University of Groningen

Essential fatty acid deficiency and the small intestineLukovac, Sabina

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2010

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Lukovac, S. (2010). Essential fatty acid deficiency and the small intestine. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 21-07-2021

Essential Fatty Acid Deficiency

and the Small Intestine

Sabina Lukovac

The research described in this thesis was carried out at the Department of

Pediatrics, Beatrix Children's Hospital, Center for Liver, Digestive and Metabolic

Diseases, University of Groningen, University Medical Center Groningen and was

financially supported by the Dutch Digestive Foundation grant (MW 04-38).

The author gratefully acknowledges the financial support for printing of this thesis by:

Dutch society for gastroenterology (NVGE) and Section Experimental Gastroenterology

©2010 Sabina Lukovac

All rights reserved. No part of this publication may be reproduced or transmitted in

any form or by any means without permission of the author and the publisher holding

the copyrights of the articles.

Cover design: Marc Daalmans en Sabina Lukovac

Layout design: Susanne Kooistra en Sabina Lukovac

Printed by: Ipskamp Drukkers B.V.

Dutch Digestieve Foundation

Graduate school for drug exploration

University of Groningen

Unilever Nederland

University Medical Center Groningen

Essential fatty acid deficiency and the small intestine

Proefschrift

ter verkrijging van het doctoraat in de

Medische Wetenschappen

aan de Rijksuniversiteit Groningen

op gezag van de

Rector Magnificus, dr. F. Zwarts,

in het openbaar te verdedigen op

woensdag 24 maart 2010

om 13.15 uur

door

Sabina Lukovac

geboren op 28 augustus 1980

te Doboj, Bosnië en Herzegovina

Promotores: Prof. dr. H.J. Verkade

Prof. dr. E.H.H.M. Rings

Beoordelingscommissie: Prof. dr. A.F. Bos

Prof. dr. A.K. Groen

Prof. dr. E. Heineman

ISBN

978-90-367-4226-9 (printed)

978-90-367-4227-6 (digital)

This tree has two million and seventy-five thousand leaves. Perhaps I missed a leaf

or two but I do feel triumphant at having persisted in counting by hand branch by

branch and marked down on paper with pencil each total. Adding them up was a

pleasure I could understand; I did something on my own that was not dependent on

others, and to count leaves is not less meaningful than to count the stars, as

astronomers are always doing. They want the facts to be sure they have them all. It

would help them to know whether the world is finite.

by David Ignatow

Paranimfen: Irma Kuipers

Hilde Herrema

CONTENTS

CHAPTER 1 9

Introduction to the thesis

CHAPTER 2 29

Essential fatty acid deficiency in mice impairs lactose digestion

CHAPTER 3 47

Essential fatty acid deficiency in mice alters jejunal cholesterol metabolism

CHAPTER 4 69

Effects of essential fatty acid deficiency on enterohepatic circulation of bile

salts in mice

CHAPTER 5 93

Functional characterization of the in vitro model of EFA deficiency

CHAPTER 6 111

Gelucire®44/14 improves fat absorption in rats with impaired lipolysis

CHAPTER 7 129

Summary and future perspectives

APPENDICES

Nederlandse samenvatting 140

Dankwoord 145

Biography 148

List of publications 149

CHAPTER 1

INTRODUCTION TO THE THESIS

S. Lukovac

Part of this chapter has been published under the title “Nutrition for children with

cholestatic liver disease” in: Nestle Nutr Workshop Ser Pediatr Program 2007; 59:

147-157

CHAPTER 1

10

INTRODUCTION CONTENTS

INTRODUCTION 11

ESSENTIAL FATTY ACIDS (EFA) 11

Essential fatty acid (EFA) deficiency 13

Essential fatty acid (EFA) deficiency in cholestasis 14

Essential fatty acid (EFA) deficiency in cystic fibrosis 15

SMALL INTESTINE 15

Crypts and villi 16

Apical and basolateral compartment in the enterocyte 16

Enterocyte function 17

Fat absorption 17

Peroxisome proliferator-activated transcription 19

factors (PPARs)

Carbohydrate absorption 20

Cholesterol absorption 20

Small intestine and the enterohepatic circulation 21

of bile salts

Manifestation of the impaired small intestinal function 22

in common intestinal disorders

AIM AND THE OUTLINE OF THE THESIS 22

REFERENCES 23

INTRODUCTION

11

1

CH

AP

TE

R

INTRODUCTION

Essential fatty acids (EFA) cannot be synthesized de novo by humans or animals and

thus can only be obtained by means of dietary intake. EFA are involved in many

biological processes; they are essential for normal neurodevelopment and regulation of

membrane function in several tissues like the brain, retina, liver, kidney, adrenal glands

and gonads.1 In addition, metabolites of EFA are precursors of eicosanoids, which

strongly modulate processes like platelet aggregation and chemotaxis of the immune

system.2 Accordingly, a shortage of EFA, also known as EFA deficiency, leads to various

clinical consequences, such as impaired cognitive and motor development, reduced

growth rate, dry skin, hair loss and functional changes in organs like hearth and liver.2

EFA deficiency is a condition that can develop due to insufficient dietary intake or

absorption, or due to enhanced metabolism of EFA. This condition is described in detail

in one of the paragraphs of this chapter. Pediatric patients with cholestatic liver disease

often encounter EFA deficiency, which is one of the determining factors for failure to

thrive in these patients. In order to improve the nutritional status of patients with

Cholestasis-induced failure to thrive (CIFTT), maintenance of intestinal absorptive

capacity is essential.

Previous studies on EFA deficiency mainly focused on its effects on the liver, brain and

heart.1 However, little is known about the effects of EFA deficiency on the function and

physiology of the small intestine. In order to improve the nutritional status of pediatric

patients, knowledge of (the effects of EFA deficiency on) the small intestinal function is

essential. Recent studies suggested that EFA deficiency by itself might deteriorate the

intestinal function, as demonstrated by EFA deficiency induced fat malabsorption.3

Rather than intraluminal effects, intracellular defects in the small intestinal enterocytes

were suggested to contribute to fat malabsorption during EFA deficiency in mice.4

The aim of this thesis was to characterize the effects of EFA deficiency on the function,

morphology and (patho)physiology of the small intestine in a murine model. To study the

intracellular effects of EFA deficiency in more detail, an in vitro model was established.

Insight into the pathophysiology of EFA deficiency, regarding the small intestinal

function, might help improve the nutritional status of patients with CIFFT and other

conditions associated with EFA deficiency.

