Phenotypic characterization of genetic variants associated with the risk of type 2 diabetes and non-alcoholic fatty liver disease

(1)

DISSERTATIONS | LILIAN FERNANDES SILVA | PHENOTYPIC CHARACTERIZATION OF GENETIC... | No 549

uef.fi

PUBLICATIONS OF

THE UNIVERSITY OF EASTERN FINLAND

Dissertations in Health Sciences

PUBLICATIONS OF

THE UNIVERSITY OF EASTERN FINLAND

LILIAN FERNANDES SILVA

Metabolic diseases, including obesity, type 2 diabetes (T2D) and non-alcoholic fatty liver disease (NAFLD), are increasing worldwide.

This thesis reports the metabolites associated with genetic variants linked to the risk of T2D and NAFLD. New associations of a GCKR

variant with lipids were found, as well as the multiple metabolic pathways disrupted

in NAFLD. Understanding the genetic architecture of T2D and NAFLD is essential

to understand the etiology and to improve diagnosis, prognosis, and therapy of

these diseases.

LILIAN FERNANDES SILVA

(2)

(3)

PHENOTYPIC CHARACTERIZATION OF GENETIC

VARIANTS ASSOCIATED WITH THE RISK OF TYPE 2

DIABETES AND NON-ALCOHOLIC FATTY LIVER DISEASE

(4)

(5)

Lilian Fernandes Silva

PHENOTYPIC CHARACTERIZATION OF GENETIC VARIANTS ASSOCIATED WITH THE RISK OF TYPE 2 DIABETES AND NON-ALCOHOLIC FATTY LIVER DISEASE

To be presented by permission of the

Faculty of Health Sciences, University of Eastern Finland for public examination in xx Auditorium, Kuopio

on February 14^th, 2020, at 12 o’clock noon Publications of the University of Eastern Finland

Dissertations in Health Sciences No 549

Department of Medicine, Institute of Clinical Medicine, School of Medicine, Faculty of Health Sciences, University of Eastern Finland

Kuopio

(6)

Series Editors

Professor Tomi Laitinen, M.D., Ph.D.

Institute of Clinical Medicine, Clinical Physiology and Nuclear Medicine Faculty of Health Sciences

Associate professor (Tenure Track) Tarja Kvist, Ph.D.

Department of Nursing Science Faculty of Health Sciences Professor Kai Kaarniranta, M.D., Ph.D.

Institute of Clinical Medicine, Ophthalmology Faculty of Health Sciences

Associate Professor (Tenure Track) Tarja Malm, Ph.D.

A.I. Virtanen Institute for Molecular Sciences Faculty of Health Sciences

Lecturer Veli-Pekka Ranta, Ph.D.

School of Pharmacy Faculty of Health Sciences

Distributor:

University of Eastern Finland Kuopio Campus Library

P.O.Box 1627 FI-70211 Kuopio, Finland

www.uef.fi/kirjasto

Grano Oy Jyväskylä, 2020

ISBN: 978-952-61-3292-1 (print/nid.) ISBN: 978-952-61-3293-8 (PDF)

ISSNL: 1798-5706 ISSN: 1798-5706 ISSN: 1798-5714 (PDF)

(7)

Author’s address: Institute of Clinical Medicine, Internal Medicine, School of Medicine, Faculty of Health Sciences

University of Eastern Finland P.O. Box 1627

70210 KUOPIO, FINLAND Email: lilian.fernandes.silva@uef.fi Doctoral program: Doctoral program in Molecular Medicine Supervisors: Professor Markku Laakso, M.D., Ph.D.

Institute of Clinical Medicine, Internal Medicine, School of Medicine, Faculty of Health Sciences

University of Eastern Finland and Kuopio University Hospital KUOPIO

FINLAND

Jagadish Vangipurapu, Ph.D.

University of Eastern Finland KUOPIO

FINLAND

Docent Tarja Kokkola, Ph.D.

University of Eastern Finland KUOPIO

FINLAND

Reviewers: Professor Hannu Järveläinen, M.D., Ph.D.

Department of Internal Medicine University of Turku

TURKU FINLAND

Docent Jorma T. Lahtela, MD, Ph.D.

Department of Internal Medicine University of Tampere

TAMPERE FINLAND

Opponent: Professor Risto Kaaja, MD, Ph.D.

Department of Department of Clinical Medicine

(8)

(9)

Fernandes Silva, Lilian

Phenotypic characterization of genetic variants associated with the risk of type 2 diabetes and non-alcoholic fatty liver disease

Kuopio: University of Eastern Finland

Publications of the University of Eastern Finland Dissertations in Health Sciences 549. 2020, 89 p.

ISBN: 978-952-61-3292-1 (print) ISSNL: 1798-5706

ISSN: 1798-5706

ISBN: 978-952-61-3293-8 (PDF) ISSN: 1798-5714 (PDF)

ABSTRACT

Metabolic diseases, including obesity, type 2 diabetes (T2D) and non-alcoholic fatty liver disease (NAFLD), are increasing worldwide in epidemic proportions.

Therefore, understanding the pathophysiology and genetics of these diseases is of great importance. Our studies focused on the glucokinase regulatory protein gene (GCKR), a risk gene for T2D, and several genetic variants associated with the risk of NAFLD. The genetic variant GCKR rs780094 has previously been associated with lactate levels in the fasting state. In Study I, we investigated the association of GCKR rs780094 with lactate levels in a frequently sampled oral glucose tolerance test in the participants of the METSIM study to evaluate the effects of increasing GCKR expression on lactate production in liver cells. The C allele of GCKR rs780094 was associated with lower lactate levels while fasting but increased lactate levels during hyperglycemia, independently of insulin levels. Increased expression of GKRP induced higher lactate levels in HepG2 cells and in human primary hepatocytes upon glucose stimulation by increasing the amount of GCK. Our results suggest that the association of GCKR rs780094 with lactate levels may involve differential GCKR expression between the carriers of the C and T alleles. In Study II, we applied a metabolomics approach to measure metabolites in the participants of the METSIM study to investigate the associations of rs780094 of GCKR with metabolites. We found novel negative associations of the T allele of GCKR rs780094 with serine and threonine, and positive associations with two metabolites of tryptophan, indolelactate and N-acetyltryptophan. We also found novel significant positive associations of this genetic variant with 12 glycerolipids and 19 glycerophospholipids, and significant negative associations with three glycerophospholipids and two sphingolipids. Our study adds new information about the pleiotropy of GCKR and shows the associations of the T allele of GCKR rs780094 with lipids. In Study III, we investigated the association of several genetic variants increasing the risk for NAFDL (PNPLA3, TM6SF2, MBOAT7, GCKR,

(10)

SAMM50, MnSOD/SOD2, PEMT, LEPR) and two genetic variants associated with decreased risk of NAFLD (PPP1R3B, HSD17B13) with metabolites in the participants of the METSIM study. We identified multiple statistically significant associations of genetic variants in seven genes (PNPLA3, TM6SF2, GCKR, MBOAT7, SAMM50, PPP1R3B, HDS17B13) with metabolites reflecting different metabolic pathways regulating liver fat content. Our study adds novel information about the pathways and metabolites affected by multiple genetic variants associated with the risk of NAFLD. These findings demonstrate that multiple metabolic pathways are disrupted in NAFLD. Understanding the pathophysiology of NAFLD is important for the planning of appropriate diagnostic and treatment approaches for this disease.

Studies I-III add important knowledge for the understanding of the pathophysiology and genetics of T2D and NAFLD.

