regulation of overall body metabolism, and leucine and isoleucine resulting in protein synthesis. Overall, these metabolites are essential for healthy development and sustenance of an infant. Along with the discussed benefits, these metabolites may also help satisfy the high glucose (energy) demands of infants by directly or indirectly being involved in glucose production. Thus, I was able to efficiently use metabolomics to highlight certain metabolite patterns of lactose feeding 112  and differentiate it from CSS feeding, elucidating the possible implications of non-lactose carbohydrates on infant metabolism. These results along with future work done in this area, can contribute to dietary recommendations for infants, focused on promoting the use of lactose as the sole carbohydrate in infant formula and discouraging use of non-lactose carbohydrates, and possible use of galactose along with glucose in intravenous nutrition for preterm infants. However, it is also important to consider the fact that as of now metabolomic analysis is exploratory in nature, and results from metabolomics alone is not sufficient to claim conclusive evidence. Further work is necessary with more focused hypothesis and objectives with perhaps additional techniques such as genomics or proteomics.  Based on results from this project, I can confidently say that we are moving in a positive direction in terms of using metabolomics technology efficiently and obtaining significant metabolite differences, with metabolites being mapped to galactose, glucose, amino acid and fatty acid metabolism. To support the inferences made, it is necessary to further study other metabolites specific to biochemical pathways of interest and relevance to this project, along with support from data of western blot analysis.  With successful development and application of mass-spectrometric tools, and identification of metabolites of galactose metabolism, and metabolite patterns differentiating lactose and CSS metabolism in rat pups, this project certainly paves way for innovation in the field of metabolomics, especially with increase in interest in use of metabolomics for high-throughput analysis. Taking this as a stepping stone, further work needs to be done to enhance knowledge on metabolic pathways, with 113  consideration of a vast range of mass spectrometric techniques to study metabolites, choosing metabolites to be closely looked at from a wide array of identified metabolites and developing techniques to improve reliability on metabolomics data. 114  REFERENCES Abdi, H., & Williams, L. J. (2010). Principal component analysis. WIRES Comp Stats, 2(4), 433-459.  Agostoni, C., Axelsson, I., Goulet, O., Koletzko, B., Michaelsen, K. F., Puntis, J., . . . Nutrition, E. C. o. (2006). Soy protein infant formulae and follow-on formulae: a commentary by the ESPGHAN Committee on Nutrition. J Pediatr Gastroenterol Nutr, 42(4), 352-361. doi:10.1097/01.mpg.0000189358.38427.cd Alexandre-Gouabau, M. C., Courant, F., Moyon, T., Küster, A., Le Gall, G., Tea, I., . . . Darmaun, D. (2013). Maternal and cord blood LC-HRMS metabolomics reveal alterations in energy and polyamine metabolism, and oxidative stress in very-low birth weight infants. J Proteome Res, 12(6), 2764-2778. doi:10.1021/pr400122v Ardeshir, A., Narayan, N. R., Méndez-Lagares, G., Lu, D., Rauch, M., Huang, Y., . . . Hartigan-O'Connor, D. J. (2014). Breast-fed and bottle-fed infant rhesus macaques develop distinct gut microbiotas and immune systems. Sci Transl Med, 6(252), 252ra120. doi:10.1126/scitranslmed.3008791 Arola, H., & Tamm, T. (1994). Metabolism of lactose in the human body. Scandinavian Journal of Gastroenterology, 202, 21-25.  Auestad, N., Korsak, R. A., Bergstrom, J. D., & Edmond, J. (1989). Milk-substitutes comparable to rat's milk; their preparation, composition and impact on development and metabolism in the artificially reared rat. Br J Nutr, 61(3), 495-518.  Auray-Blais, C., Maranda, B., & Lavoie, P. (2014). High-throughput tandem mass spectrometry multiplex analysis for newborn urinary screening of creatine synthesis and transport disorders, Triple H syndrome and OTC deficiency. Clin Chim Acta, 436, 115  249-255. doi:10.1016/j.cca.2014.05.024 Barallobre-Barreiro, J., Chung, Y. L., & Mayr, M. (2013). Proteomics and metabolomics for mechanistic insights and biomarker discovery in cardiovascular disease. Rev Esp Cardiol (Engl Ed), 66(8), 657-661. doi:10.1016/j.rec.2013.04.009 Bechmann, L. P., Hannivoort, R. A., Gerken, G., Hotamisligil, G. S., Trauner, M., & Canbay, A. (2012). The interaction of hepatic lipid and glucose metabolism in liver diseases. J Hepatol, 56(4), 952-964. doi:10.1016/j.jhep.2011.08.025 Ben, X. M. (2008). Nutritional management of newborn infants: practical guidelines. World J Gastroenterol, 14(40), 6133-6139.  Berg JM, Tymoczko JL, Stryer L. Biochemistry. 