ESSENTIALS FATTY ACIDS (EFA)

The two EFA, also known as “parental” EFA, are linoleic acid (C18:2ω-6, LA) and α-

linolenic acid (C18:3ω-3, ALA). By means of a cascade of desaturation and elongation of

the carbon chain, LA and ALA can be converted into their long chain fatty acid

metabolites (LCPUFA: long chain polyunsaturated fatty acids) of the ω-6 and the ω-3

families, respectively (Figure 1).5,6,7

Enzymes responsible for desaturation steps are being competed for by different

LCPUFA. The enzymes have preferred affinity for the ω-3 family of LCPUFA over the ω-

6 family members. The affinity for these two EFA families, on the other hand, is preferred

over the affinity for non-EFA of ω-9 and ω-7 fatty acids. Desaturation and elongation of

fatty acids depend on the needs and availability of the LCPUFA in the organism.8 LA and

CHAPTER 1

12

ALA are not only converted to LCPUFA; part is used as substrates for β-oxidation,

representing a source for energy for the organism.9 Another relevant function of EFA and

their LCPUFA metabolites is their role as constituents of the membrane lipids (mainly of

phospholipids). Within the membrane, they regulate its fluidity, but also the function and

localization of the proteins within these membranes. EFA and their LCPUFA metabolites

can also serve as precursors of eicosanoids and leukotrienes which are important

signaling molecules in inflammation and second messengers of the central nervous

system. Recently, EFA and LCPUFA (along with other non-essential fatty acids) were

reported as potent ligands for nuclear receptors, which regulate the gene expression of

genes involved in several metabolic processes.10

Figure 1 Essential fatty acids of the ω-3 and ω-6 family, with their source and long chain polyunsaturated metabolites (LCPUFA) and enzymes involved in desaturation and elongation of EFA and LCPUFA.

Under certain circumstances, for example during excessive intake of dietary LA or during

low metabolism of LA, LA can be stored in adipose tissue for future use.11

Since

(preterm) infants have limited amounts of adipose tissue and are rapidly growing and

developing, they are highly dependent on sufficient and continuous intake of dietary

EFA.

Within the enterocytes of the small intestine, absorbed EFA and LCPUFA are mainly re-

acylated into triglycerides and subsequently assembled into chylomicrons in order to be

excreted into the lymph. However, resident EFA are incorporated in membrane

phospholipids, which are mainly rich in LA and its metabolite arachidonic acid (C20:4ω-

6, AA). The relatively short lifespan of the enterocytes requires a continuous and rapid

supply of EFA and their metabolites, either from dietary, from biliary or from systemic

origin. Around 40% of bile consists of EFA- or LCPUFA-acyl chains, making it a very

important supplier of intestinal EFA.12

High EFA requirements are needed for

INTRODUCTION

13

1

CH

AP

TE

R

morphological and dynamic structural changes in the intestinal mucosa. Therefore, the

intestinal mucosa is highly sensitive and adaptive to dietary changes in EFA.

As stated above, EFA deficiency in the small intestine can develop in times of low dietary

intake, enhanced metabolism, and/or malabsorption of (essential) fatty acids. In general,

severe EFA deficiency can lead to growth retardation, skin lesions, reduced vision,

impaired cognitive development and steatosis. The symptoms caused by ω-6 fatty acid

deficiency are more obvious than those caused by ω-3 fatty acid deficiency.

Essential fatty acid (EFA) deficiency

The supply of EFA in the Western diet is usually sufficient to fulfill the metabolic needs.

Some chronic intestinal disorders can lead to severe fat malabsorption and thus to EFA

deficiency. However, most common is the EFA deficiency due to reduced fat absorption

as a consequence of reduced bile secretion in patients with cholestatic liver diseases or

reduced activity of pancreatic enzymes, like for example in patients with cystic fibrosis

(CF).13,14,15

EFA deficiency itself aggravates the fat malabsorption in these patients

leading to even more severe symptoms.4,3,16

Symptoms of EFA deficiency are usually not immediately obvious, especially not for

isolated ALA deficiency. It is therefore important to have another, preferably biochemical,

marker to assess EFA deficiency in (pediatric) patients. Plasma measurements of total

lipid LCPUFA are relatively easy and can function as an indication of EFA status. Yet,

plasma EFA composition may not correspond to the EFA status of various organs, but

may rather correlate more closely with recent dietary EFA intake. The EFA composition

in membrane phospholipids of erythrocytes may be a better indicator of body EFA status,

based on their relatively long half lives.17

This, on the other hand, might only be relevant

during EFA assessment in severe, long lasting EFA deficiency. Neither plasma nor red

blood cell phospholipid measurements are likely completely representative for complete

EFA status, since different tissues are known to have their own specific requirements

and metabolism of EFA. Unfortunately, it is clinically impossible to determine the EFA

status in the most relevant tissues, such as for example the central nervous system, and

therefore plasma or erythrocyte composition of LCPUFA is the most commonly used

parameter to assess EFA status. For estimation of the severity of combined deficiency of

ω-3 and ω-6 EFA, the so called triene:tetraene ratio has been introduced by Holman in

1960.18

In case of reduction of both ω-3 and ω-6 EFA, the synthesis of non-essential

fatty acids of the ω-9 family increases, leading to an enhanced production of the long

chain metabolite eicosanoic acid (C20:3ω-9, also known as mead acid) from oleic acid

(C18:1ω-9). The increase of the mead acid is an indicator of LA and ALA deficiency.

Since sufficient supply of one of the two EFA will prevent an increase in mead acid, this

ratio can only be used when the concentrations of both EFA are decreased. Although the

triene:tetraene ratio has been regarded for long as “the biochemical marker of EFA

deficiency”, it does not provide an overall picture of the EFA and their LCPUFA

metabolites.19

More common is the use of triene:tetraene ratio in combination with other

determinations of EFA (e.g. plasma profile) in order to obtain a more complete and

accurate picture of the EFA status in patients. The nature of the disease may influence

what the best clinically relevant marker is for a certain disease. Magbool et al. have

CHAPTER 1

14

recently demonstrated that in pediatric patients with CF, assessment of serum LA status

as the clinical indicator of EFA status is more relevant than the triene:tetraene ratio.19

Fatty acid compositions are most often represented as molar percentages, which

indicate the percentage of an individual fatty acid (or a group of fatty acids) as the

percentage of total fatty acids in plasma or other compartments. The relative, molar

percentages are often more relevant than absolute concentrations, since the latter do not

indicate the changes in membranes, which are mainly influenced by the composition.

As stated above, a high incidence of EFA deficiency has been reported during

cholestasis or CF. In both conditions, the small intestinal function seems to be

affected.13,14,15,20

The association of cholestasis and CF in relation to EFA deficiency will

be discussed in more detail in the next two paragraphs.

Essential fatty acid (EFA) deficiency in cholestasis

Cholestatic liver diseases (CLD) are characterized by decreased or absent hepatic

secretion of bile into the intestine, either caused by congenital or acquired diseases.21,22

CLD are associated with several nutritional complications, including EFA deficiency.23