National Library of Medicine Classification: QU 477, QU 500, WI 710, WK 810 Medical Subject Headings: Diabetes Mellitus, Type 2; Non-alcoholic Fatty Liver Disease; Risk; Genes; Genetic Variation; Genetics; Gene Expression; Alleles;

Phenotype; Metabolomics; Metabolic Networks and Pathways; Glucokinase; Glucose Tolerance Test; Lactic Acid; Fasting; Hyperglycemia; Hep G2 Cells; Hepatocytes.

(11)

Fernandes Silva, Lilian

Phenotypic characterization of genetic variants associated with the risk of type 2 diabetes and non-alcoholic fatty liver disease

Kuopio: Itä-Suomen yliopisto

Publications of the University of Eastern Finland Dissertations in Health Sciences 549. 2020, 89 s.

ISBN: 978-952-61-3292-1 (print) ISSNL: 1798-5706

ISSN: 1798-5706

ISBN: 978-952-61-3293-8 (PDF) ISSN: 1798-5714 (PDF)

TIIVISTELMÄ

Metaboliset sairaudet, ylipaino, tyypin 2 diabetes ja ei-alkoholiperäinen rasvamaksatauti lisääntyvät maailmanlaajuisesti epidemian tavoin. Tämän vuoksi näiden sairauksien patofysiologian ja genetiikan ymmärtäminen on tärkeää.

Tutkimme tyypin 2 diabeteksen riskigeeniä, glukokinaasia säätelevää proteiinia (glucokinase regulatory protein gene, GCKR), sekä rasvamaksataudin riskiin liittyviä geenejä. GCKR rs780094-C geenimuutoksen on aikaisemmissa tutkimuksissa havaittu lisäävän laktaatin pitoisuutta paastossa. Ensimmäisessä tutkimuksessamme tutkimme laktaattipitoisuutta glukoosirasitustestissä METSIM-kohortissa. Samassa tutkimuksessa selvitimme myös, miten GCKR-ekspression lisääntyminen liittyy laktaatin tuottamiseen maksassa. GCKR rs780094-geenimuutoksen C-alleeli liittyi pienentyneeseen laktaattipitoisuuteen paastossa, mutta lisääntyneeseen laktaattipitoisuuteen glukoosistimulaation aikana riippumatta insuliinipitoisuudesta. GKRP-ekspression lisääminen glukokinaasin avulla liittyi laktaattipitoisuuden lisääntymiseen HepG2- maksasoluissa sekä ihmisen maksasoluissa. Tulokset viittaavat siihen, että GCKR rs780094-geenimuutoksen vaikutus laktaattipitoisuuteen on erilainen C- ja T-alleelien välillä. Toisessa tutkimuksessa selvitimme GCKR rs780094-geenimuutoksen vaikutuksia metaboliittin pitoisuuksiin METSIM-tutkimuksessa. Geenimuutoksen T-alleeli liittyi negatiivisesti aminohappojen seriini ja treoniini pitoisuuksiin, mutta positivisesti tryptofaanin pitoisuuteen sekä tämän hajoamistuotteiden pitoisuuteen. Lisäksi mainittu geenimuutos liittyi positivisesti 12 glyserolipidin ja 19 glyserofosfolipidin, ja negatiivisesti kolmen glyserofosfolipidin, ja kahden sfingolipidin pitoisuuksiin.

Tutkimuksemme osoittaa, että GCKR-geeni on pleiotrofinen eli liittyy erittäin monen metaboliittien pitoisuuksiin, jotka säätelevät erityisesti lipidimetaboliaa.

Kolmannessa tutkimuksessa selvitimme rasvamaksataudin riskiin liittyvien geenimuutosten (PNPLA3, TM6SF2, MBOAT7, GCKR, SAMM50, MnSOD/SOD2, PEMT, LEPR), ja siltä suojaavien geenimuutosten (PPP1R3B, HSD17B13) liittymistä metaboliittien pitoisuuksiin METSIM-tutkimuksessa. Tutkimuksemme mukaan useat geenimuutokset seitsemässä geenissä (PNPLA3, TM6SF2, GCKR, MBOAT7,

(12)

SAMM50, PPP1R3B, HDS17B13) liittyivät useisiin metaboliitteihin, jotka säätelevät maksan rasvoittumisen mekanismeja. Tutkimuksemme osoitti, että maksan rasvoittumista säätelevät useat geenit, jotka vaikuttavat moneen metaboliatiehen.

Tutkimuksemme auttaa ymmärtämään maksan rasvoittumisen mekanismia, joista on mahdollisesti hyötyä rasvamaksataudin hoidon kehittämisessä. Yhteenvetona voidaan todeta, että tutkimuksemme antaa tärkeää lisätietoa sekä tyypin 2 diabeteksen että rasvamaksan patofysiologiasta ja genetiikasta.

Luokitus: QU 477, QU 500, WI 710, WK 810

Yleinen suomalainen ontologia: aikuistyypin diabetes; rasvamaksa; riskitekijät;

geenit; geneettinen muuntelu; geeniekspressio; fenotyyppi; aineenvaihdunta;

aineenvaihduntatuotteet; maitohappo; paasto; hyperglykemia; soluviljely.

(13)

ACKNOWLEDGEMENTS

This study was performed in the Institute of Clinical Medicine, School of Medicine, Faculty of Health Sciences, University of Eastern Finland.

I would like to thank Professor Markku Laakso who besides being my supervisor, was also my mentor on this journey. The door to his office was always open whenever I ran into trouble or had a question about my research or writing. Not only that he consistently encouraged me to develop my own ideas and to improve my abilities, but he also steered me in the right direction whenever he thought it was needed. His expert guidance and enthusiasm towards research helped me to keep motivated during the hard times. His constant effort in nurturing my self-confidence helped me to look beyond the horizons with optimism.

I feel extremely grateful for my second supervisor, Ph.D. Jagadish Vangipurapu for his guidance and endless patience to teach me statistics. I could not have imagined accomplishing this work without his help. I would like to express my gratitude also for my third supervisor Docent Tarja Kokkola who was always ready to help in anything I needed. My appreciation extends to my co-workers and laboratory technicians, for all the coffee breaks and for their support during this time.

My sincere thanks to my reviewers, Professor Hannu Järveläinen and Docent Jorma Lahtela who gave me constructive comments that improved the quality of my thesis. I extend my gratitude to Professor Risto Kaaja who promptly accepted to be my opponent, despite of the short time available before my defense. I want to express my gratitude also to Antti and Tyyne Soininen Foundation and Sigrid Jusélius Foundation for the financial support of this study.

Special thanks to my friends, Mohamed, Sarang, Moataz, Maria, Dorota, Amro, Vanessa, Massoud, Kristina, Rami, Mesi, Beni, Rand and Elisa, without whom I could not have achieved this goal. All the coffees, the afternoons, the nights, the lunches and the dinners we shared made me understand that together we are stronger. Also to my Brazilian friends, Bruno, Mary and Fernanda with whom I learned that great friendships continues to grow even over long distances. My deepest thanks to my parents who supported me to initiate this journey without hesitation at the time when everything was uncertain. Finally, last but by no means least, thanks to Davi, who has always been my sunshine in the dark days!

Kuopio, February 2020 Lilian Fernandes Silva

(14)

(15)

LIST OF ORIGINAL PUBLICATIONS

This dissertation is based on the following original publications:

I López Rodríguez M, Fernandes Silva L, Vangipurapu J, Modi S, Kuusisto J, Kaikkonen MU, Laakso M. Functional Variant in the GCKR Gene Affects Lactate Levels Differentially in the Fasting State and During Hyperglycemia.

Sci Rep. 8:15989, 2018

II Fernandes Silva L, Vangipurapu J, Kuulasmaa T, Laakso M. An intronic variant in the GCKR gene is associated with multiple lipids. Sci Rep. 9:10240, 2019

III Fernandes Silva L, Vangipurapu J, Kuulasmaa T, Laakso M. Phenotypic characterization of genetic variants associated with non-alcoholic fatty liver disease (submitted)

The publications were adapted with the permission of the copyright owners.