5th edition. New York: W H Freeman; 2002. Available from: http://www.ncbi.nlm.nih.gov/books/NBK21154/ Bhatia, J., Greer, F., & Nutrition, A. A. o. P. C. o. (2008). Use of soy protein-based formulas in infant feeding. Pediatrics, 121(5), 1062-1068. doi:10.1542/peds.2008-0564 Bird, I. M. (1989). High performance liquid chromatography: principles and clinical applications. BMJ, 299(6702), 783-787.  Boden, G., Cheung, P., Stein, T. P., Kresge, K., & Mozzoli, M. (2002). FFA cause hepatic insulin resistance by inhibiting insulin suppression of glycogenolysis. Am J Physiol Endocrinol Metab, 283(1), E12-19. doi:10.1152/ajpendo.00429.2001 Brew, K., Vanaman, T. C., & Hill, R. L. (1968). The role of alpha-lactalbumin and the A protein in lactose synthetase: a unique mechanism for the control of a biological reaction. Proc Natl Acad Sci U S A, 59(2), 491-497.  Brosnan, J. T., & Brosnan, M. E. (2006). Branched-chain amino acids: enzyme and substrate regulation. J Nutr, 136(1 Suppl), 207S-211S.  116  Bruno, R.-D. (1983). Human milk. Endeavour, 7(2), 80-87.  Cahill, G. F., Ashmore, J., Renold, A. E., & Hastings, A. B. (1959). Blood glucose and the liver. Am J Med, 26(2), 264-282.  Cai, Z., Zhao, J. S., Li, J. J., Peng, D. N., Wang, X. Y., Chen, T. L., . . . Xie, D. (2010). A combined proteomics and metabolomics profiling of gastric cardia cancer reveals characteristic dysregulations in glucose metabolism. Mol Cell Proteomics, 9(12), 2617-2628. doi:10.1074/mcp.M110.000661 Cassis, L. A. (2000). Fat cell metabolism: insulin, fatty acids, and renin. Curr Hypertens Rep, 2(2), 132-138.  Cesare Marincola, F., Dessì, A., Corbu, S., Reali, A., & Fanos, V. (2015). Clinical impact of human breast milk metabolomics. Clin Chim Acta. doi:10.1016/j.cca.2015.02.021 Chandrasekera, P. C., & Pippin, J. J. (2014). Of rodents and men: species-specific glucose regulation and type 2 diabetes research. ALTEX, 31(2), 157-176.  Chawla, D., Thukral, A., Agarwal, R., Deorari, A. K., & Paul, V. K. (2008). Parenteral nutrition. Indian J Pediatr, 75(4), 377-383. doi:10.1007/s12098-008-0042-5 Cherrington, A. D., Moore, M. C., Sindelar, D. K., & Edgerton, D. S. (2007). Insulin action on the liver in vivo. Biochem Soc Trans, 35(Pt 5), 1171-1174. doi:10.1042/BST0351171 Clark, P. M., Flores, G., Evdokimov, N. M., McCracken, M. N., Chai, T., Nair-Gill, E., . . . Witte, O. N. (2014). Positron emission tomography probe demonstrates a striking concentration of ribose salvage in the liver. Proc Natl Acad Sci U S A, 111(28), E2866-2874. doi:10.1073/pnas.1410326111 Coelho, A. I., Berry, G. T., & Rubio-Gozalbo, M. E. (2015). Galactose metabolism and health. Curr Opin Clin Nutr Metab Care, 18(4), 422-427. 117  doi:10.1097/MCO.0000000000000189 Cohn, R. M., & Segal, S. (1973). Galactose metabolism and its regulation. Metabolism, 22(4), 627-642.  Cole, J.T. (2010). Metabolism of BCAAs. In R. R. e. al (Ed.), Branched Chain Amino Acids in Clinical Nutrition (Vol. 1). New York: Springer Science+Business Media. Cornblath, M., Wybregt, S. H., & Baens, G. S. (1963). Studies of carbohydrate metabolism in newborn infant. VII. Tests of carbohydrate tolerance in premature infants. Pediatrics, 32, 1007-1024.  Cuatrecasas, P., & Segal, S. (1966). Mammalian galactose dehydrogenase. II. Properties, substrate specificity, and developmental changes. J Biol Chem, 241(24), 5910-5918.  Dasarathy, S., Kasumov, T., Edmison, J. M., Gruca, L. L., Bennett, C., Duenas, C., . . . Kalhan, S. C. (2009). Glycine and urea kinetics in nonalcoholic steatohepatitis in human: effect of intralipid infusion. Am J Physiol Gastrointest Liver Physiol, 297(3), G567-575. doi:10.1152/ajpgi.00042.2009 Dashty, M. (2013). A quick look at biochemistry: carbohydrate metabolism. Clin Biochem, 46(15), 1339-1352. doi:10.1016/j.clinbiochem.2013.04.027 Dessì, A., Puddu, M., Ottonello, G., & Fanos, V. (2013). Metabolomics and fetal-neonatal nutrition: between "not enough" and "too much". Molecules, 18(10), 11724-11732. doi:10.3390/molecules181011724 Dettmer, K., Aronov, P. A., & Hammock, B. D. (2007). Mass spectrometry-based metabolomics. Mass Spectrom Rev, 26(1), 51-78. doi:10.1002/mas.20108 DeSesso, J. M., & Jacobson, C. F. (2001). Anatomical and physiological parameters affecting gastrointestinal absorption in humans and rats. Food Chem Toxicol, 39(3), 118  209-228.  Dobbing, J., & Sands, J. (1973). Quantitative growth and development of human brain. Arch Dis Child, 48(10), 757-767.  Donnell, G. N., Ng, W. G., Hodgman, J. E., & Bergren, W. R. (1967). Galactose metabolism in the newborn infant. Pediatrics, 39(6), 829-837.  Dudley, E., Yousef, M., Wang, Y., & Griffiths, W. J. (2010). Targeted metabolomics and mass spectrometry. Adv Protein Chem Struct Biol, 80, 45-83. doi:10.1016/B978-0-12-381264-3.00002-3 Dunn, W. B., & Ellis, D. I. (2005). Metabolomics: Current analytical platforms and methodologies. Trends in analytical chemistry, 24(4), 285-294.  Faa, G., Sciot, R., Farci, A. M., Callea, F., Ambu, R., Congiu, T., . . . Costa, V. (1994). Iron concentration and distribution in the newborn liver. Liver, 14(4), 193-199.  