In

general, neonatal and pediatric patients are more affected by CLD than adults. EFA

deficiency is one of the contributors to “failure to thrive” in pediatric patients with

cholestasis, known as cholestasis induced failure to thrive (CIFTT). Several treatment

options aim to reduce the cholestatic symptoms in pediatric patients, as well as their

negative impact on the nutritional condition. However, CIFFT can be very resistant to

treatment options, particularly in young children with end stage liver disease who require

liver transplantation.24,25

The most common cause of CLD in children requiring liver transplantation is biliary

atresia. Biliary atresia is a progressive disorder characterized by an inflammatory

reaction towards the extrahepatic and intrahepatic bile ducts, leading to destruction and

subsequent replacement of the normal tissue by fibrotic scar tissue. The etiology of

biliary atresia remains unknown, although an inflammatory reaction to a detrimental

stimulus seems to play an initiating role.26

Another group of causes for pediatric CLD

involve Progressive Familial Intrahepatic Cholestasis (PFIC). PFIC constitute a group of

genetically transmitted disorders, inherited in an autosomal recessive fashion. Three

phenotypic forms of PFIC have been characterized and attributed to gene defects in

three different genes (PFIC1-3; official symbols: ATP8B1, ABCB11, ABCB4).27

Another

cause of CLD is non-syndromic paucity of the intrahepatic bile ducts, whose etiology is

still enigmatic, infections, chromosomal disorders and metabolic disorders have been

suggested to play a role.28

Inborn errors in bile acid synthesis account for another part of

the children with CLD. Defects have been identified in enzymes catalyzing cholesterol

catabolism and bile acid synthesis.28

CLD in adolescents and young adults is often due

to autoimmune hepatitis, primary biliary cirrhosis or primary sclerosing cholangitis.28,29

Poor dietary intake is an important factor in the pathophysiological basis of malnutrition

in children with CLD. The nutritional status may be further compromised by decreased

absorption of the macronutrients, fat, carbohydrates and proteins. At infant age, fat

accounts for the most important dietary energy source (up to 50% of total ingested

energy). It is therefore, not surprising that up to 70% of children with CLD requiring liver

transplantation have biochemical indications of EFA and LCPUFA deficiency.24,25

INTRODUCTION

15

1

CH

AP

TE

R

Several studies demonstrated the decreased uptake and/or intracellular processing in

the enterocyte as the main reason for decreased EFA and LCPUFA concentrations

during EFA deficiency,3,4

rather than the decreased activity of desaturases and/or

elongases as proposed earlier by Socha et al.14

In addition, cholestasis has been

proposed to impair the β-oxidation pathway and can therefore interfere with the last step

of DHA and DPA metabolism from their precursors.2,30

However, in an animal model for

cholestasis (rats with bile duct ligation), Minich et al. showed no major difference in LA

oxidation, thus showing no support for this concept.31

Essential fatty acid (EFA) deficiency in cystic fibrosis

Cystic fibrosis (CF) is still one of the most common genetic disorders among the

Caucasian population.20

It is an autosomal recessive disease caused by a mutation in

the cystic fibrosis transmembrane conductance regulator (CFTR) gene. The encoded

CFTR protein mainly functions as a chloride channel.20

Over 1500 mutations have been

identified in the CFTR gene. However, for only a small number of these mutations the

functional importance has been elucidated. Symptoms of CF are age- and patient-

dependent, but most of them involve gastrointestinal, pulmonary, endocrine and

reproductive disorders.32

Gastrointestinal problems include meconium ileus (obstructive condition of the small

intestine) and pancreatic insufficiency, which both lead to malnutrition and failure to

thrive.33,34

In a sub selection of patients, cirrhosis and cholestatic symptoms may develop

in CF patients which contribute to an even further detoriation of the nutritional status.35,36

EFA deficiency has been common in CF, particularly in the era that many patients were

treated with low-fat diets to counteract the steatorrhoea, and was mainly characterized

by low plasma levels of linoleic acid (C18:2ω-6, LA).37,38

Number of events has been

suggested to contribute to EFA deficiency in CF patients, like pancreatic insufficiency,

solubilization defects, altered intestinal microclimate and altered enzyme activity of

desaturases and elongases involved in EFA metabolism. Additionally, increased energy

expenditure is thought to contribute to the poor nutritional status in CF patients.39

Several attempts to correct for EFA deficiency, with pancreatic enzyme replacement

therapy and linoleic acid supplementations, in CF patients have shown variable

effects.40,41,42,43

SMALL INTESTINE

The small intestine is one of the largest and most metabolically active tissues. It is

continuously renewed by processes of proliferation and differentiation, leading to a highly

ordered tissue architecture.44

Enterocytes are responsible for the absorption of dietary

and endogenous compounds. Enterocyte differentiation can be studied by assessment of

the expression of brush border enzymes, such as lactase and sucrase-isomaltase. The

three transcription factors Gata-4, Hnf1α and Cdx2 regulate the expression of the

corresponding genes.45

Cdx2 also plays an essential role in the organogenesis of the

midgut into the small intestine.46

Enterocytes located within different regions of the small

intestine (duodenum, jejunum and ileum) vary in their functional capacities; while for

example the carbohydrate absorption mainly takes place in the more proximal part, bile

CHAPTER 1

16

salt uptake occurs mainly in the terminal ileum. Small intestinal morphology is

characterized by two distinct axes, the horizontal axis, i.e. the proximal to distal (or

anterior to posterior) small intestine, and the vertical axis, representing the crypt-to-villus

distinction in the enterocytes.47

Crypts and villi

Already during the formation of the primitive gut at gestational age of 9 weeks in

humans, the morphogenesis of nascent villi and crypts occurs within the epithelium

(Figure 2). At this point, cellular proliferation is concentrated mainly within the intervillus

region. During development, the intervillus regions are transformed into crypts by means

of cellular penetration into the mesenchyme.

Figure 2 Histological staining of mouse small intestine indicating the crypt and villus regions. Crypts are located at the bottom and contain stem cells which migrate towards the upper located villus region. Fully differentiated cells represent mainly the absorptive cells, the enterocytes.

It has been accepted that the small intestinal epithelium is maintained by a population of

tissue-specific stem cells.48

Developed crypts contain a small, proliferating group of stem

cells which give rise to different intestinal cell types, subsequently migrating towards the

adjacent villi.48,49

In most mammals, re-differentiation of the intestine starts after birth,

simultaneous with increased proliferation. Increased proliferation eventually leads to

development of larger crypt depth and increased villus height. Parallel with the

development of the crypt and villi, different cell lineages develop from the immature cells,

including absorptive cells (enterocytes), mucus secreting cells (goblet cells), various

enteroendocriene cells and enzyme- and antibacterial peptides secreting cells (Paneth

cells).48,50,51,52

All the epithelial cells, originate from the same multipotent stem cells that

proliferate from the bottom of the crypt. Enterocytes account for almost 90% of all

epithelial cells within the small intestine. Research described in this thesis focuses on the

enterocytes, the most relevant intestinal cell type with regard to absorption and

metabolism of dietary compounds.

Apical and basolateral compartment in the enterocyte

Enterocytes are absorptive intestinal cells characterized by two domains within the cell,

Villus region

(differentiating cells)

Crypt region (stem cells)

INTRODUCTION

17

1

CH

AP

TE

R

namely the apical and the basolateral domain,53

separated by the tight junctions.54

It is

the existence of these domains that plays an important role in the maintenance of the

intestinal barrier function. One of the most remarkable features of the absorptive

enterocytes is the presence of the so called brush border membrane (BBM) at the apical

site of the cell which consists of many closely packed microvilli.53

Apical and basolateral

domains differ in their expression of different enzymes and transporters.55

The

histocompatibility antigens are specifically located at the basolateral membrane of the

enterocyte.56

Enzymes (hydrolases) appear only within the BBM at the end of the

physiological differentiation process of the enterocyte, i.e. when the proliferating cells

reach the villi during their migration from the crypts. For this reason, many hydrolytic

enzymes, like lactase and sucrase isomaltase, are used as the markers for the

differentiation status of the absorptive enterocytes.57

The exact pathways by which

enterocytes deliver different newly synthesized proteins from the Golgi apparatus to the

apical or basolateral site are still the subject of intense cell biological research.