(16)

(17)

ABBREVIATIONS

ADP Adenosine diphosphate Akt Protein kinase B

ALT Alanine aminotransferase ApoB100 Apolipoprotein B100

AST Aspartate

aminotransferase

ATP Adenosine triphosphate BMI Body mass index

Ca Calcium

cAMP Cyclic adenosine monophosphate

CoA Coenzyme A

DAG Diacylglycerol DNL De novo lipogenesis

FA Fatty acid

FFA Free fatty acid

FOXA2 Forkhead box protein A2 G3P Glycerol-3-phosphate

GCK Glucokinase

GCKR Glucokinase regulatory protein gene

GKRP Glucokinase regulatory protein

GL Glycerolipid

GLP-1 Glucagon-like peptide-1 GLUT2 Glucose transporter 2 GLUT4 Glucose transporter 4 GPC Glycerophosphatidyl- choline

GPE Glycerophosphatidyl- ethanolamine

GPL Glycerophospholipid GWAS Genome-wide association studies

HbA1c Hemoglobin A1c HDLC High-density lipoprotein cholesterol

HPH Human primary

hepatocytes hs-CRP High-sensitivity C- reactive protein HSD17B13 17β-Hydroxysteroid dehydrogenase type 13 IGF Insulin-like growth

factor

(20)

IL-6 Interleukin 6 IR Insulin receptor IRS Insulin receptor substrate

LC3 Microtubule-associated protein 1A/1B light chain 3 LDLC Low-density lipoprotein

cholesterol M Molar

MAG Monoacylglycerol MBOAT7 Membrane-bound O acyltransferase domain- containing protein 7 METSIM Metabolic syndrome in men

mRNA Messenger ribonucleic acid

NAD+ Nicotinamide adenine dinucleotide

NADPH Nicotinamide adenine dinucleotide phosphate

NAFLD Non-alcoholic fatty liver disease

OGTT Oral glucose tolerance test PC Phosphatidylcholine

PE Phosphatidyl-

ethanolamine

PEMT Phosphatidyl- ethanolamine e N- methyltransferase PI Phosphatidylinositol PIP3 Phosphatidylinositol 3,4,5- Trisphosphate

PKC Protein kinase C PNPLA3 Patatin-like

phospholipase domain- containing protein 3 PPARϒ Peroxisome proliferator activated receptor gamma PPP1R3B Protein phosphatase

regulatory subunit 3B PS Phosphatidylserine

PSS1/2 CDP-diacylglycerol serine O-phosphatidyl- transferase ½ PTP1B Protein tyrosine phosphatase-1B

ROS Reactive oxygen species Sam50 Sorting assembly machinery protein SAMM50 Sorting and assembly machinery component 50 SUR Sulfonylurea receptor T2D Type 2 diabetes

TAG Triacylglycerol

(21)

TCA Tricarboxylic acid TM6SF2 Transmembrane 6 superfamily member 2 TNF-α Tumor necrosis factor alpha

VLDL Very-low-density lipoprotein

(22)

(23)

1 INTRODUCTION

Obesity is a complex and multifactorial disease affecting over a third of the world’s population today. The epidemic of obesity presents a major challenge to healthcare around the world. By 2030 an estimated 20% of the world’s adult population will be obese (Hruby and Hu, 2015). Changes in our environment, economic growth, industrialization, sedentary lifestyle, and high calorie diets over the last 30 years are important contributors to this epidemic, resulting in increased morbidity and mortality. Obesity is also a driving force for the metabolic syndrome, type 2 diabetes (T2D), non-alcoholic fatty liver disease (NAFLD), and cardiovascular disease.

NAFLD is strongly associated with insulin resistance and T2D. In morbid obesity, almost all patients have steatosis, and T2D is 5-9 times more frequent in patients with NAFLD than in the general population (Anstee et al., 2013). Thus, obesity, T2D, and NAFLD are tightly linked. T2D is a chronic disease caused by genetic, lifestyle and environmental factors and their interactions. The prevalence and incidence of this disease is constantly increasing worldwide. In 2017, approximately 425 million adults had diabetes, and by 2045 this will rise to 629 million. The proportion of T2D of all cases of diabetes is about 80% and is increasing in most countries.

Genome-wide association studies (GWAS) have made it possible to identify thousands of common genetic variants associated with different diseases, clinical traits, laboratory measurements, and biological pathways. These studies have made it possible to start to understand the genetic architecture of complex diseases.

Combining GWAS with detailed phenotypic and -omics data in large population- based cohorts will help to make new fundamental discoveries in human genetics (Visscher et al., 2017).

The aim of the study is to perform association studies of genetic variants with detailed phenotyping, including metabolomics and other clinically relevant traits, to better understand the metabolic pathways associated with the risk of T2D and NAFLD.

(24)

(25)

2 REVIEW OF THE LITERATURE

2.1 TYPE 2 DIABETES

Type 2 diabetes (T2DM) and its complications have been an enormous contributor to the burden of mortality and disability worldwide, being the ninth major factor responsible for reducing life expectancy (GDB, 2013). Diabetes is predicted to affect 439 million individuals between 20 and 79 years old by 2030 (Shaw et al., 2010).

T2DM is characterized by impaired insulin secretion and insulin resistance in peripheral tissues, especially in skeletal muscle, liver and adipose tissue (Kahn, 1998). Diagnosis of diabetes is established when fasting plasma glucose is ≥7.0 mmol/l in repeated tests or 2-h plasma glucose is ≥11.1 mmol/l in oral glucose tolerance test (OGTT), or HbA1c level is ≥48mmol/mol or random plasma glucose

≥11.1 mmol/l) if symptoms of hyperglycemia are present (ADA, 2010). About 80% of all patients with diabetes have T2D.

Risk factors for the development of T2D are increased calorie intake, overweight, abdominal obesity, sedentary lifestyle, smoking, aging, and genetic susceptibility.

Over 400 genetic loci have been associated with the development of T2D in GWAS (Mahajan et al., 2018). Therefore, identifying the risk variants and metabolic pathways affected in T2D is of great importance for the understanding of the pathophysiology and etiology of this disease.

2.1.1 Pathophysiology

Impaired insulin secretion is the most important defect in T2D. Insulin is a peptide hormone secreted by the pancreatic islets of Langerhans, specifically by the β-cells.

Insulin stimulates cellular glucose uptake and keeps the blood glucose in the normal range. Insulin is involved in the regulation of carbohydrate, protein and lipid metabolism. The most important stimulus for insulin secretion is glucose, but also amino acids and fatty acids (FAs) increase insulin secretion. In insulin-resistant states insulin’s effects are subnormal in insulin-sensitive tissues (Reaven, 2004).

Figure 1 shows the hyperbolic association between insulin secretion and insulin sensitivity. In insulin resistant states more insulin is needed to keep glucose levels in the normal range.

Hyperglycemia is a result of the loss of β-cell mass (Matveyenko and Butler, 2008) or function (Ferrannini et al., 2005), attributable to genetic or environmental factors.

T2D develops when insulin secretion cannot compensate for insulin resistance, leading to a decreased disposition index, which is calculated as a product of insulin secretion and insulin sensitivity (Basu et al., 2009). First-phase insulin secretion is severely impaired or lost in T2D (van Haeften et al., 2000), whereas the ratio of proinsulin to insulin is increased (Yoshioka et al., 1988). Hyperglycemia becomes

(26)

more severe and the treatment of diabetes more challenging in the course of the disease due to the continuous deterioration of β-cell function.

Figure 1. Graph showing relationship between insulin secretion and insulin sensitivity (Ader and Bergman, 1987).