Fanos, V., Antonucci, R., Barberini, L., Noto, A., & Atzori, L. (2012). Clinical application of metabolomics in neonatology. J Matern Fetal Neonatal Med, 25 Suppl 1, 104-109. doi:10.3109/14767058.2012.663198 Fanos, V., Iacovidou, N., Puddu, M., Ottonello, G., Noto, A., & Atzori, L. (2013). Metabolomics in neonatal life. Early Hum Dev, 89 Suppl 1, S7-10. doi:10.1016/S0378-3782(13)70003-3 Fehr, M., Takanaga, H., Ehrhardt, D. W., & Frommer, W. B. (2005). Evidence for high-capacity bidirectional glucose transport across the endoplasmic reticulum membrane by genetically encoded fluorescence resonance energy transfer nanosensors. Mol Cell Biol, 25(24), 11102-11112. doi:10.1128/MCB.25.24.11102-11112.2005 Felig, P., Wahren, J., Sherwin, R., & Palaiologos, G. (1977). Amino acid and protein 119  metabolism in diabetes mellitus. Arch Intern Med, 137(4), 507-513.  Fenn, J. B., Mann, M., Meng, C. K., Wong, S. F., & Whitehouse, C. M. (1989). Electrospray ionization for mass spectrometry of large biomolecules. Science, 246(4926), 64-71.  Ficicioglu, C., Hussa, C., Gallagher, P. R., Thomas, N., & Yager, C. (2010). Monitoring of biochemical status in children with Duarte galactosemia: utility of galactose, galactitol, galactonate, and galactose 1-phosphate. Clin Chem, 56(7), 1177-1182. doi:10.1373/clinchem.2010.144097 Fiehn, O. (2002). Metabolomics--the link between genotypes and phenotypes. Plant Mol Biol, 48(1-2), 155-171.  Filhoulaud, G., Guilmeau, S., Dentin, R., Girard, J., & Postic, C. (2013). Novel insights into ChREBP regulation and function. Trends Endocrinol Metab, 24(5), 257-268. doi:10.1016/j.tem.2013.01.003 Fomon, S. (2001). Infant feeding in the 20th century: formula and beikost. J Nutr, 131(2), 409S-420S.  Freund, H. R., & Hanani, M. (2002). The metabolic role of branched-chain amino acids. Nutrition, 18(3), 287-288.  Gao, X., Pujos-Guillot, E., & Sébédio, J. L. (2010). Development of a quantitative metabolomic approach to study clinical human fecal water metabolome based on trimethylsilylation derivatization and GC/MS analysis. Anal Chem, 82(15), 6447-6456. doi:10.1021/ac1006552 García, Y., & Díaz-Castro, J. (2013). Advantages and disadvantages of the animal models v. in vitro studies in iron metabolism: a review. Animal, 7(10), 1651-1658. doi:10.1017/S1751731113001134 120  Girard, J., Ferré, P., & Foufelle, F. (1997). Mechanisms by which carbohydrates regulate expression of genes for glycolytic and lipogenic enzymes. Annu Rev Nutr, 17, 325-352. doi:10.1146/annurev.nutr.17.1.325 Goresky, C. A., Bach, G. G., & Nadeau, B. E. (1973). On the uptake of materials by the intact liver. The transport and net removal of galactose. J Clin Invest, 52(5), 991-1009. doi:10.1172/JCI107300 Gowda, G. A., & Djukovic, D. (2014). Overview of mass spectrometry-based metabolomics: opportunities and challenges. Methods Mol Biol, 1198, 3-12. doi:10.1007/978-1-4939-1258-2_1 Hall, J. C. (1998). Glycine. JPEN J Parenter Enteral Nutr, 22(6), 393-398.  Hall, R. T., & Carroll, R. E. (2000). Infant feeding. Pediatr Rev, 21(6), 191-199; quiz 200.  Hawdon, J. M., Aynsley-Green, A., Alberti, K. G., & Ward Platt, M. P. (1993). The role of pancreatic insulin secretion in neonatal glucoregulation. I. Healthy term and preterm infants. Arch Dis Child, 68(3 Spec No), 274-279.  Heath, D. F., & Threlfall, C. J. (1968). The interaction of glycolysis, gluconeogenesis and the tricarboxylic acid cycle in rat liver in vivo. Biochem J, 110(2), 337-362.  Hiatt, H. H. (1957). Glycogen formation via the pentose phosphate pathway in mice in vivo. J Biol Chem, 224(2), 851-859.  Hillgartner, F. B., Salati, L. M., & Goodridge, A. G. (1995). Physiological and molecular mechanisms involved in nutritional regulation of fatty acid synthesis. Physiol Rev, 75(1), 47-76.  Ho, C. S., Lam, C. W., Chan, M. H., Cheung, R. C., Law, L. K., Lit, L. C., . . . Tai, H. L. (2003). Electrospray ionisation mass spectrometry: principles and clinical 121  applications. Clin Biochem Rev, 24(1), 3-12.  Holcapek, M., Kolárová, L., & Nobilis, M. (2008). High-performance liquid chromatography-tandem mass spectrometry in the identification and determination of phase I and phase II drug metabolites. Anal Bioanal Chem, 391(1), 59-78. doi:10.1007/s00216-008-1962-7 Holden, H. M., Rayment, I., & Thoden, J. B. (2003). Structure and function of enzymes of the Leloir pathway for galactose metabolism. J Biol Chem, 278(45), 43885-43888. doi:10.1074/jbc.R300025200 Holecek, M. (2013). Branched-chain amino acids and ammonia metabolism in liver disease: therapeutic implications. Nutrition, 29(10), 1186-1191. doi:10.1016/j.nut.2013.01.022 Iizuka, K., Bruick, R. K., Liang, G., Horton, J. D., & Uyeda, K. (2004). Deficiency of carbohydrate response element-binding protein (ChREBP) reduces lipogenesis as well as glycolysis. Proc Natl Acad Sci U S A, 101(19), 7281-7286. doi:10.1073/pnas.0401516101 Iizuka, K. (2013). Recent progress on the role of ChREBP in glucose and lipid metabolism. Endocr J, 60(5), 543-555.  