Enterocyte function

Small intestine is one of the first barriers to be encountered by nutrients after their dietary

ingestion. Absorption of most important dietary and hepatobiliary compounds is

described below. Figure 3 shows a short schematic summary of enzymes and proteins

involved in processes of absorption and metabolism of the small intestine.

Fat absorption

Dietary fat is mainly absorbed as triglycerides in the human diet and in smaller amounts

as phospholipids (~10%).58

Intestinal absorption of fat can be separated in intraluminal

and intracellular events, and these have been reviewed extensively.58,59,60,61

Intraluminal

steps of fat absorption can be divided into emulsification, lipolysis, and solubilization,

followed by translocation across the epithelial apical membrane. Emulsification involves

mechanical disruption and partial hydrolysis of triglycerides within the stomach and

results in increasing the oil-water surface area by decreasing the median size of the fat

droplets from the diets. This process is stimulated mechanically by shear force in the

stomach and the pylorus and biochemically by generating the lipolytic products of

triglycerides, namely diacylglycerol and free fatty acids.

The emulsified dietary fat subsequently enters the first part of the small intestine, the

duodenum, where it is subjected to lipolysis by pancreatic lipases into monoacylglycerlol

and free fatty acids. Triglyceride lipolysis by pancreatic lipases requires a co-factor,

pancreatic co-lipase, which is able to facilitate proper binding of the lipase to the oil-

water surface of the fat emulsion. The secretion of the pancreatic lipase into the

duodenum is often associated with gallbladder contraction and cholecystokinin release.

Digestion of phospholipids derived from diet and bile, occurs within the duodenum by the

enzyme phospholipase A2. However, as demonstrated by studies in PLA2-deficient mice,

additional enzyme(s) can compensate for pancreatic PLA2 in catalyzing phospholipid

digestion.62

Hydrolysis of phospholipids results in production of lyso-phospholipids and

free fatty acids, which can subsequently be translocated across the enterocyte apical

membrane. Lipolytic products must be solubilized in order to be soluble and thus

transportable in the aqueous phase of the intestinal lumen and across the so called

CHAPTER 1

18

unstirred water layer, which is the border between the luminal site of the intestine and

the BBM of the enterocytes. Solubilization of the lipolytic products is performed by biliary

bile salts and phospholipids by means of their mixed micellar formation with the products

of lipolysis.63

Mixed micellar solubilization increases the solubility of the lipolytic products

up to 1000-fold.64

Compared to diglycerides and fatty acids, phospholipids are more

independent of bile salts for the mucosal uptake, since they can interact more easily with

water molecules.

Figure 3 Major small intestinal enzymes, proteins and nuclear receptors involved in fatty acid, carbohydrate, cholesterol and bile salt absorption and metabolism. On the left, the apical site with the brush border membrane and on the right the basolateral site of the enterocyte is indicated. ABCA1, Abc-transporter a1; ABCG5/8, Abc-transporter g5/g8; ACAT2, Acyl-coenzyme A:cholesterol Acyltransferase 2; ASBT, Apical sodium dependent bile acid transporter; DGAT1/2, Acyl coenzyme A:diacylglycerol acyltransferase 1/2; FABP, Fatty acid bindind protein; FAT, Fatty acid transporter; FATP4,fatty acid transport protein; FGF15, Fibroblast growth factor 15; FGFR4, Fibroblast growth factor receptor 4; FXR, Farnesoid X receptor, GLUT2/5, Glucose transporters 2/5; IBABP, Ileal bile acid binding protein; LDLR, Low density lipoprotein receptor; LXR, Liver X receptor; MGAT, monoacylglycerol O-acyltransferase 1; MTTP, microsomal triglyceride transfer protein; NPC1L1, Niemann-Pick C1 like 1; OSTα/ß, Heteromeric organic solute transporter alpha-beta; PPAR, Peroxisome proliferator-activated receptors; RXR, Retinoid X receptor; SI, Sucrase isomaltase; SGLT, Sodium glucose cotransporter; SHP, short heterodimer partner; SRBI, scavenger receptor BI.

After lipolysis and solubilization, fatty acids dissociate from different lipid classes

(micelles, liposomes, liquid crystalline vesicles or free phospholipids) within the unstirred

INTRODUCTION

19

1

CH

AP

TE

R

water layer.65

Translocation of the free fatty acids and across the BBM subsequently

occurs. Whether this translocation occurs only via passive diffusion, or in addition via fat

transporters in the enterocytes, remains unclear. Several candidate transporters have

been identified to facilitate fat transport across the BBM of the enterocyte, including the

fatty acid transport protein 4 (FATP4; official symbol SLC27A4) and the fatty acid

translocase (FAT; official symbol CD36), both located at the BBM of the

enterocytes.66,67,68

However, FATP4 has been shown to localize within the enterocyte as

well and thus not exclusively at the BBM.69

More importantly, several studies in mice with

deletions in these transporters clearly have indicated that these transporters are not

essential. Rather, they might co-facilitate in dietary fatty acid absorption or influence their

intracellular processing, as these mice do not show severe signs of fat

malabsorption.70,71

Within the enterocyte re-esterification and chylomicron formation

occur, starting with the binding of fatty acids to the intestinal fatty acid binding protein

(IFABP; official symbol FABP2) or liver fatty acid binding protein (LFABP; official symbol

FABP1) which escort them to the endoplasmatic reticulum.72

Interestingly, I-FABP

deficiency in mice does not lead to fat malabsorption, indicating that I-FABP is not

essential for sufficient absorption of dietary fat.

Within the smooth endoplasmatic reticulum absorbed fatty acids are acylated into

triglycerides via two different pathways. Under physiological conditions, the so called

monoacylglycerol pathway is the predominant one in which 1 acetylated fatty acid

molecule and 2 molecules of monoacylglycerol are re-esterificated into triglycerides. The

enzymes involved in the two steps of monoacylglycerol pathway are acyl-

CoA:monoacylglycerol acyltransferases (MGATs). These enzymes convert

monoacylglycerol and fatty acyl-CoA into diacylglycerol. Acyl-CoA: diacylglycerol

acyltransferases (DGATs), on the other hand, convert diacylglycerol intro triglycerides.

The second, physiologically less prominent route is the α-glycerophosphate pathway,

which becomes of major importance under conditions of fat malabsorption. In the first

two steps, glycerol-3-phosphate is converted into phosphatidic acid by means of

glycerol-3-phosphate acyltransferases and 1-acylglycerol-3-phosphate O-

acyltransferases. Subsequently, phosphatidic acid is converted to diacylglycerol by PA

phosphatases. Diacylglycerol produced by this pathway is preferentially used to

synthesize new phospholipids. The rest of the diacylglycerol is used to produce

triglycerides, which are thought to be slightly different from triglycerides produced by the

monoacylglycerol pathway, since the triglycerides from the latter pathway are

transported faster across the basolateral membrane of the enterocytes.