2.1.2 Glucose metabolism

The plasma glucose level is regulated by the absorption of nutrients from the intestine in the postprandial state, and by glycogenolysis and gluconeogenesis during the fasting state. After feeding, glycolysis takes place in the cytoplasm, generating. Pyruvate can be converted to acetyl-CoA, which enters the Kreb's cycle in aerobic conditions, or it can be converted to lactate under anaerobic conditions (Cori cycle) (Phypers and Pierce, 2006). Conversion of pyruvate to lactate generates NAD+, which can be used for additional glycolysis. The liver takes up 70% of the lactate generated from oxidation of pyruvate in gluconeogenesis. Glucose is then released into the bloodstream and taken up by the muscles and other insulin- sensitive tissues (Phypers and Pierce, 2006). After 12, 20, and 40 hours of fasting, gluconeogenesis accounts for 41, 71, and 92% of glucose production, while Cori cycle accounts for 18, 35, and 36%, respectively (Katz and Tayek, 1998). Glycogenolysis is the main mechanism responsible for providing glucose availability during the first 8-12 hours of fasting. Gluconeogenesis takes place after longer periods of fasting, releasing glucose from the liver into the bloodstream (Rui, 2014). Kidney is also a gluconeogenic organ and has all the gluconeogenic enzymes. It is responsible for about 20% of overall glucose endogenous release, whereas liver accounts for about 45% (Gerich et al., 2001), the intestines are responsible for 20-25% (Mithieux and

(27)

Gautier-Stein, 2014), and the remaining amount comes from muscles and in small proportions from the brain.

The key hormones regulating glucose homeostasis are insulin and glucagon.

When glucose concentration in the blood is >3.3 mmol/l, β-cells secrete insulin into the bloodstream (Gerich, 1993). Insulin regulates postprandial glucose levels by sending signals to insulin-sensitive peripheral tissues, mainly skeletal muscle, resulting in increased glucose uptake (Gerich et al., 1974). In the liver insulin promotes glycogenesis (the synthesis of glycogen) and inhibits glycogenolysis (the breakdown of glycogen) and gluconeogenesis (the synthesis of glucose). In pancreatic alpha cells insulin inhibits glucagon secretion (Wilson and Foster, 1992).

Insulin is involved in several metabolic processes, such as triacylglyceride (TAG) storage in fat tissue, protein synthesis and mitogenic processes (Wilson and Foster, 1992). Increased plasma concentrations of some amino acids, such as arginine, leucine and lysine, also stimulate insulin secretion. In the fed state, glucose levels reach their peak, and then decrease during the following hours until glucose reaches its fasting level. The liver generates glucose when the level of glucose decreases in the plasma (Figure 2). The kidneys also contribute to glucose generation (Gerich, 2010).

(28)

Figure 2. Glucose regulation by insulin and glucagon hormones (Hædersdal et al., 2018) β-cells also secrete amylin into the bloodstream in response to feeding (Koda et al., 1992). Amylin supresses glucagon secretion in the fed state (Gedulin et al., 1997) via efferent vagal signals and also by slowing the gastric emptying, and absorption of nutrients (Samson et al., 2000). Amylin cannot prevent glucagon secretion in hypoglycaemia induced by insulin (Heise et al., 2004).

Glucagon is a hormone secreted by pancreatic α-cells. It contains 29 amino acids and is involved in catabolic reactions. Glucagon stimulates hepatic glycogenolysis and gluconeogenesis to keep normal glucose levels in the fasting state. Glucagon also stimulates gluconeogenesis in the kidneys and increases the glomerular filtration rate and natriuresis (Bankir et al., 2016). Glucagon secretion is suppressed in the fed state, resulting in a decrease in gluconeogenesis. Glucagon is not suppressed adequately in T2D, resulting in increased hepatic glucose production (Orci et al., 1975).

(29)

Gut peptides also play an important role in glucose homeostasis, inducing incretin hormones and releasing insulin upon feeding (Kreymann et al., 1987). The L cells located in the distal intestine release glucagon-like peptide-1 (GLP-1), and the K cells from the upper small intestine secrete gastric inhibitory peptide (GIP) in response to feeding, which stimulates the parasympathetic system (Drucker, 2001).

Gut peptides also regulate gastric emptying and gut motility (Drucker, 2003).

2.1.3 Insulin secretion

Insulin gene is located on chromosome 11. Insulin synthesis happens in the endoplasmic reticulum (ER) to produce proinsulin (Schroder and Zuhlke, 1982).

Proinsulin is then transferred to the Golgi apparatus, where the zinc-containing proinsulin hexamers are generated (Dodson and Steiner, 1998). Once released from the Golgi apparatus, proinsulin is converted to insulin and C-peptide, which are secreted into the blood by exocytosis (Malaisse, 1997).

Insulin is released in response to the metabolic demand in healthy individuals.

An increase in plasma glucose concentration leads to a release of insulin by the β- cells (Schmitz et al., 2008). β-cells are grouped in islets connected to the vascular system by fenestrated capillaries, which enhances permeability between the circulation and the islets, allowing fast insulin release into the blood (Suckale and Solimena, 2008). β-cells express glucose transporter 2 (GLUT2), which balances glucose levels in β-cells via facilitated diffusion.

Glucose enters β-cells via GLUT2 and is phosphorylated by GCK, which is the rate-limiting enzyme in the β-cells (Suckale and Solimena, 2008), generating glucose- 6-phosphate and further pyruvate (Figure 3). Pyruvate is the end product of glycolysis and is used to produce ATP in β-cells via anaplerosis and acetyl-CoA.

Pyruvate oxidation is the most important pathway involved in the release of insulin- containing granules. Pyruvate oxidation increases the ATP:ADP ratio in the cytoplasm, closing the ATP-sensitive potassium channels (KATP), depolarizing the plasma membrane and opening voltage-gated Ca2+ channels. The influx of Ca2+

triggers the exocytosis of insulin from the granules (Ashcroft et al., 1984).

Anaplerosis generates NADPH, malonyl-CoA, and glutamate, which can amplify the release of insulin-containing granules (Chang and Goldberg, 1978) (Figure 3).

(30)

Figure 3. Glucose-stimulated insulin secretion (Jameson et al., 2018). Abbreviations: Ca, calcium; cAMP, cyclic adenosine monophosphate; K, potassium.

The combination of specific amino acids in plasma can generate metabolites that replenish the TCA cycle and consequently increase insulin secretion from β-cells (Dixon et al., 2003). β-cell insulin secretion can be indirectly activated by free amino acids released into the bloodstream during fasting. Free amino acids can also stimulate glucagon secretion, increase glucose levels and stimulate insulin secretion (Dixon et al., 2003). In addition dietary amino acids present in the gut can stimulate the incretin hormones GLP-1 and GIP, which in turn stimulate insulin secretion (MacDonald et al., 2002).

Glycerol-3-phosphate (G3P) and FFAs are other signalling molecules for glucose via the generation of long-chain acyl-CoA and diacylglycerol (DAG), which increase insulin secretion and provide NAD+ to mitochondria, triggering mitochondrial energy metabolism (Bender et al., 2006). Long-chain acyl-CoA acylates synaptosomal-associated protein-25 (SNAP-25) and synaptogamin, which are insulin granule fusion proteins (Chapman et al., 1996; Gonzalo and Linder, 1998).

DAG activates protein kinase C (PKC), and it can also bind to protein Munc-13, triggering insulin secretion (Rhee et al., 2002).

(31)

Some other hormones can also influence insulin secretion, including 17β- estradiol (Brussaard et al., 1997), melatonin (Kemp et al., 2002), leptin (Ahren et al., 1999), and growth hormone (Sonksen, 2001).