Ingram, C. J., Mulcare, C. A., Itan, Y., Thomas, M. G., & Swallow, D. M. (2009). Lactose digestion and the evolutionary genetics of lactase persistence. Hum Genet, 124(6), 579-591. doi:10.1007/s00439-008-0593-6 Janečková, H., Hron, K., Wojtowicz, P., Hlídková, E., Barešová, A., Friedecký, D., . . . Adam, T. (2012). Targeted metabolomic analysis of plasma samples for the diagnosis of inherited metabolic disorders. J Chromatogr A, 1226, 11-17. doi:10.1016/j.chroma.2011.09.074 122  Jenness, R. (1979). The composition of human milk. Semin Perinatol 3(3), 225-239.  Jeong, Y. S., Kim, D., Lee, Y. S., Kim, H. J., Han, J. Y., Im, S. S., . . . Cha, J. Y. (2011). Integrated expression profiling and genome-wide analysis of ChREBP targets reveals the dual role for ChREBP in glucose-regulated gene expression. PLoS One, 6(7), e22544. doi:10.1371/journal.pone.0022544 Johnson, C. H., & Gonzalez, F. J. (2012). Challenges and opportunities of metabolomics. J Cell Physiol, 227(8), 2975-2981. doi:10.1002/jcp.24002 Jones, M. O., Pierro, A., Hammond, P., Nunn, A., & Lloyd, D. A. (1993). Glucose utilization in the surgical newborn infant receiving total parenteral nutrition. J Pediatr Surg, 28(9), 1121-1125.  Kahn, S. E., Prigeon, R. L., McCulloch, D. K., Boyko, E. J., Bergman, R. N., Schwartz, M. W., . . . Palmer, J. P. (1994). The contribution of insulin-dependent and insulin-independent glucose uptake to intravenous glucose tolerance in healthy human subjects. Diabetes, 43(4), 587-592.  Kalhan, S. C., & Kiliç, I. (1999). Carbohydrate as nutrient in the infant and child: range of acceptable intake. Eur J Clin Nutr, 53 Suppl 1, S94-100.  Kamel, A., & Prakash, C. (2006). High performance liquid chromatography/atmospheric pressure ionization/tandem mass spectrometry (HPLC/API/MS/MS) in drug metabolism and toxicology. Curr Drug Metab, 7(8), 837-852.  Katz, J., Abraham, S., Hill, R., & Chaikoff, I. L. (1955). The occurrence and mechanism of the hexose monophosphate shunt in rat liver slices. J Biol Chem, 214(2), 853-868.  Keen, C. L., Lönnerdal, B., Clegg, M., & Hurley, L. S. (1981). Developmental changes in composition of rat milk: trace elements, minerals, protein, carbohydrate and fat. J 123  Nutr, 111(2), 226-236.  Kersten, S., Seydoux, J., Peters, J. M., Gonzalez, F. J., Desvergne, B., & Wahli, W. (1999). Peroxisome proliferator-activated receptor alpha mediates the adaptive response to fasting. J Clin Invest, 103(11), 1489-1498. doi:10.1172/JCI6223 Khitan, Z., & Kim, D. H. (2013). Fructose: a key factor in the development of metabolic syndrome and hypertension. J Nutr Metab, 2013, 682673. doi:10.1155/2013/682673 Kikuchi, G., Motokawa, Y., Yoshida, T., & Hiraga, K. (2008). Glycine cleavage system: reaction mechanism, physiological significance, and hyperglycinemia. Proc Jpn Acad Ser B Phys Biol Sci, 84(7), 246-263.  Kliegman, R. M., Morton, S., & Miettinen, E. (1984). Galactose alimentation in the pup of a diabetic mother. Pediatric Research, 18, 140A-140A.  Kliegman, R. M., & Sparks, J. W. (1985). Perinatal galactose metabolism. The journal of Paediatrics, 107(6), 831-841.  Kogure, K., Ishizaki, M., Nemoto, M., Kuwano, H., & Makuuchi, M. (1999). A comparative study of the anatomy of rat and human livers. J Hepatobiliary Pancreat Surg, 6(2), 171-175.  Kohn, B. D., Hoffman, R. C., & Kohn, P. (1963). A study of the mechanism of conversion of galactose to glucose in the intact rat. J Biol Chem, 238, 1193-1195.  Koletzko, B., Baker, S., Cleghorn, G., Neto, U. F., Gopalan, S., Hernell, O., . . . Zong-Yi, D. (2005). Global standard for the composition of infant formula: recommendations of an ESPGHAN coordinated international expert group. J Pediatr Gastroenterol Nutr, 41(5), 584-599.  Kowalski, G. M., & Bruce, C. R. (2014). The regulation of glucose metabolism: implications 124  and considerations for the assessment of glucose homeostasis in rodents. Am J Physiol Endocrinol Metab, 307(10), E859-871. doi:10.1152/ajpendo.00165.2014 Kretzschmar, M. (1996). Regulation of hepatic glutathione metabolism and its role in hepatotoxicity. Exp Toxicol Pathol, 48(5), 439-446.  Kuhn, N. J. (1972). The lactose and neuraminlactose content of rat milk and mammary tissue. Biochem J, 130(1), 177-180.  Kuhn, N. J., & White, A. (2009). The topography of lactose synthesis. 1975. J Mammary Gland Biol Neoplasia, 14(3), 213-220. doi:10.1007/s10911-009-9136-6 Kunst, C., Kliegman, R., & Trindade, C. (1989). The glucose-galactose paradox in neonatal murine hepatic glycogen synthesis. Am J Physiol, 257(5 Pt 1), E697-703.  Lacy, W. W. (1973). Role of insulin in amino acid metabolism. South Med J, 66(1), 79-84.  Lai, Y. S., Chen, W. C., Kuo, T. C., Ho, C. T., Kuo, C. H., Tseng, Y. J., . . . Sheen, L. Y. (2015). Mass-Spectrometry-Based Serum Metabolomics of a C57BL/6J Mouse Model of High-Fat-Diet-Induced Non-alcoholic Fatty Liver Disease Development. J Agric Food Chem, 63(35), 7873-7884. doi:10.1021/acs.jafc.5b02830 Lebenthal, E., Heitlinger, L., Lee, P.C., Nord, K.S., Hodge, C., Brooks, S.P., & George, D. (1983). Corn syrup sugars: In vitro and in vivo digestibility and clinical tolerance in acute diarrhea of infancy. The Journal of Paediatrics, 103(1), 29-34.  Ligon, W. V., & May, R. J. (1984). Target compound analysis by two-dimensional gas chromatography-mass spectrometry. Journal of chromatography, 294, 77-89.  