Nutritional modulation of hepatic lipid metabolism in health and disease Richard Johnston, BmedSci, BM BS, MRCP Thesis submitted to the University of Nottingham for the degree of Doctor of Medicine. February 2012 1 Abstract The objective of this thesis was to assess the impact of altering macronutrient intakes on hepatic lipid metabolism. Two separate studies were performed, with liver triglyceride content being the principal outcome of both. In the first study 32 healthy and centrally overweight males were randomised to 2 periods, each of 2 weeks, of either a high fructose or glucose intake in a non-crossover fashion. Isoenergetic status was maintained by providing foodstuffs during the first period, followed by a 6 week washout and then a second period of ad libitum overfeeding. In the second study 55 patients with biopsy proven non-alcoholic fatty liver disease were randomised to 3 months 5g a day of capsules containing either n-3 polyunsaturated fatty acid or oleic enriched sunflower oil. The main findings are summarised. High intakes of fructose and glucose in the isoenergetic period resulted in a stable weight, and no change in hepatic, serum and ectopic triglyceride content. There was a raised serum uric acid with fructose. During the hyperenergetic period there was a tendency for greater uric acid with fructose, whilst both groups had a matched weight gain, elevation of liver biochemistry and an increase in hepatic, serum and muscle triglycerides. Changes in liver biochemistry and triglycerides were associated with changes in weight. During both periods there was calorimetric evidence for a shift in whole body metabolism towards that reflective of a high carbohydrate intake. There was no alteration in renal function or cardiovascular haemodynamic parameters or consistent change in insulin resistance. The n-3 polyunsaturated versus oleic acid study resulted in significant alterations of serum fatty acid profiles between the groups, which were in line with the capsules’ contents. These changes however failed to translate, in the whole group, to any detected metabolic or hepatic changes beyond a reduction in serum triglyceride with n-3 polyunsaturated fatty acids. Only 43 of the 55 patients had elevated liver triglycerides on baseline MRI. Amongst this 43 there was a reduction in liver triglyceride with n-3 polyunsaturated fatty acids, but no other associated metabolic changes. The uric acid findings support the notion of fructose and glucose differing in their pre triose metabolism. There was however no differing outcomes in terms 2 of lipid synthesis or storage. There was a suggestion of reduced liver triglycerides with n-3 polyunsaturated fatty acids though this was an isolated result only found amongst those with a steatotic liver at baseline. Ultimately the exquisite sensitivity of the liver to nutrient intakes was highlighted by the 0.8% gain in weight in the fructose / glucose study resulting in a 24% increase in liver lipid. This affirms the notion that dietary energy intakes have a profound influence on hepatic metabolism, but there is no evidence from this thesis that this influence is macronutrient specific. In the future macronutrient comparisons need to be made. 3 Acknowledgements I am very grateful for the hard work of many involved in this thesis. The process of leading a large and varied team down a long and complex road has been hugely challenging, educational and ultimately rewarding. The work presented was carried out by me, with support as acknowledged below. I was integrally involved in the development and design of the studies, and personally applied for financial and ethical support for them. I personally contacted, assessed and recruited all the subjects. I designed the foodstuff plans and devised the home delivery orders. I was present at every assessment, though was assisted in performing some of the hyperinsulinaemic euglycaemic clamps by Ian Macdonald. I performed all the statistical analyses and analysed the raw data originating from the clinical assessments, food diaries, satiety assessments, 1H MRS liver and calf data, indirect calorimetry, finometry, and hyperinsulinaemic euglycaemic clamps. First, and foremost, I would have got nowhere without the endless belief and encouragement from Rosie, my wife. She has made an enormous contribution to my completion of the thesis by supporting me through the anti-social hours of data collection and analyses. The next key players in these studies were my supervisors: Ian Macdonald, Guru Aithal, and Moira Taylor. Without such excellent stewardship of my, sometimes overly ambitious, plans things would never have got going. I am privileged to have received such teaching, support and encouragement. I look forward to future work together. In order to minimise disruption to the subjects’ day most magnetic