2.1.4 Insulin resistance

Insulin regulates metabolic processes in the fed state. In situations of an excess of exogenous energy, insulin release is stimulated, which suppresses lipolysis and stimulates adipose tissue synthesis. In myocytes, glucose entrance promotes synthesis and storage of glycogen, allowing carbohydrates to be used as the first energy source for muscle contraction (Karam, 1997).

The insulin receptor (IR) contains 2 α-subunits and 2 β-subunits. For insulin to exert its actions it binds to the extracellular α subunits, inducing the transphosphorylation of one β subunit by another on specific tyrosine residues, increasing the kinase catalytic activity (Saltiel and Pessin, 2003). This process activates the IR, which phosphorylates tyrosine residues on insulin receptor substrates (IRS) (Watson et al., 2004). IRS proteins interact with phosphatidylinositol (PI) 3-kinase, activating the enzyme, which generates PI3,4,5-trisphosphate (PIP3) (Shepherd, 2005). PIP3 activates phosphoinositide-dependent kinase 1 (PDK1), which phosphorylates and activates protein kinase B (Akt), mediating the translocation of glucose transporter 4 (GLUT4) to the plasma membrane (Mora et al., 2004). The manifestation of insulin resistance at the cellular level involves defects in insulin signaling, including disruption in the pathways involving the tyrosine phosphorylation of the insulin receptor, insulin receptor substrate protein or PI3- kinases, or disturbances in GLUT4 function (Wheatcroft et al., 2003). Among these mechanisms, protein tyrosine phosphatase 1 B (PTP1B) seems to play an important role in insulin resistance by targeting tyrosine-phosphorylated IR and IRS-1, hence negatively attenuating their signal (Kenner et al., 1996). Important roles of PTP1B in obesity and diabetes were confirmed by a deletion of the PTP1B gene in mice resulting in a decrease in insulin resistance, most likely due to an increase in tyrosine phosphorylation of IR (Elchebly et al., 1999).

The effects of insulin resistance are different depending on the function of the tissues and organs (Figure 4). Insulin-dependent tissues include adipose tissue, liver and muscle. Muscles can take up around 60-70% of the available glucose in the body via GLUT4 (Reaven, 2004). In the postprandial state insulin activates glycogen synthase, producing glycogen storage that can be further used as an energy source via glycogenolysis. In the basal state myocytes are not dependent on glucose or glycogen as energy sources. Insulin is an anabolic hormone, exerting its anabolic effects through growth hormone and insulin-like growth factor 1 (IGF-1) (Giorgino et al., 2005). Insulin resistance results in decreased glycogen synthesis in muscles (Karam, 1997).

Adipose tissue is responsible for 10% of total body glucose uptake via GLUT4

(32)

lipolysis, and releases FFAs into the bloodstream, which are taken up by the liver.

Augmented influx of FFAs into the liver result in the production of very-low-density lipoprotein (VLDL) particles (Grundy, 2004), whereas the uptake of TAG is decreased, likely attributable to the impairment in the lipoprotein lipase activity.

Adipose tissue secretes several cytokines that increase insulin resistance, including angiotensinogen, interleukin 6, TNF-α, plasminogen activator inhibitor 1, and leptin (Devaraj et al., 2004). Adipocytes from patients with T2D have reduced IRS-1 gene and protein expression, diminished GLUT4 translocation, and defects in PIP-3 kinase and Akt pathways (Smith, 2002).

Glucose uptake into the liver is independent of insulin. The liver counts for about 30% of total body insulin-mediated glucose disposal. During starvation gluconeogenesis is increased. Insulin stimulates glycogen synthesis and suppresses gluconeogenesis and ketone body generation (Smith, 2002; Rui, 2014). Insulin also exerts mitogenic effects mediated via the generation of IGFs by the liver (Hunter et al., 1998) and supresses Sex Hormone Binding Globulin (SHBG) generation (Arcidiacono et al., 2012). Other effects of insulin resistance in the liver are changes in lipoprotein metabolism, such as accumulation of TAG in the liver and increased VLDL secretion (Krauss, 2004).

The brain is not an insulin-dependent organ. Insulin in the brain may act as a neuropeptide, regulating olfaction, memory, satiety, appetite and cognition (Gerozissis, 2004). Insulin has effects on other appetite-regulating neurotransmitters and peptides, including leptin, which share the same pathway as insulin in the hypothalamus (Devaraj et al., 2004). The kidneys have sufficent gluconeogenic capacity and glucose-6-phosphatase activity, and insulin is not needed for glucose transport in the organ. The kidneys also control mineral transport (Nagasaka and Kaneko, 1992). Glomerular hyperfiltration has been associated with insulin resistance and microalbuminuria, and it also predicts incident diabetes (Bell et al., 2006, Mykkänen et al., 1998). In the bone, insulin has an anabolic function by stimulating bone formation in osteoblasts (Thomas et al., 1998).

(33)

Figure 4. Effects of insulin resistance in different tissues and organs (modified from Artunc et al, 2016). Abbreviations: FFA, free fatty acid; VLDL, very-low-density lipoprotein.

Insulin also exerts its effects in intestinal cells. In the fed state intestinal epithelial L cells secrete GLP-1, a 30-amino acid peptide, which enhances glucose-stimulated secretion of insulin from the β-cells (Lim and Brubaker, 2006). Decreased levels of GLP-1 have been reported in T2D (Toft-Nielsen et al., 2001), but the clearance rate of GLP-1 does not differ between controls and individuals with diabetes, indicating a possible secretory defect in the L cells (Vilsboll et al., 2003).

Insulin is important for endothelial function, especially for nitric oxide generation. Insulin stimulates endothelial NOS (eNOS) via PIP-3 kinase and Akt, resulting in an increase in nitric oxide, which mediates endothelial relaxation in large arteries, and prevents cell adhesion, platelet aggregation, and smooth muscle cell proliferation (Chen et al., 2008). Insulin also stimulates the release of the vasoconstrictor endothelin, which is increased in insulin-resistant states (Chen et al., 2008). Pathways stimulating eNOS are impaired in insulin-resistant states, and the responses to vasodilator stimuli and to cholinergic agonists are disrupted (Muniyappa et al., 2013).

2.1.5 Genetic factors

The risk of developing T2D depends on genetic and environmental factors.

Individuals having one parent with T2D have a 40% increased risk, and individuals having both parents affected have a 70% increased risk of developing T2D

(34)

populations, and this variability is explained by environmental, lifestyle and genetic factors. Previous studies have identified genetic variants increasing the risk of developing T2D by applying different approaches, linkage studies, candidate gene studies and recently GWAS studies.

Linkage studies aim to investigate the genetic variants that are inherited together.

This method is a good tool to detect rare variants of large effect, but it is not effective in detecting genes involved in complex polygenic disorders, including T2D. Linkage studies have discovered only one gene consistently linked with T2D, the transcription factor 7-like 2 (TCF7L2) (Table 1). The candidate gene study focuses on genes involved in known mechanisms that possibly influence T2D development, such as glucose metabolism, insulin secretion, insulin receptors, post-receptor signaling, and lipid metabolism. Only a few genes have shown significant associations with T2D in candidate gene studies (Table 1).

Recently, a GWAS approach has been applied. This approach includes a genome- wide set of genetic variants to investigate whether they are associated with a specific trait in large population samples. GWAS studies have identified more than 400 loci increasing the risk of developing T2D (Mahajan et al., 2018). The first genetic variants have been found in TCF7L2, HHEX, SLC30A8, CDKN2A/B, IGF2BP2 and GCKR genes (Table 1).

Table 1. Genes associated with T2D in linkage studies, candidate gene studies and GWAS studies.