López-Corcuera, B., Geerlings, A., & Aragón, C. (2001). Glycine neurotransmitter transporters: an update. Mol Membr Biol, 18(1), 13-20.  Lu, W., Bennett, B. D., & Rabinowitz, J. D. (2008). Analytical strategies for LC-MS-based 125  targeted metabolomics. J Chromatogr B Analyt Technol Biomed Life Sci, 871(2), 236-242. doi:10.1016/j.jchromb.2008.04.031 Lu, W., Clasquin, M. F., Melamud, E., Amador-Noguez, D., Caudy, A. A., & Rabinowitz, J. D. (2010). Metabolomic analysis via reversed-phase ion-pairing liquid chromatography coupled to a stand alone orbitrap mass spectrometer. Anal Chem, 82(8), 3212-3221. doi:10.1021/ac902837x Luo, B., Groenke, K., Takors, R., Wandrey, C., & Oldiges, M. (2007). Simultaneous determination of multiple intracellular metabolites in glycolysis, pentose phosphate pathway and tricarboxylic acid cycle by liquid chromatography-mass spectrometry. J Chromatogr A, 1147(2), 153-164. doi:10.1016/j.chroma.2007.02.034 Lönnerdal, B., & Lien, E. L. (2003). Nutritional and physiologic significance of alpha-lactalbumin in infants. Nutr Rev, 61(9), 295-305.  Ma, B., Zhang, Q., Wang, G. J., A, J. Y., Wu, D., Liu, Y., . . . Zheng, Y. Y. (2011). GC-TOF/MS-based metabolomic profiling of estrogen deficiency-induced obesity in ovariectomized rats. Acta Pharmacol Sin, 32(2), 270-278. doi:10.1038/aps.2010.196 Magnusson, I., Chandramouli, V., Schumann, W. C., Kumaran, K., Wahren, J., & Landau, B. R. (1988). Pentose pathway in human liver. Proc Natl Acad Sci U S A, 85(13), 4682-4685.  Malarkey, D. E., Johnson, K., Ryan, L., Boorman, G., & Maronpot, R. R. (2005). New insights into functional aspects of liver morphology. Toxicol Pathol, 33(1), 27-34. doi:10.1080/01926230590881826 Maniam, J., Antoniadis, C., & Morris, M. J. (2014). Early-Life Stress, HPA Axis Adaptation, and Mechanisms Contributing to Later Health Outcomes. Front Endocrinol 126  (Lausanne), 5, 73. doi:10.3389/fendo.2014.00073 Martin-Calama, J., Buñuel, J., Valero, M. T., Labay, M., Lasarte, J. J., Valle, F., & de Miguel, C. (1997). The effect of feeding glucose water to breastfeeding newborns on weight, body temperature, blood glucose, and breastfeeding duration. J Hum Lact, 13(3), 209-213.  McDevitt, R. M., Bott, S. J., Harding, M., Coward, W. A., Bluck, L. J., & Prentice, A. M. (2001). De novo lipogenesis during controlled overfeeding with sucrose or glucose in lean and obese women. Am J Clin Nutr, 74(6), 737-746.  Meaney, M. J., & Aitken, D. H. (1985). The effects of early postnatal handling on hippocampal glucocorticoid receptor concentrations: temporal parameters. Brain Res, 354(2), 301-304.  Md., M., & Afroza, F. (2014). Metabolic and Physiological Roles of Branched-Chain Amino Acids. Advances in Molecular Biology, 2014, 6 pages.  Moore, M. C., Coate, K. C., Winnick, J. J., An, Z., & Cherrington, A. D. (2012). Regulation of hepatic glucose uptake and storage in vivo. Adv Nutr, 3(3), 286-294. doi:10.3945/an.112.002089 Mussap, M., Antonucci, R., Noto, A., & Fanos, V. (2013). The role of metabolomics in neonatal and pediatric laboratory medicine. Clin Chim Acta, 426, 127-138. doi:10.1016/j.cca.2013.08.020 Neville, M. C., McFadden, T. B., & Forsyth, I. (2002). Hormonal regulation of mammary differentiation and milk secretion. J Mammary Gland Biol Neoplasia, 7(1), 49-66.  Newsholme, E. A., & Dimitriadis, G. (2001). Integration of biochemical and physiologic effects of insulin on glucose metabolism. Exp Clin Endocrinol Diabetes, 109 Suppl 2, 127  S122-134. doi:10.1055/s-2001-18575 Nguyen, D., Samson, S. L., Reddy, V. T., Gonzalez, E. V., & Sekhar, R. V. (2013). Impaired mitochondrial fatty acid oxidation and insulin resistance in aging: novel protective role of glutathione. Aging Cell, 12(3), 415-425. doi:10.1111/acel.12073 Nicholas, K. R., & Hartmann, P. E. (1981). Progressive changes in plasma progesterone, prolactin and corticosteroid levels during late pregnancy and the initiation of lactose synthesis in the rat. Aust J Biol Sci, 34(4), 445-454.  Nyamundanda, G., Brennan, L., & Gormley, I. C. (2010). Probabilistic principal component analysis for metabolomic data. BMC Bioinformatics, 11, 571. doi:10.1186/1471-2105-11-571 Oliver, W. T., Mathews, S. A., Phillips, O., Jones, E. E., Odle, J., & Harrell, R. J. (2002). Efficacy of partially hydrolyzed corn syrup solids as a replacement for lactose in manufactured liquid diets for neonatal pigs. J Anim Sci, 80(1), 143-153.  O'Sullivan, A., He, X., McNiven, E. M., Hinde, K., Haggarty, N. W., Lönnerdal, B., & Slupsky, C. M. (2013). Metabolomic phenotyping validates the infant rhesus monkey as a model of human infant metabolism. J Pediatr Gastroenterol Nutr, 56(4), 355-363. doi:10.1097/MPG.0b013e31827e1f07 Page-Goertz, S. (2007). Hypoglycemia in the breastfeeding newborn. International Lactaction Consultant Association, 1-9. Pereira, P. C. (2014). Milk nutritional composition and its role in human health. Nutrition, 30(6), 619-627. doi:10.1016/j.nut.2013.10.011 Pitt, J. J. (2009). Principles and applications of liquid chromatography-mass spectrometry in clinical biochemistry. Clin Biochem Rev, 30(1), 19-34.  