The last step of lipid absorption involves the assembly of newly produced triglycerides,

phospholipids, cholesterol, cholesteryl esters and apolipoproteins (mainly apoB48) into

pre-chylomicrons.73

This process requires the microsomal triglyceride transfer protein

(MTTP) within the smooth endoplasmatic reticulum. Afterwards, these pre-chylomicrons

are transported towards the Golgi apparatus, where they transform into mature

chylomicrons, These are eventually released in the cytoplasm and exocytosed into the

interstitium, ending up in lymphe.

Peroxisome proliferator-activated transcription factors (PPARs)

Recently, fatty acids have been identified as the natural ligands for peroxisome

CHAPTER 1

20

proliferator-activated transcription factors (PPARs) α, β/δ and γ which, like the other

nuclear receptors, heterodimerizes with retinoid X receptor (RXR)74

. Although the

functions of PPARs have been studied extensively in the liver, their role in the intestine is

still emerging. PPARα activation in the intestine has recently been demonstrated to

activate the transcription of several genes involved in fatty acid, triacylglycerol, sterol and

bile acid metabolism.75

PPARδ activation, on the other hand, was recently shown to

reduce intestinal cholesterol absorption efficiency. 76

PPARγ within the intestine has

recently been implied in modulating epithelial and mucosal inflammation.

Carbohydrate absorption

Carbohydrates in diet are derived from starch (polysaccharides, 75%) or sugars (di- and

monosaccharides). Starch is composed of amylase and amylopectin, and is digested by

salivary and pancreatic amylases. Afterwards, final hydrolysis to glucose at the brush

border of the enterocytes in the proximal part of the small intestine occurs by sucrase-

isomaltase and maltaseglycoamylase. Glucose can subsequently be taken up by the

sodium-dependent glucose transporter (SGLT1; official symbol SLC5A1).55

Lactose and

sucrose are the quantitatively most important dietary disaccharides. They are hydrolyzed

into glucose and galactose or fructose, respectively. Hydrolysis of lactose and sucrose is

catalyzed by the enzymes lactase and sucrase isomaltase, respectively, anchored within

the brush border of the enterocytes. Monosaccharides are transported directly across the

BBM by means of SGLT1 (glucose and galactose) or GLUT5 (fructose; official symbol

SLC2A5), without requiring hydrolysis.77

Subsequently, basolateral transport of all

carbohydrates occurs via the universal GLUT2 (official symbol SLC2A2) transporter.77,78

Studies in rats with bile duct ligation demonstrated that cholestasis is not associated with

severely affected absorption and digestion of carbohydrates.79,80

However, whether EFA

deficiency affects digestion and absorption of dietary carbohydrates is not known.

Cholesterol absorption

Between 25% and 85% of dietary cholesterol is absorbed from the small intestine in

humans.81,82

Once in the lumen of proximal small intestine, cholesterol and plant sterols

are most likely transported into the enterocyte by means of the recently identified, apical

transporter Niemann-Pick C1-like 1 protein (NPC1l1).81

The function of this protein can

be illustrated by the phenotype of mice lacking NPC1l1 protein, showing severely

reduced cholesterol absorption compared to their wild type littermates.81

Within the

enterocytes, cholesterol is esterified into cholesteryl esters by means of the acyl-

coenzyme A:cholesterol acyltransferase 2 (ACAT2), which has a high affinity for

cholesterol, but not for plant sterols.83

This results in packaging of the cholesteryl esters

into chylomicrons, which are subsequently secreted into the circulation. Recent studies

demonstrated that a fraction of the enterocytic cholesterol can be secreted into the

circulation independent from the chylomicron pathway. Direct secretion across the

basolateral membrane occurs in monomeric form, to be subsequently incorporated into

the HDL particles.84

This basolateral transport occurs via the ATP binding cassette

transporter 1 (ABCA1).84

Scavenger receptor class B, member 1 (SR-BI; official symbol

SCARB1) and LDL receptor (LDLR), localized at the basolateral site of the enterocyte,

can reabsorb selectively the cholesteryl esters, without absorption of the remnants of the

INTRODUCTION

21

1

CH

AP

TE

R

HDL particle.85

Unesterified plant sterols are not assembled for basolateral secretion, but

are transported back to the intestinal lumen along with unesterified cholesterol. This

apical transport of plant sterols and unesterified cholesterol from the enterocyte into the

lumen is facilitated by an ABC heterodimeric transporter ABCG5/ABCG8.86

Within the

enterocyte, the nuclear liver X receptor (LXRα and LXRβ; official symbols NR1H3 and

NR1H2) is expressed, which tightly regulates cholesterol and fatty acid metabolism by

inducing the transcription of genes involved in these metabolic pathways (ABC

transporters, SREBP1c and SREBP2; official symbols SREBF1 and SREBF2).87

Until recently, hepatobiliary secretion of cholesterol has been thought as the most

prominent way of cholesterol excretion from the body. This is rather peculiar, since

already in 1927 an alternative pathway has been proposed, involving direct secretion

from the intestine. However, this latter pathway has never been validated or paid

sufficient scientific attention. Recently, the alternative pathway has become re-

appreciated, since in various conditions and models the fecal excretion of neutral sterols

was higher than the sum of dietary and biliary cholesterol entering the intestinal

lumen.88,89

Direct transintestinal pathway for cholesterol excretion (TICE) has been

demonstrated in mice by van der Velde et al.90

The capacity of the intestinal cholesterol

excretion pathway was exactly sufficient to account for the missing cholesterol and twice

as high as the quantitative hepatobiliary secretion. This observation indicated the

relevance of TICE in excretion of cholesterol in mice.90

Importance of TICE in other

species has not been studied in detail so far. TICE was demonstrated to depend on the

dietary fat content.91

The EFA deficiency might, therefore, be associated with alterations

in TICE. However, the effects of EFA deficiency on cholesterol metabolism in the

intestine have not been studied so far.

Small intestine and the enterohepatic circulation of bile salts

Bile salts are synthesized in the liver from cholesterol via the neutral or the acidic

pathway.92

Under physiological conditions, bile salts are subsequently secreted via bile

into the intestine. Within the intestine the bile salts are almost completely reabsorbed;

only around 5% of the endogenous bile salts escape the reabsorption and is excreted via

the feces every day. Unconjugated bile salts in the small intestine and in colon can be

transported passively.93,94

However, conjugated bile salts require facilitated transport

across the BBM. This is achieved by means of the apical sodium-dependent bile salt

transporter (ASBT/ISBT; official symbol SLC10A2), mainly expressed in the terminal

ileum. The intracellular transport of bile acids from the apical to the basolateral

compartment was thought to be facilitated by the ileal bile acid binding protein (IBABP;

official symbol FABP6);95,96

however, the exact role of IBABP in the intracellular

trafficking of bile salts is still under debate.97

Within the cell, bile salts can bind to and

activate the nuclear hormone farnesoid receptor (FXR; official symbol NR1H4), which is

an important regulator of bile salt homeostasis.98,99

Activated FXR initiates the

transcription of a whole cascade of genes important for bile salt metabolism. One of

these genes is the small heterodimer partner (SHP; official symbol NR0B2), which leads

to subsequent ASBT repression.100,101

Another intestinal protein which is tightly regulated

by the activated FXR is the fibroblast growth factor 19 (FGF19, mouse homologue is