TYPE OF STUDY GENE ASSOCIATION WITH T2D

LINKAGE STUDIES

TCF7L2

(Transcription factor 7- like 2)

Encodes a transcription factor that is active in beta cells

This finding was confirmed later in the Mexican-American, Icelandic, American and Danish cohorts, making TCF7L2 the most consistent and strongly associated T2D gene (Tong et al., 2009). Among several genes TCF7L2 presented the strongest signal for T2D in GWAS study, being replicated in several human population studies (Peng et al., 2013). This gene seems to modify the action of incretins, impairing insulin secretion (Schäfer et al., 2007).

PPARG

(Peroxisome proliferator- activated receptor gamma)

Encodes the molecular target of

thiazolidenediones

The risk of developing T2D is increased in 20%

in presence of the Pro12Ala substitution in the PPARG gene (Deeb, et al. 1998). This association has been confirmed in several meta-analyses.

(35)

CANDIDATE GENE STUDIES

IRS1 and IRS-2 (Insulin receptor substrate 1/2) Encode peptides involved in insulin signal transduction

Insulin receptor substrate IRS-1 and IRS-2 genes. Polymorphisms in these genes are associated with decreased insulin sensitivity (Clausen et al., 1995).

KCNJ11

(Potassium Voltage- Gated Channel Subfamily J Member 11 Encodes the Kir6ATP- sensitive potassium channel)

It plays a role in the regulation of insulin secretion by β-cells. Mutations in this gene leads to neonatal diabetes. A polymorphism in KCNJ11 is associated with T2D (Hani et al., 1998) and also with decreased insulin secretion (Gloyn et al., 2003).

WFS-1 (Wolframin ER Transmembrane Glycoprotein) Encodes Wolframin

The WFS1 gene seems to regulate β-cell function. Two polymorphisms in WFS-1 were associated with T2D in different populations (Sandhu et al., 2007).

HNF1A, HNF1B and HNF4A

(HNF homeobox 1A/1B/4A)

These genes play a role in the development and functioning of the liver and β-cells.

Polymorphisms in these genes are associated with impaired insulin secretion and an increased risk of developing T2D (Furuta et al., 2002).

GWAS STUDIES

IGF2BP2

(Insulin-like growth factor 2 mRNA-binding protein 2)

Encodes a protein that regulates IGF2 translation [69,70].

The mechanisms leading to increased risk of T2D development are not yet known, but it may affect β-cell function (Grarup et al., 2007).

SLC30A8

(Solute Carrier Family 30 Member 8)

Encodes for a protein that regulates insulin granules

It is expressed only in the islets of Langerhans (Lefebvre et al., 2012). This association has been replicated in several populations (Strawbridge et al., 2011).

GCKR

(Glucokinase regulatory protein)

It inactivates GCK during fasting state.

Polymorphisms in GCKR gene have been associated with the development of T2D (Ling et al., 2011) and NAFLD (Speliotes et al.,

(36)

Encodes for glucokinse regulatory protein

HHEX

(Hematopoietically- expressed homeobox protein)

Encodes a transcription factor involved in Wnt signaling

The association was identified in Caucasian and Asian populations (Li et al., 2012a). The mechanisms how the risk of T2D is increased remains unknown.

CDKN2A/B

(cyclin-dependent kinase Inhibitor 2A/2B)

SNPs were associated with the risk of T2D in multiple GWAS studies, conferring 20-50%

increased risk for development of T2D (Li et al., 2012b).

Multiple other genes Currently more than 400 loci increasing risk of developing T2D were identified in GWAS studies (Udler et al., 2019).

The GCKR gene, which is in the focus in our studies, encodes for glucokinase regulatory protein (GKRP), which inhibits GCK, an enzyme that converts glucose to glucose-6-phosphate in the cytoplasm of the cells. GCKR is mainly expressed in the liver and pancreas. Differently from other hexokinases, GCK is not inhibited by physiological levels of glucose-6-phosphate (Agius, 2008), acting as a sensor for insulin secretion in the pancreas and facilitating glycogen storage in the liver.

During fasting GKRP forms an inactive complex with GCK in the nucleous, inactivating the enzyme (van Schaftingen et al., 1992), triggering glycogenolysis and gluconeogenesis. In the fed state GCK translocates to the cytoplasm, triggering glycolysis, glycogen storage and de novo lipogenesis (DNL).

The intronic GCKR variant rs780094-C and the coding region variant Pro446Leu, have been shown to be risk variants for several diseases and conditions in GWAS studies, including T2D, NAFLD, familial combined hyperlipidemia, gout, and chronic kidney diseases (Dupuis et al., 2010; Speliotes et al., 2011; Wang et al., 2012;

Bonetti et al., 2011). GCKR variants have also been associated with a wide range of metabolites, such as carbohydrates (Suhre et al., 2011), lipids (Orho-Melander et al., 2008), amino acids (Stančáková et al., 2012), fasting insulin (Ingelsson et al., 2010), C-peptide (Perez-Martinez et al., 2011), homeostasis model assessment of insulin resistance (Perez-Martinez et al., 2011), liver enzyme γ-glutamyl transferase (Chambers et al., 2011), and urate (Kolz et al., 2009). Whereas the GCKR rs780094-C is associated with increased levels of glucose and ketone bodies (Mahendran et al., 2013), the rs780094-T has been associated with lower levels of plasma glucose,

(37)

increased TAG levels (Orho-Melander et al., 2008), lower levels of high-density lipoprotein cholesterol (HDLC) (Dupuis et al., 2010), and higher C-reactive protein (CRP) levels (Orho-Melander et al., 2008). These associations are likely due to the Pro446Leu substitution in GCKR, which results in destabilization of the GCK binding interface, explaining inverse correlation between fasting glucose and

triglycerides for this variant (Zelent et al., 2014).

GCKR rs780094-T has been previously associated with increased fasting lactate levels (Tin et al., 2016). Glucose and alanine are the main sources for lactate generation in humans. In the fasting state, the liver takes up lactate released by adipose tissue and skeletal muscle for gluconeogenesis. In the fed state glucose is metabolized to lactate (Ferguson et al., 2018). Thus, plasma lactate levels indicate the rate of hepatic glucose metabolism.

2.2 NON-ALCOHOLIC FATTY LIVER DISEASE

The epidemic of obesity and diabetes have substantially contributed to the increase of NAFLD worldwide. NAFLD is characterized by lipid deposition (>5% of the liver weight) in the liver not related to alcohol consumption. A recent clustering study has identified NAFLD as a diabetes subtype in four distinct diabetes cohorts, confirming the significance of disrupted liver lipid metabolism in patients with T2D (Udler et al., 2018). NAFLD is a progressive disease, ranging from simple steatosis to nonalcoholic steatohepatitis (NASH), cirrhosis and hepatocellular carcinoma (Hazlehurst et al., 2016).

Serum biomarkers (alanine aminotransferase (ALT), aspartate aminotransferase (AST) and imaging by ultrasound or MRI are used in the diagnosis of NAFLD. The NAFLD fibrosis score is a validated, noninvasive diagnostic tool for identifying patients having liver fibrosis (Angulo et al., 2007). It is calculated based on selected laboratory values (glucose, platelet count, albumin, the AST/ALT ratio) and patient characteristics (age, BMI, and diabetes status). Fibrosis scores range from 0 to 4, 0 indicating no signs of fibrosis and 4 showing the presence of cirrhosis. However, the liver biopsy is still considered the gold standard method for final diagnosis for NAFLD (NICE, 2016).

NAFLD increases the risks of liver and cardiovascular morbidity and mortality, and this disease is predicted to become the leading cause for liver transplantation in United States by 2030 (Pais et al., 2016). NAFDL is a multifactorial disease and results in impairment of lipid metabolism, mitochondrial derangement, hormonal disturbance and tissue repairing.

(38)

2.2.1 Patophysiology

The pathophysiology of NAFLD includes nutritional, hormonal and genetic factors.