128  Postic, C., Shiota, M., & Magnuson, M. A. (2001). Cell-specific roles of glucokinase in glucose homeostasis. Recent Prog Horm Res, 56, 195-217.  Puiman, P., & Stoll, B. (2008). Animal models to study neonatal nutrition in humans. Curr Opin Clin Nutr Metab Care, 11(5), 601-606. doi:10.1097/MCO.0b013e32830b5b15 Pérez-Cano, F. J., Franch, À., Castellote, C., & Castell, M. (2012). The suckling rat as a model for immunonutrition studies in early life. Clin Dev Immunol, 2012, 537310. doi:10.1155/2012/537310 Quinn, R. (2005). Comparing rat's to human's age: how old is my rat in people years? Nutrition, 21(6), 775-777. doi:10.1016/j.nut.2005.04.002 Kliegman, R.M., & Sparks, J.W. (1985). Perinatal galactose metabolism. The journal of Paediatrics, 107(6), 831-841.  Ringnér, M. (2008). What is principal component analysis? Nat Biotechnol, 26(3), 303-304. doi:10.1038/nbt0308-303 Roberts, L. D., Souza, A. L., Gerszten, R. E., & Clish, C. B. (2012). Targeted metabolomics. Curr Protoc Mol Biol, Chapter 30, Unit 30.32.31-24. doi:10.1002/0471142727.mb3002s98 Rogers, S., Guerra, M., & Segal, S. (1983). Galactose metabolism in suckling and adult isolated rat hepatocytes. Pediatr Res, 17(8), 609-616. doi:10.1203/00006450-198308000-00001 Roser, M., Josic, D., Kontou, M., Mosetter, K., Maurer, P., & Reutter, W. (2009). Metabolism of galactose in the brain and liver of rats and its conversion into glutamate and other amino acids. J Neural Transm (Vienna), 116(2), 131-139. doi:10.1007/s00702-008-0166-9 129  Rui, L. (2014). Energy metabolism in the liver. Compr Physiol, 4(1), 177-197. doi:10.1002/cphy.c130024 Ruiz-Matute, A. I., Hernández-Hernández, O., Rodríguez-Sánchez, S., Sanz, M. L., & Martínez-Castro, I. (2011). Derivatization of carbohydrates for GC and GC-MS analyses. J Chromatogr B Analyt Technol Biomed Life Sci, 879(17-18), 1226-1240. doi:10.1016/j.jchromb.2010.11.013 Saltiel, A. R., & Kahn, C. R. (2001). Insulin signalling and the regulation of glucose and lipid metabolism. Nature, 414(6865), 799-806. doi:10.1038/414799a Sanz, M. L., Diez-Barrio, M. T., Sanz, J., & Martínez-Castro, I. (2003). GC Behavior of disaccharide trimethylsilyl oximes. J Chromatogr Sci, 41(4), 205-208. doi:NO_DOI Schauer, N., Steinhauser, D., Strelkov, S., Schomburg, D., Allison, G., Moritz, T., . . . Kopka, J. (2005). GC-MS libraries for the rapid identification of metabolites in complex biological samples. FEBS Lett, 579(6), 1332-1337. doi:10.1016/j.febslet.2005.01.029 Segal, S., & Foley, J. (1958). The metabolism of D-ribose in man. J Clin Invest, 37(5), 719-735. doi:10.1172/JCI103658 Smilowitz, J. T., O'Sullivan, A., Barile, D., German, J. B., Lönnerdal, B., & Slupsky, C. M. (2013). The human milk metabolome reveals diverse oligosaccharide profiles. J Nutr, 143(11), 1709-1718. doi:10.3945/jn.113.178772 Staley, M. D., Gibson, C. A., Herbein, J. F., Grosvenor, C. E., & Baumrucker, C. R. (1998). Rat milk and dietary long arginine3 insulin-like growth factor I promote intestinal growth of newborn rat pups. Pediatr Res, 44(4), 512-518. doi:10.1203/00006450-199810000-00008 Stoffels, B., Türler, A., Schmidt, J., Nazir, A., Tsukamoto, T., Moore, B. A., . . . Bauer, A. J. 130  (2011). Anti-inflammatory role of glycine in reducing rodent postoperative inflammatory ileus. Neurogastroenterol Motil, 23(1), 76-87, e78. doi:10.1111/j.1365-2982.2010.01603.x Sumner, L. W., Mendes, P., & Dixon, R. A. (2003). Plant metabolomics: large-scale phytochemistry in the functional genomics era. Phytochemistry, 62(6), 817-836.  Syggelou, A., Iacovidou, N., Atzori, L., Xanthos, T., & Fanos, V. (2012). Metabolomics in the developing human being. Pediatr Clin North Am, 59(5), 1039-1058. doi:10.1016/j.pcl.2012.07.002 Tsugawa, H., Bamba, T., Shinohara, M., Nishiumi, S., Yoshida, M., & Fukusaki, E. (2011). Practical non-targeted gas chromatography/mass spectrometry-based metabolomics platform for metabolic phenotype analysis. J Biosci Bioeng, 112(3), 292-298. doi:10.1016/j.jbiosc.2011.05.001 Turck, D. (2007). Soy protein for infant feeding: what do we know? Curr Opin Clin Nutr Metab Care, 10(3), 360-365. doi:10.1097/MCO.0b013e3280fa821b Urashima, T., Fukuda, K., & Messer, M. (2012). Evolution of milk oligosaccharides and lactose: a hypothesis. Animal, 6(3), 369-374. doi:10.1017/S1751731111001248 Villas-Bôas, S. G., Mas, S., Akesson, M., Smedsgaard, J., & Nielsen, J. (2005). Mass spectrometry in metabolome analysis. Mass Spectrom Rev, 24(5), 613-646. doi:10.1002/mas.20032 Villas-Bôas, S. G., Rasmussen, S., & Lane, G. A. (2005). Metabolomics or metabolite profiles? Trends Biotechnol, 23(8), 385-386. doi:10.1016/j.tibtech.2005.05.009 Wang, W., Wu, Z., Dai, Z., Yang, Y., Wang, J., & Wu, G. (2013). Glycine metabolism in animals and humans: implications for nutrition and health. Amino Acids, 45(3), 463-131  477. doi:10.1007/s00726-013-1493-1 Weckwerth, W., & Fiehn, O. (2002). Can we discover novel pathways using metabolomic analysis? Curr Opin Biotechnol, 13(2), 156-160.  