Fgf15).102

Upon FXR activation, FGF19 is released into the circulation, in order to be

CHAPTER 1

22

transported to the liver.102,103

In the liver, FGF19 binds to the fibroblast growth factor

receptor 4 (FGFR4) on the hepatocyte cell membrane. This binding leads to the

activation of the JNK pathway and repression of cholesterol 7-α-hydroxylase (CYP7A1)

and sterol 12-α-hydroxylase (CYP8B1), resulting in decreased bile salt synthesis.102,103

Recent studies demonstrated that in addition to intestinal/hepatic FGF19/FGFR4

signaling pathway, liver FGFR4/FGF19 pathway might exist to protect the liver under

conditions of bile salt accumulation.103,104

Another study reported expression of FGFR4

at the basolateral site of the enterocytes and in cholangiocytes, suggesting the existence

of a feedback loop mechanism of FGF19/FGFR4 within the intestine and bile ducts.

Excretion of bile salts in the enterocytes occurs via basolaterally localized heterodimeric

organic solute transporter OSTα-OSTβ.105,106

Manifestation of the impaired small intestinal function in common intestinal

disorders

Two common small intestinal disorders with a profound impact at pediatric age are celiac

disease and Crohn’s disease. Both conditions can severely affect small intestinal

morphology and function, and lead to malabsorption to nutrients and to growth failure.

Celiac disease is a form of autoimmune disease of the small intestine leading to nutrient

malabsorption and immune reaction to transglutaminidase in genetically predisposed

subjects. It is a life-long condition characterized by villous atrophy (blunted villi),

enhanced cell proliferation, increased number of crypts and increased infiltration of

lymphocytes upon ingestion of gluten. Symptoms vary largely among the patients and

disappear upon a gluten-free diet.107

Crohn’s disease is anti-inflammatory disease which

can affect the whole gastrointestinal tract. Within the small intestine neutrophil infiltration

into the epithelium can occur along with atypical crypt branching and finally with villous

blunting.108

Intestinal permeability might also be profoundly increased, associated with

an impaired barrier function.109

Together, these pathophysiological factors can lead to

malabsorption of nutrients and growth failure. The exact factors involved in

pathophysiology of EFA deficiency in the small intestine which lead to nutrient absorption

remain unclear. Therefore, it is useful to study how EFA deficiency affects the small

intestinal function and morphology.

AIM AND THE OUTLINE OF THE THESIS

Clinical conditions associated with EFA deficiency are accompanied by impaired

nutritional status. In children with cholestasis, EFA deficiency aggravates the cholestasis

induced failure to thrive (CIFTT). In animal models, EFA deficiency by itself is associated

with malabsorption of fat, even in absence of cholestasis or CF. Previous studies

suggested that defects in the small intestine during EFA deficiency were located at the

intracellular level. We aimed to characterize and unravel the effects of EFA

deficiency on the pathophysiology and the function of the small intestine.

First, we studied the epithelial histology and function by analyzing the morphology and

nutrient absorption of the small intestine during EFA deficiency (chapter 2). We describe

the effects of EFA deficiency in mice on the absorption of carbohydrates and on the

expression of lactose, relevant small intestinal differentiation marker. By means of the

INTRODUCTION

23

1

CH

AP

TE

R

administration of stably labeled glucose and lactose, we determined the absorption and

digestion of these compounds in vivo. In chapter 3 we further characterized the effects

of EFA deficiency on intestinal physiology by determining the jejunal cholesterol

absorption and metabolism during EFA deficiency. The results obtained are based on

the physiological parameters and the microarray analysis of mouse jejunal tissue.

In chapter 4 we determined the effects of EFA deficiency on the enterohepatic

circulation (EHC) of bile salts. Bile salt (re)absorption is a small intestinal function which

does not depend on the jejunal intestinal epithelium, but rather on that of the terminal

ileum. In order to study whether EFA deficiency differentially affects different small

intestinal segments, we studied the EHC in EFA-deficient mice. Small intestine plays an

important role in the EHC of bile salts by regulating the feedback mechanism of the

hepatic bile salt synthesis. Previous studies in EFA-deficient mice revealed elevated bile

salt secretion and bile flow. The underlying mechanism of this finding remained unclear.

We determined several parameters of the enterohepatic circulation of bile salts using the

stable isotope dilution technique, combined with bile duct cannulation. Small intestinal

regulatory mechanisms of the enterohepatic circulation were assessed by analyzing the

expression of the intestinal genes implicated in bile salt metabolism.

In order to study in more detail the intracellular effects of EFA deficiency on the small

intestine, an in vitro model of EFA deficiency has been established. Differentiating Caco-

2 cells cultured in EFA-deficient or control medium were characterized and validated as

a model for EFA deficiency (chapter 5). We described the effects of EFA deficiency on

cell differentiation, gene expression and morphology, based on several in vitro

experiments in EFA-deficient Caco-2 cells.

To optimize nutritional condition during cholestatic liver disease, one could aim to

decrease the fat malabsorption by administration of exogenous absorption enhancers.

Chapter 6 describes experiments in two different rat models of fat malabsorption; one

with impaired lipolysis (pancreatic insufficiency model) and one with reduced

solubilization (cholestatic model). In these rat models we studied the effects of the

compound Gelucire®44/14 on fat malabsorption in vivo. Gelucire

®44/14 is currently used

to improve the absorption of poorly soluble drugs. Fat absorption was assessed in both

models, at the level of lipolysis and solubilization, respectively, after the administration of

Gelucire®44/14.

Chapter 7 provides a summary of the most relevant findings in this thesis and future

perspectives for EFA deficiency-related research.

REFERENCES

1 Das UN. Biotechnology Journal 2006; 1(2): 420-439.

2 Smit E, Muskiet F, and Boersma ER. Prostaglandins, Leukotrienes and Essential Fatty Acids 2004; 71(4):

241-250.

3 Levy E, Garofalo C, Thibault L, Dionne S, Daoust L, Lepage G, and Roy CC. Am J Physiol Gastrointest

Liver Physiol 1992; 262(2): G319-G326.

4 Werner A, Minich DM, Havinga H, Bloks VW, van Goor H, Kuipers F, and Verkade HJ. Am J Physiol

Gastrointest Liver Physiol 2002; 283(4): G900-G908.

5 Das UN. Curr Pharm Biotechnol 2006; 7(6): 467-482.

6 Rosenthal MD. Progress in Lipid Research 1987; 26(2): 87-124.

CHAPTER 1

24

7 Cinti DL, Cook L, Nagi MN, and Suneja SK. Progress in Lipid Research 1992; 31(1): 1-51.

8 Brenner RR. Progress in Lipid Research 1981; 20: 61-7.

9 Cunnane SC. Progress in Lipid Research 2003; 42(6): 544-568.

10 Bensinger SJ and Tontonoz P. Nature 2008; 454(7203): 470-477.

11 Hodson L, Skeaff CM, and Fielding BA. Progress in Lipid Research 2008; 47(5).

12 Melin T, Qi C, and Nilsson A. Prostaglandins Leukot Essent Fatty Acids 1996; 55(5): 337-343.

13 Cabré E and Gassull MA. Nutrition 2009; 12(7-8): 542-548.

14 Socha P, Koletzko B, Swiatkowska E, Pawlovska J, Stolarczyk A, and Socha J. Acta Peadiatr 1998;

87(3): 278-283.