The liver receives signals from the adipose tissue, pancreas, brain and gut to maintain metabolic homeostasis (Maher, 2016). Several factors, such as hormonal levels, nutritional state, intestinal dysbiosis and lipotoxicity can disturb liver homeostasis, thus affecting the development and progression of this disease (Maher, 2016). Insulin resistance and hyperinsulinemia together with an excess of calories intake and obesity are currently considered as the key mechanisms in the development of NAFLD (Marchesini et al., 2001). Decreased insulin sensitivity in the whole body, adipose tissue and liver is a hallmark of NAFLD. An increased influx of FFAs into the liver together with increased glucose levels trigger anabolic processes, leading to fat accumulation in the liver.

Hormonal changes related to the reward center, decreased gut-derived hormones promoting satiety, such as GLP-1 (Page et al., 2013), and increases in gut hormones stimulating appetite, such as ghrelin (Teff et al., 2004), are relevant factors in the development of NAFLD, because they increase the amount of nutrient intake, resulting in increased TAG levels in blood circulation (Teff et al., 2004). The interplay between adipose-derived hormones leptin and adiponectin are likely involved in the development of steatosis. Adiponectin increases hepatic insulin sensitivity and decreases body fat (Buechler, et al., 2011), whereas leptin reduces food intake and increases caloric expenditure (Ahima et al., 2000). Individuals with NAFLD show low levels of adiponectin (Targher et al., 2006) and high levels of leptin, indicating leptin resistance (Chitturi et al., 2002).

The imbalance of gut microbiome can disrupt the functioning of the immunologic barrier in the gut. Bacteria dysbiosis has been associated with inflammation, impaired energy homeostasis, and disturbance in choline and bile acid metabolism (Boursier et al., 2016), factors involved in the development and progression of NAFDL. Dysbiosis leads to the activation of Kupffer cells and hepatic stellate cells (Compare et al., 2012), which activate Toll-like receptors (TLRs) and tumor necrosis factor-alpha (TNF-α) receptors, causing inflammation and liver injury (Boursier et al., 2016; He et al., 2016).

Nutritional factors affect gut microbiota by influencing the development of or protection from NAFLD. High-fat, high-protein, and high-fructose diets and diets that are low in omega–3 FAs may lead to dysbiosis in the gut microbiota and increase the risk for the development of NAFLD (Mokhtari et al., 2017). Gut microbiota regulates changes in substrate supply, predisposing to the development of NAFLD by increasing the conversion of choline to methylamine, decreasing VLDL synthesis, leading to fat accumulation in the liver, and by lowering FXR signaling resulting in an increase in liver DNL (Jasirwan et al., 2019). Conversely, a diet rich in prebiotics, probiotics, and polyphenols can protect from NAFLD (Mokhtari et al., 2017).

Lipotoxicity has a pivotal role in the development of NAFLD, promoting lipoapoptosis (Kusminski et al., 2009). Excess FFA delivery to the liver induces ER stress, activating caspase, signaling pathways such as C/EBP-homologous protein

(39)

(CHOP) and c-Jun-N-terminal kinase (JNK), and causing mitochondrial dysfunction, insulin resistance, oxidative stress, and lipoapoptosis (Ibrahim et al., 2011). The severity of NAFLD correlates with the degree of hepatocyte lipoapoptosis (Feldstein et al., 2003) and is reflected by an increase in serum caspase-cleaved cytokeratin–fragments (Feldstein et al., 2009). Lipoapoptosis correlates with a high ratio of AST/ALT and hepatic fibrosis (Feldstein et al., 2003). NAFLD is a multifactorial disease that entails a complex interaction of genetic factors, hormones, diet, and lifestyle, all of which contribute to the NAFLD phenotype.

2.2.2 Genetic factors

Non-acoholic fatty liver disease (NAFLD) shows a heritability ranging from 20% to 70% in different studies (Sookoian and Pirola, 2017). Although several genetic variants involved in NAFLD have been identified in GWAS studies, including PNPLA3 (Speliotes et al., 2011), TM6SF2, MBOAT7 (Buch et al., 2015), GCKR (Yang et al., 2011), and SAMM50 (Kitamoto et al., 2013) (Table 2), the NAFLD European guidelines recommends to genotype selected patients only for the two major genetic variants of NAFLD (PNPLA3 and TM6SF2) (EASL, 2016). Genetic variants preventing NAFLD in the PPP1R3B (Dongiovanni et al., 2018) and HSD17B13 (Abul- Husn et al., 2018) genes have also been identified (Table 3).

Table 2. Genes associated with an increase in the risk of NAFLD.

GENE ASSOCIATION WITH NAFLD

PNPLA3 (Patatin-like

phospholipase domain- containing 3)

PNPLA3 is the strongest and most replicated genetic variant predisposing to NAFLD (Singal et al., 2014). The PNPLA3 rs738409-G encodes for a nonsynonymous change, isoleucine to methionine substitution at position 148 (I148M) of the protein (Romeo et al., 2008), likely inactivating the protein function. Thus, accumulation of inactive PNPLA3 on the surface of lipid droplets leads to fatty accumulation in the liver. This variant has been reported to increase ALT levels (Larrieta- Carrasco et al., 2014), insulin resistance (Wang et al., 2011) and retinol levels (Mondul et al., 2015).

TM6SF2

(Transmembrane 6 superfamily 2)

TM6SF2 encodesan ER membrane protein. TM6SF2 rs58542926-T leads to misfolded protein and increasesTM6SF2 degradation (Kozlitina et al., 2014), resulting in reduced secretion of hepatic VLDL, TAG, and apolipoprotein B (ApoB), and increased accumulation of hepatic TAG (Anstee and Day, 2015). An Argentinean study showed a close association between the TM6SF2 variant and the severity of hepatic steatosis (diagnosed by liver biopsy) (Sookoian et al., 2015).

(40)

MBOAT7

(Membrane-bound O- acyltransferase domain- containing protein 7)

MBOAT7 rs641738-T variant is associated with increased hepatic fat content and increased risk of fibrosis in two cohorts (Mancina et al., 2016). MBOAT7 rs641738-T is associated with reduced protein expression in the liver and disturbances in plasma PI species. MBOAT7 catalyses the transfer of an acyl-CoA to lysophosphatidylinositol species and has a preference for arachidonoyl-CoA. Thus, MBOAT7 likely regulates free arachidonic acid and eicosanoid levels, which are inflammatory mediators (Gijón et al., 2008).

GCKR

(Glucokinase regulatory protein)

Genetic variants in GCKR have been associated with NAFLD in different studies (Speliotes et al., 2011; Di Costanzo et al., 2018). The T allele of rs780094 was associated with increased TAG levels, lower fasting glucose, insulin levels and insulin resistance in different association studies (Teslovich et al., 2010; Sparso et al., 2008; Willer et al., 2008).

A GCKR variant was also associated with carbohydrates (Suhre et al., 2011), lipids (Orho-Melander et al., 2008), amino acids (Stančáková et al., 2012), fasting insulin (Ingelsson et al., 2010), C peptide (Perez- Martinez et al., 2011), homeostatic model assessment insulin resistance (HOMA-IR) (Perez-Martinez et al., 2011), the liver enzyme γ-glutamyl transferase (Chambers et al., 2011), urate (Kolz et al., 2009), ketone bodies (Mahendran et al., 2013), CRP levels (Orho-Melander et al., 2008), andlactate levels (Tin et al., 2016).