Weckwerth, W., & Morgenthal, K. (2005). Metabolomics: from pattern recognition to biological interpretation. Drug Discov Today, 10(22), 1551-1558. doi:10.1016/S1359-6446(05)03609-3 Weinert, D. J. (2009). Nutrition and muscle protein synthesis: a descriptive review. J Can Chiropr Assoc, 53(3), 186-193.  Wight, N., Marinelli, K. A., & Medicine, A. o. B. (2014). ABM clinical protocol #1: guidelines for blood glucose monitoring and treatment of hypoglycemia in term and late-preterm neonates, revised 2014. Breastfeed Med, 9(4), 173-179. doi:10.1089/bfm.2014.9986 Wineman, P. L., & Keto, R. O. (1994). Target-compound method for the analysis of accelerant residues in fire debris. Analytica Chimica Acta, 288(1-2), 97-110.  World Health Organization. Exclusive Breastfeeding for six months best for babies everywhere.(2011)Retrievedfrom https://www.who.int/mediacentre/news/statements/2011/breastfeeding-20110115/en/ Wright, E. M., Hirayama, B. A., & Loo, D. F. (2007). Active sugar transport in health and disease. J Intern Med, 261(1), 32-43. doi:10.1111/j.1365-2796.2006.01746.x Wu, G. (2010). Functional amino acids in growth, reproduction, and health. Adv Nutr, 1(1), 31-37. doi:10.3945/an.110.1008 Xu, X., So, J. S., Park, J. G., & Lee, A. H. (2013). Transcriptional control of hepatic lipid metabolism by SREBP and ChREBP. Semin Liver Dis, 33(4), 301-311. doi:10.1055/s-0033-1358523 132  Yamamoto, H., Fujimori, T., Sato, H., Ishikawa, G., Kami, K., & Ohashi, Y. (2014). Statistical hypothesis testing of factor loading in principal component analysis and its application to metabolite set enrichment analysis. BMC Bioinformatics, 15, 51. doi:10.1186/1471-2105-15-51 Yee, J. K., Wahjudi, P. N., Vega, J., Lim, S., Martin, A., Patterson, M. E., . . . Lee, W. N. (2013). Stearoyl-CoA desaturase enzyme 1 inhibition reduces glucose utilization for de novo fatty acid synthesis and cell proliferation in 3T3-L1 adipocytes. Metabolomics, 9(4), 809-816. doi:10.1007/s11306-013-0511-3 Zeisel, S. H., Freake, H. C., Bauman, D. E., Bier, D. M., Burrin, D. G., German, J. B., . . . Rasmussen, K. M. (2005). The nutritional phenotype in the age of metabolomics. J Nutr, 135(7), 1613-1616.  Zhang, A., Sun, H., Wang, P., Han, Y., & Wang, X. (2012). Modern analytical techniques in metabolomics analysis. Analyst, 137(2), 293-300. doi:10.1039/c1an15605e Zhao, Y. Y., Wang, H. L., Cheng, X. L., Wei, F., Bai, X., Lin, R. C., & Vaziri, N. D. (2015). Metabolomics analysis reveals the association between lipid abnormalities and oxidative stress, inflammation, fibrosis, and Nrf2 dysfunction in aristolochic acid-induced nephropathy. Sci Rep, 5, 12936. doi:10.1038/srep12936       133  APPENDIX A  NUTRIENT COMPOSITION OF RAT MILK FORMULA  Component  Amount per liter formula   acid casein 70.00 g/L whey 47.00 g/L coconut oil 20 g/L corn oil  41 g/L soy oil 26 g/L MCT oil 53 g/L lactose 35 g/L corn starch 1.87 g/L ZnSO4 113.79 mg/L carnitine 4.00 mg/L creatine 70.00 mg/L ethanolamine 1.13 mg/L sodium hydroxide 1.27 g/L potassium phosphate 5.05 g/L citric acid 1.04 g/L Magnesium sulfate 0.91 g/L Iron sulfate 50.52 mg/L potassium iodide 3.67 mg/L sodium fluoride 3.11 mg/L aluminum sulfate 1.97 mg/L Manganese sulfate 0.54 mg/L Calcium carbonate 6.031 g/L calcium gluconate 1.365 g/L iron sulfate 20 mg/L riboflavin 9.31 mg/L niacin 14.50 mg/L pyridoxal 7.75 mg/L inositol 518.43 mg/L arginine 500.00 mg/L glycine 300.00 mg/L taurine 150.00 mg/L picolinic acid 20.00 mg/L    p-Aminobenzoic acid 44.05 mg/L vitamin C (coated) 406.64 mg/L biotin 0.18 mg/L 134  vitamin B12 11.89 mg/L Calcium pantothenate 26.43 mg/L choline dihydrogen citrate 1398.77 mg/L folic acid 0.79 mg/L inositol 44.05 mg/L vitamin K3 19.82 mg/L niacin 39.65 mg/L pyridoxine hydrochloride 8.81 mg/L riboflavin 8.81 mg/L thiamin hydrochloride 8.81 mg/L vitamin A palmitate 7929.60 IU/L Vitamin D3 881.00 IU/L vitamin E 48.46 IU/L                        135  APPENDIX B  WESTERN BLOTTING ANALYSIS DATA Day 10  ChREBP  Figure 42a. Western blot showing abundance of ChREBP in day 10 rat liver samples, with no difference in expression between lactose-fed and CSS-fed group. β-Actin was used as internal control.  Figure 42b. Day 10 liver bar graph (ChREBP/Actin ratio) depicting ChREBP expression in lactose-fed group (0.605) compared to CSS-fed group of rat pups (0.660), by Western Blotting. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) No difference found between groups (p=7268). ChREBP, carbohydrate response element-binding protein, CSS, glucose-corn syrup solids. SD, standard deviation   0.6050.66000.10.20.30.40.50.60.70.80.9ChREBP/Actin Lactose CSS92 kDa 43 kDa  136  SREBP1-c  Figure 43a. Western blot showing abundance of SREBP1-c in day 10 rat liver samples, with no difference in expression between lactose-fed and CSS-fed group. β-Actin was used as internal control. Figure 43b. Day 10 liver bar graph (SREBP1-c/Actin ratio) depicting SREBP1-c expression in lactose-fed group (1.031) compared to CSS-fed group of rat pups (1.076), by Western Blotting. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) No difference found between groups (p=0.5669). SREBP, sterol regulatory element-binding protein, CSS, glucose-corn syrup solids. SD, standard deviation   1.031 1.07600.20.40.60.811.21.41.61.82SREBP1-c/Actin ratioLactose CSS68 kDa 43 kDa  137  SCD  Figure 44a. Western blot showing abundance of SCD in day 10 rat liver samples, with no difference in expression between lactose-fed and CSS-fed group. β-Actin was used as internal control.  Figure 44b. Day 10 liver bar graph (SCD/Actin ratio) depicting SCD expression in lactose-fed group (1.233) compared to CSS-fed group of rat pups (1.300), by Western Blotting. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) No difference found between groups (p=0.7320). SCD, stearoyl CoA desaturase. CSS, glucose-corn syrup solids. SD, standard deviation   1.233 1.30000.20.40.60.811.21.41.61.82SCD/Actin ratioLactose CSS40 kDa 43 kDa  138  Day 14 ChREBP  Figure 45a. Western blot showing abundance of ChREBP in day 14 rat liver samples, with no difference in expression between lactose-fed and CSS-fed group. β-Actin was used as internal control.  Figure 45b. Day 14 liver bar graph (ChREBP/Actin ratio) depicting ChREBP expression in lactose-fed group (1.000) compared to CSS-fed group of rat pups (1.051), by Western Blotting. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) No difference found between groups (p=0.1988). ChREBP, carbohydrate response element-binding protein, CSS, glucose-corn syrup solids. SD, standard deviation  1.0001.05100.20.40.60.811.21.41.61.82ChREBP/Actin ratioLactose CSS92 kDa   43 kDa  139  SREBP1-c    Figure 46a. Western blot showing abundance of SREBP1-c in day 14 rat liver samples, with no difference in expression between lactose-fed and CSS-fed group. β-Actin was used as internal control. Figure 46b. Day 14 liver bar graph (SREBP1-c/Actin ratio) depicting SREBP1-c expression in lactose-fed group (0.965) compared to CSS-fed group of rat pups (1.001), by Western Blotting. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) No difference found between groups (p=0.4045). SREBP, sterol regulatory element-binding protein. CSS, glucose-corn syrup solids. SD, standard deviation     0.9651.00100.20.40.60.811.21.41.61.82SREBP1-c/Actin ratioLactose CSS68 kDa 43 kDa  140  SCD  Figure 47a. Western blot showing abundance of SCD in day 14 rat liver samples, with no difference in expression between lactose-fed and CSS-fed group. β-Actin was used as internal control.  Figure 47b. Day 14 liver bar graph (SCD/Actin ratio) depicting SCD expression in lactose-fed group (1.055) compared to CSS-fed group of rat pups (1.118), by Western Blotting. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) No difference found between groups (p=0.4725). SCD, stearoyl CoA desaturase. CSS, glucose-corn syrup solids. SD, standard deviation   1.0551.11800.20.40.60.811.21.41.61.82SCD/Actin ratioLactose CSS40 kDa 43 kDa  141  Day 18            ChREBP           Figure 48a. Western blot showing abundance of ChREBP in day 18 rat liver samples, with no difference in expression between lactose-fed and CSS-fed group. β-Actin was used as internal control  Figure 48b. Day 18 liver bar graph (ChREBP/Actin ratio) depicting ChREBP expression in lactose-fed group (0.605) compared to CSS-fed group of rat pups (0.660), by Western Blotting. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) No difference found between groups (p=0.8520). ChREBP, carbohydrate response element-binding protein, CSS, glucose-corn syrup solids. SD, standard deviation  0.896 0.91600.20.40.60.811.21.41.61.82ChREBP/Actin ratioLactose CSS92 kDa 43 kDa  142  SREBP1-c    Figure 49a. Western blot showing abundance of SREBP1-c in day 18 rat liver samples, with no difference in expression between lactose-fed and CSS-fed group. β-Actin was used as internal control   Figure 49b. Day 18 liver bar graph (SREBP1-c/Actin ratio) depicting SREBP1-c expression in lactose-fed group (1.038) compared to CSS-fed group of rat pups (1.036), by Western Blotting. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) No difference found between groups (p=0.9862). SREBP, sterol regulatory element-binding protein. CSS, glucose-corn syrup solids. SD, standard deviation  1.038 1.03600.20.40.60.811.21.41.61.82SREBP1-c/Actin ratioLactose CSS68 kDa 43 kDa  143  SCD  Figure 50a. Western blot showing abundance of SCD in day 18 rat liver samples, with no difference in expression between lactose-fed and CSS-fed group. β-Actin was used as internal control   Figure 50b. Day 18 liver bar graph (SCD/Actin ratio) depicting SCD expression in lactose-fed group (0.966) compared to CSS-fed group of rat pups (0.953), by Western Blotting. Data presented as mean ± SD and comparison made based on independent sample’s t-test (n=6 per group) No difference found between groups (p=0.8059). SCD, stearoyl CoA desaturase. CSS, glucose-corn syrup solids. SD, standard deviation  0.966 0.95300.20.40.60.811.21.41.61.82SCD/Actin ratioLactose CSS40 kDa 43 kDa  

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.24.1-0223696/manifest

Comment

Related Items