15 Levy E, Roy C, Lacaille F, Lambert M, Messier M, Gavino V, Lepage G, and Thibault L. Am J Clin Nutr

1993; 57(4): 573-579.

16 Lukovac S, Los EL, Stellaard F, Rings EHHM, and Verkade HJ. Am J Physiol Gastrointest Liver Physiol

2008; 295(3): G605-G613.

17 Siguel EN and Lerman RH. Metabolism 1996; 45(1): 12-23.

18 Holman R. J Nutr 1960; 70(3): 405-410.

19 Maqbool A, Schall JI, Garcia-Espana JF, Zemel BS, Strandvik B, and Stallings VA. Journal of Pediatric

Gastroenterology and Nutrition 2008; 47(5).

20 Peretti N, Marcil V, Drouin E, and Levy E. Nutr Metab (Lond) 2005; 2(1): 11.

21 Elferink RO. Gut 2003; 52(Suppl 2): ii42-ii48.

22 O'Leary JG and Pratt DS. Curr Opin Gastroenterol 2009; 23(3): 232-236.

23 Los EL, Lukovac S, Werner A, Dijkstra T, Verkade HJ, and Rings EH. Nestle Nutr Workshop Ser Pediatr

Program 2007; 59: 147-157.

24 Chin SE, Sheperd RW, Thomas BJ, Cleghom GJ, Patrick MK, Wilcox JA, Ong TH, Lynch SV, and Strong

R. Am J Clin Nutr 1992; 56(1): 164-168.

25 Protheroe SM and Kelly DA. Baillieres Clin Gastroenterol 1998; 12(4): 823-841.

26 Middlesworth W and Altman RP. Curr Opin Pediatr 1997; 9(3): 265-269.

27 Davit-Spraul A, Gonzales E, Baussan C, and Jacquemin E. Orphanet Journal of Rare Diseases 2009;

4(1): 1.

28 Poupon R, Chazouillères O, and Poupon RE. J Hepatology 2000; 32(1 Suppl): 129-140.

29 Narkewicz MR. Current Opinion in Pediatrics 2001; 13(5).

30 Krähenbühl S, Talos C, and Reichen J. Hepatology 1994; 19(5): 1272-1281.

31 Minich D, Havinga R, Stellaard F, Vonk R, Kuipers F, and Verkade H. Am J Physiol Gastrointest Liver

Physiol 2000; 279(6): G1242-G1248.

32 Welsh MJ and Fick RB. J Clin Invest 1987; 80(6): 1523-1526.

33 Strandvik B. Acta Paediatr Scand Suppl 1989; 363: 58-63.

34 Lai HC, Kosorok MR, Laxova A, Davis LA, FitzSimmon SC, and Farrell PM. Pediatrics 2000; 105(1): 53-

61.

35 Sanko-Resmer J, Paczek L, Wyzgal J, Ziólkowski J, Ciszek M, Alsharabi A, Grzelak I, Paluszkiewicz R,

Patkowski W, and Krawczyk M. Transplant Proc 2006; 38(1): 212-214.

36 Colombo C and Battezzati PM. Eur J Gastroenterol Hepatol 1996; 8(8): 748-754.

37 Farrell PM, Mischler EH, Engle MJ, Brown DJ, and Lau SM. Pediatr Res 1985; 19(1): 104-109.

38 Hubbard VS, Dunn GD, and di Sant'Agnese PA. Lancet 1977; 2(8052-8053): 1302-1304.

39 Landon C, Kerner JA, Castillo R, Adams L, Whalen R, and Lewiston NJ. JPEN J Parenter Enteral Nutr

1981; 5(6): 501-504.

40 Kusoffsky E, Strandvik B, and Troell S. J Pediatr Gastroenterol Nutr 1983; 2(3): 434-438.

41 Chase HP, Cotton EK, and Elliot RB. Pediatrics 1979; 64(2): 207-213.

42 Baker SS, Borowitz D, and Baker RD. Curr Gastroenterol Rep 2005; 7(3): 227-233.

43 Gow R, Bradbear R, Francis P, and Sheperd R. Lancet 1981; 2(8255): 1071-1074.

44 de Santa Barbara P, van den Brink GR, and Roberts DJ. Cellular and Molecular Life Sciences 2003;

60(7): 1322-1332.

INTRODUCTION

25

1

CH

AP

TE

R

45 Boudreau F, Rings E, van Wering H, Kim R, Swain G, Krasinski S, Moffett J, Grand R, Suh E, and Traber

P. J Biol Chem 2002; 277(35): 31909-31917.

46 Gao N, White P, and Kaestner KH. Developmental Cell 2009; 166(4): 588-599.

47 Pacha J. Physiol Rev 2000; 80(4): 1633-1667.

48 Cheng H and Leblond CP. Am J Anat 1974; 141(4): 503-519.

49 Yen TH and right NA. Stem Cell Rev 2006; 2(3): 203-212.

50 Cheng H. Am J Anat 1974; 141(4): 521-535.

51 Cheng H. Am J Anat 1974; 141(4): 481-501.

52 Gordon JI, Schmidt GH, and Roth KA. FASEB J 1992; 6(12): 3039-3050.

53 Massey-Harroche D. Microsc Res Tech 2000; 49(4): 353-362.

54 Albers TM, Lomakina I, and Moore RP. Lab Invest 1995; 73(1): 139-148.

55 Thomson ABR, Keelan M, Thiesen A, Clandinin MT, Ropeleski M, and Wild GE. Digestive Diseases and

Sciences 2001; 46(12): 2567-2587.

56 Gonnella PA and Wilmore DW. J Cell Sci 1993; 106(3): 937-940.

57 Rings EH, de Boer PA, Moorman AF, van Beers EH, Dekker J, Montgomery RK, Grand RJ, and Buller

HA. Gastroenterology 1992; 103(4): 1154-1161.

58 Carey MC, Small DM, and Bliss CM. Annual Review of Physiology 1983; 45(1): 651-677.

59 Mu H and Høy CE. Prog Lipid Res 2004; 43(2): 105-133.

60 Minich DM, Vonk RJ, and Verkade HJ. J Lipid Res 1997; 38(9): 1709-1721.

61 Friedman HI and Nyland B. Am J Clin Nutr 1980; 33(5): 1108-1139.

62 Richmond BL, Boileau AC, Zheng S, Huggins KW, Granholm NA, Tso P, and Hui DY. Gastroenterology

2001; 120(5): 1193-1202.

63 Hofmann A and Borgström. J Clin Invest 1964; 43: 247-257.

64 Gerloff T, Stieger B, Hagenbuch B, Madon J, Landmann L, Roth J, Hofmann A, and Meier P. J Biol Chem

1998; 273(16): 10046-10050.