resonance scans were performed very early in the morning. For this I am very grateful. This data was mainly collated by Mary Stephenson who was aided by Elisa Placidi and Eleanor Cox. Mary also analysed the liver volume and 31P data. Professor Peter Morris aided Mary in local development of some of the protocols. The support and adoption of these studies by the Biomedical Research Unit was instrumental to their successful completion. In particular, I would like to thank Paul Roach and Rosemary Dainty whose ever-warm personalities and logistical advice helped keep me focused on the business side of the projects. Vic Shepherd and Lisa Chalkley were extremely accommodating to patient reviews occurring at all times of the day. Tracey Wildsmith put in many long hours and early starts. She provided immense practical support, and made numerous cups of tea, during the arduous running of the fructose versus glucose study. I would like to thank all the team from the human physiology laboratory for their superb skills and facilities. In particular I would like to thank Sara Brown for her assistance which often went beyond her designated duties. She performed all of the DEXA scans and smoothed the complex logistics of delivering nearly 2,000 meals to individuals’ doorsteps and preparing nearly 200 breakfast or lunches within the laboratory. Always heard before she is seen, her bubbly personality kept the subjects and I entertained for hours during the study visits. 4 The advice of Liz Simpson was critical in aiding a gastroenterologist to perform metabolic physiology assessments. My studies were heavily reliant on the clinical trials pharmacy. Sheila Hodgson and team counted up numerous capsules into individual bottles for the n- 3 PUFA study, whereas Paul Douglas prepared endless tracer infusions early in the morning. Within the University, Sally Cordon performed all the fructose versus glucose serum and plasma analyses with her customary efficiency, calmness and warmth. Hannah Crossland, under the supervision of Kenny Smith, performed highly efficient analyses of deuterated glucose enrichment of serum samples. Within the hospital support came from Doctor Philip Kaye reviewing the n-3 PUFA liver biopsy slides, and the hepatologists who facilitated the recruitment of their patients. The n-3 PUFA serum data was analysed in the hospital laboratories. External support and assistance came from Professor Philip Calder and Annette West in Southampton University. They analysed the serum fatty acid profiles for the n-3 PUFA study. The financial support that I have obtained for this work comes from a variety of sources. Internally Professors Ian Macdonald and Chris Hawkey have been extremely generous. Professor Bardhan provided a large external support for the n-3 PUFA work. I am so grateful to him and his enthusiasm. His enablement of the n-3 PUFA project was a clear stepping stone to me securing a fellowship with core charity and the nutrition research foundation. The fellowship provided me with great encouragement. Reto Muggli and Peter Clough from Wassen International supplied the n-3 PUFA and placebo capsules. Finally, I need to thank all the subjects and patients involved in the studies. Ultimately it is them that made the work possible, and indeed so fun. 5 Contents 1 Introduction 13 1.1 The function of adipose tissue 13 1.2 Overview of hepatic fatty acid metabolism in health 15 1.3 Introduction to Non-Alcoholic Fatty Liver Disease (NAFLD) 18 1.4 NAFLD epidemiology 18 1.4.1. Prevalence of NAFLD 18 1.4.2 Natural history of NAFLD 19 1.5. NAFLD aetiology 22 1.5.1. 1st hit – Hepatic steatosis development 22 1.5.1.1. Hepatic steatosis and insulin resistance 24 1.5.1.2. Hepatic steatosis and obesity 27 1.5.1.3. Hepatic steatosis and genetics 27 1.5.2. 2nd hit – Steatohepatitis development 28 1.6 Medical and weight loss therapy for NAFLD 29 1.6.1. Lifestyle and weight loss therapy 30 1.6.2. Promoting insulin sensitivity an antioxidants 31 1.6.3. Dyslipidaemia treatments 32 1.7. Interactions between diet and NAFLD 32 1.7.1. Dietary history assessments in NAFLD 33 1.7.2. Dietary assessment in NAFLD employing liver biopsy lipid analyses 36 1.8. Polyunsaturated fatty acids 38 1.8.1. Classification of polyunsaturated fatty acids (PUFAs) 38 1.8.2. Dietary sources of n-3 PUFAs 38 1.8.2.1. Intake trends for PUFAs and the n-6 to n-3 ratio 39 1.8.2.2. Determining adequacy of long-chain n-3 PUFA intakes 40 1.8.3. Bioavailability of long chain PUFAs 41 1.8.4 Biological activity of PUFAs in relation to hepatic fatty acid 42 metabolism 1.8.4.1. PUFAs and insulin resistance 42 1.8.4.2. PUFAs and immunomodulation 44 1.8.4.3. PUFAs and gene expression 45 1.8.4.3.1. n-3 PUFAs and peroxisome proliferator-activated receptors (PPARs) 45 1.8.4.3.2. n-3 PUFAs and sterol regulatory element binding proteins (SREBPs) 46 1.8.5.