SAMM50

(Sorting and assembly machinery component 50)

SAMM50 encodes the SAM50 protein and has a role in integrating beta- barrel proteins into the outer mitochondrial membrane (Humphries et al., 2005). SAMM50 is involved in keeping mitochondrial shape, morphology of mitochondrial cristae and assembly of respiratory complexes (Ott et al., 2012; Ott et al., 2015). A decrease in Sam50 leads to mitochondrial dysfunction, such as changes in mitochondrial shape, morphology of the cristae, and assembly of respiratory complexes (Ott et al., 2012), resulting in an increase of insulin resistance (Lowell and Shuman, 2005; Ma et al., 2012). SAMM50 was strongly associated with the presence and severity of NAFLD in a Korean population (Chung et al., 2018). In a Mexican population SAMM50 was associated with higher ALT levels and a higher risk of NAFLD (Larrieta-Carrasco et al., 2018).

SOD2

(Manganese-dependent superoxide dismutase)

SOD2 encodes manganese-dependent superoxide dismutase (MnSOD), a mitochondrial protein that binds to the superoxide byproducts of oxidative phosphorylation and converts them to hydrogen peroxide and diatomic oxygen, protecting cells from oxidative stress (MacMillan-Crow & Cruthirds, 2001). The MnSOD/SOD2 rs4880-T variant has been shown to play an important role in the pathogenesis of advanced NAFLD (Namikawa et al., 2004; Al-Serri et al., 2011).

(41)

PEMT

(Phosphatidylethanola- mine N-methyltransferase)

PEMT encodes phosphatidylethanolamine N-methyltransferase, enzyme that converts phosphatidylethanolamine (PE) to phosphatidylcholine (PC) by sequential methylation in the liver and plays a role in the

ssembling and secretion of VLDL particles in the liver (Bremer and Greenberg, 1961). Another synthetic pathway in nucleated cells converts intracellular choline to PC by a three-step process. PEMT rs7946-T has been implicated in NAFLD, due to the impairment in the VLDL assembling process (Dong et al., 2007; Song et al., 2005).

LEPR

(Leptin receptor gene)

LEPR encodes a protein that is a receptor for leptin and is involved in the regulation of fat metabolism. LEPR rs1137101-G, a variant in leptin receptor gene, has also been associated with the risk of NAFLD (Zain et al., 2013; An et al., 2016).

Table 3. Genes associated with a decrease in the risk of NAFLD.

GENE ASSOCIATION WITH NAFLD

PPP1R3B

(Protein Phosphatase 1 Regulatory Subunit 3B)

PPP1R3B regulates glycogen metabolism in the liver. PPP1R3B dephosphorylates and activates glycogen synthase, and inactivates glycogen phosphorylase (Agius, 2015). An increase in glycogen synthesis and a decrease in glycogen breakdown result in glycogen accumulation in the liver (Weber and Cantero, 1958). PPP1R3B was associated with decreased cholesterol levels in three cohorts. In obese individuals, it was associated with increases in PPP1R3B mRNA expression and lipid oxidation and a decrease in lipid metabolism pathways, inflammation, and the cell cycle, thus lowering the risk of NAFLD (Dongiovanni et al., 2018). The PPP1R3B variant has previously been associated with glycine (Teslovich et al., 2018).

HDS17B13 (17β-Hydroxysteroid dehydrogenase type 13)

HSD17B13 rs72613567:TA is a splice variant that encodes hepatic lipid droplet protein hydroxysteroid 17-beta desydrogenase 13, which catalyzes the conversion between 17-keto- and 17-hydroxysteroids (Wen et al., 2019). The HSD17B13 variant generates a truncated protein with a reduced enzymatic activity (Abul et al., 2018). HSD17B13 variant has been associated with decreased PNPLA3 mRNA expression, and decreased ALT levels (Abul et al., 2018). This variant has been associated with decreased levels of ALT and AST and a decreased risk of alcoholic liver disease, NAFLD, alcoholic cirrhosis, and nonalcoholic cirrhosis. Associations were replicated in two independent cohorts (Abul et al., 2018). The rs72613567:TA variant also decreased liver injury associated with a PNPLA3 variant (Abul et al., 2018).

(42)

2.2.3 Lipid metabolism

Lipids belong to a large class of molecules that have several metabolic functions, such as energy generation, signaling molecules and plasma membrane elements.

Lipids are the key components in the adipose tissue, liver and intestine, as well as for energy storage or lipid turnover (Gross and Silver, 2014). Lipids accumulate in skeletal muscles, adrenal cortex and mammary glands (Walther and Farese, 2012).

In poor energy conditions, the lipids stored are used for energy production.

Disrupted lipid metabolism is associated with several diseases, including T2D, coronary artery disease, NAFLD, and cancer (Adams et al., 2006; Pischon et al., 2008).

The lipid sources of the body include the diet, the adipose tissue, and the DNL in the liver. Fats coming from the diet are digested in the small intestine, where TAGs are broken down into monoacylglycerides and FFAs, which are absorbed through the intestinal mucosa, and are further resynthesized as TAGs (Lehninger et al., 2000), and transported into the liver or adipose tissue. During the fasting state lipolysis takes place, breaking down TAGs into glycerol and FAs, which are used as energy sources of the body (Lehninger et al., 2000).

FAs undergo β-oxidation, resulting in acetyl-CoA molecules, which generate ATP in the Krebs cycle (Lehninger et al., 2000). An excess of acetyl-CoA can be used to synthesize ketone bodies. If glucose is not available, ketone bodies can be oxidized to produce energy. Acetyl-CoA generated from excess glucose or carbohydrates can be used to synthesize lipids, TAGs, steroid hormones, cholesterol, and bile salts (Lehninger et al., 2000).

Glycerolipid (GL)/FFA cycling is a process by which FFAs are esterified to G3P to synthetize GL, which can be hydrolyzed to FFAs that can be re-esterified (Prentki and Madiraju, 2008). This cycle is continuous within the cell, and generates neutral GL (monoacylglycerol (MAG), DAG, TAG) and phospholipids. All the metabolites of the GL/FFA cycle are signaling molecules, except for TAG (Prentki and Madiraju, 2008).

DAG can be generated through the phosphorylation of phosphatidic acid (PA), and via the hydrolysis of phospholipids by phospholipase-C (Figure 5). Once DAG is generated it can be acetylated by diacylglycerol acyltransferases (DGAT) to form TAG (reversible reaction), or it can be redirected to generate phospholipids (e.g. PC, PE) (Prentki and Madiraju, 2008) (Figure 5). If DAG is generated from the hydrolysis of TAG, it ends up in the lipid droplet, whereas DAG coming from phospholipid hydrolysis is most likely to be in the plasma membrane, and DAG coming from PA hydrolysis usually ends up in the ER (Prentki and Madiraju, 2008). Therefore, DAG can participate in different signaling pathways, attributable to its source of production (Prentki and Madiraju, 2008).

Phenotypic characterization of genetic variants associated with the risk of type 2 diabetes and non-alcoholic fatty liver disease

uef.fi

Dissertations in Health Sciences

LILIAN FERNANDES SILVA

LILIAN FERNANDES SILVA

PHENOTYPIC CHARACTERIZATION OF GENETIC

VARIANTS ASSOCIATED WITH THE RISK OF TYPE 2

DIABETES AND NON-ALCOHOLIC FATTY LIVER DISEASE

Lilian Fernandes Silva

PHENOTYPIC CHARACTERIZATION OF GENETIC VARIANTS ASSOCIATED WITH THE RISK OF TYPE 2 DIABETES AND NON-ALCOHOLIC FATTY LIVER DISEASE

ABSTRACT

TIIVISTELMÄ

ACKNOWLEDGEMENTS

LIST OF ORIGINAL PUBLICATIONS

CONTENTS

ABBREVIATIONS

1 INTRODUCTION

2 REVIEW OF THE LITERATURE

2.1 TYPE 2 DIABETES

2.2 NON-ALCOHOLIC FATTY LIVER DISEASE