65 Wilson FA, Sallee VL, and Dietschy JM. Science 1971; 174(13): 1031-1033.

66 Drover VA, Nguyen DV, Bastie CC, Darlington YF, Abumrad NA, Pessin JE, London E, Sahoo D, and

Phillipis MC. J Biol Chem 2008; 283(19): 13108-13115.

67 Chen M, Yang Y, Braunstein E, Georgeson KE, and Harnom CM. Am J Physiol Endocrinol Metab 2001;

281(5): E916-E923.

68 Stahl A, Hirsch DJ, Gimeno RE, Punreddy S, Ge P, Watson N, Patel S, Kotler M, Raimondi A, Rataglia

LA, and Lodisch HF. Mol Cell 1999; 4(3): 299-308.

69 Milger K, Herrmann T, Becker C, Gotthardt D, Zickwolf J, Ehehalt R, Watkins PA, Stremmel W, and

Fullekrug J. J Cell Sci 2006; 119(22): 4678-4688.

70 Goudriaan J, Dahlmans V, Febbraio M, Teusink B, Romijn J, Havekes L, and Voshol P. Molecular and

Cellular Biochemistry 2002; 239(1 - 2): 199-202.

71 Coe NR, Simpson MA, and Bernlohr DA. J Lipid Res 1999; 40(5): 967-972.

72 Ockner RK and Manning JA. J Clin Invest 1974; 54(2): 326-338.

73 Hussain MM. Atherosclerosis 2000; 148(1): 1-15.

74 Berger J and Moller DE. Annual Review of Medicine 2002; 53(1): 409-435.

75 Bunger M, van den Bosch H, van der Meijde J, Kersten S, Hooiveld G, and Muller M. Physiol Genomics

2007; 30(2):192-204.

76 van der Veen JN, Kruit JK, Havinga R, Baller JFW, Chimini G, Lestavel S, Staels B, Groot PHE, Groen

AK, and Kuipers F. J Lipid Res 2005; 46(3): 526-534.

77 Keelan M, Walker K, and Thomson AB. Comp Biochem Physiol A 1985; 82(1): 83-89.

78 Swallow D, Poulter M, and Hollox E. Drug Metab Dispos 2001; 29(4): 513-516.

79 Los EL, Wolters H, Stellaard F, Kuipers F, Verkade HJ, and Rings EHHM. Am J Physiol Gastrointest

Liver Physiol 2007; 293(3): G615-G622.

80 Borges EL, Braga AA, and Petroianu A. Digestive Diseases and Sciences 1998; 43(10): 2196-2200.

CHAPTER 1

26

81 Altmann SW, Davis HRJr, Zhu LJ, Yao X, Hoos LM, Tetzloff G, Iyer SP, Maguire M, Golovko A, Zeng M,

Wang L, Murgolo N, and Graziano MP. Science 2004; 303(5661): 1201-1204.

82 Mettinen TA, Tilvis RS, and Kesaniemi YA. Am J Epidemiol 1990; 131(1): 20-31.

83 Lee RG, Willingham MC, Davis MA, Skinner KA, and Rudel LL. J Lipid Res 2000; 41(12): 1991-2001.

84 Brunham LR, Kruit JK, Verchere CB, and Hayden MR. J Clin Invest 2008; 118(2): 403-408.

85 Hui D and Howles P. Seminars in Cell & Developmental Biology 2005; 16(2): 183-192.

86 Berge KE, Tian H, Graf GA, Yu L, Grishin NV, Schultz J, Kwiterovich P, Shan B, Barnes R, and Hobbs

HH. Science 2000; 290(5497): 1771-1775.

87 Joseph SB and Tontonoz P. Current Opinion in Pharmacology 2003; 3(2): 192-197.

88 Kruit JK, Plosch T, Havinga R, Boverhof R, Groot PHE, Groen AK, and Kuipers F. Gastroenterology

2005; 128(1): 147-156.

89 van der Veen JN, van Dijk TH, Vrins CLJ, van Meer H, Havinga R, Bijsterveld K, Tietge UJF, Groen AK,

and Kuipers F. J Biol Chem 2009; 284(29): 19211-19219.

90 van der Velde AE, Vrins CLJ, van den Oever K, Kunne C, Oude Elferink RPJ, Kuipers F, and Groen AK.

Gastroenterology 2007; 133(3): 967-975.

91 van der Velde AE, Vrins CLJ, van den Oever K, Seemann I, Oude Elferink RPJ, van Eck M, Kuipers F,

and Groen AK. Am J Physiol Gastrointest Liver Physiol 2008; 295(1): G203-G208.

92 Lefebvre P, Cariou B, Lien F, Kuipers F, and Staels B. Physiol Rev 2009; 89(1): 147-191.

93 McClintock C and Shiau YF. Am J Physiol Gastrointest Liver Physiol 1983; 244(5): G507-G514.

94 Bahar RJ, Stolz A. Gastroenterol Clin North Am 1999; 28(1): 27-58.

95 Trauner M and Boyer JL. Physiol Rev 2003; 83(2): 633-671.

96 Lin MC, Kramer W, and Wilson FA. J Biol Chem 1990; 265(25): 14986-14995.

97 Kok T, Hulzebos C, Wolters H, Havinga R, Agellon L, Stellaard F, Shan B, Schwarz M, and Kuipers F. J

Biol Chem 2003; 278(43): 41930-41937.

98 Makishima M, Okamoto AY, Repa JJ, Tu H, Learned RM, Luk A, Hull MV, Lustig KD, Mangelsdorf DJ,

and Shan B. Science 1999; 284(5418): 1362-1365.

99 Dawson P, Haywood J, Craddock A, Wilson M, Tietjen M, Kluckman K, Maeda N, and Parks J. J Biol

Chem 2003; 278(36): 33920-33927.

100 Goodwin B, Jones SA, Price RR, Watson MA, McKee DD, Moore LB, Galardi C, Wilson JG, Lewis MC,

Roth ME, Maloney PR, Willson TM, and Kliewer SA. Molecular Cell 2000; 6(3): 517-526.

101 Neimark E, Chen F, Li X, and Shneirder BL. Hepatology 2004; 40(1): 149-156.

102 Holt JA, Luo G, Billin AN, Bisi J, McNeill YY, Kozarsky KF, Donahee M, Wang DY, Mansfield TA, Kliewer

SA, Goodwin B, and Jones SA. Genes Dev 2003; 17(13): 1581-1591.

103 Song KH, Li T, Owsley E, Strom S, and Chiang JY. Hepatology 2009; 49(1): 297-305.

104 Schaap FG, van der Gaag NA, Gouma DJ, and Jansen PL. Hepatology 2009; 49(4): 1228-1235.

105 Rao A, Haywood J, Craddock AL, Belinsky MG, Kruh GD, and Dawson PA. PNAS 2008; 105(10): 3891-

3896.

106 Ballatori N, Fang F, Christian WV, Li N, and Hammond CL. Am J Physiol Gastrointest Liver Physiol 2008;

295(1): G179-G186.

107 Kagnoff MF. Gastroenterology 2005; 128(4, Supplement 1): S10-S18.

108 Head K and Jurenka JS. Altern Med Rev 2004; 9(4): 360-401.

109 Laukoetter MG, Nava P, and Nusrat P. World J Gastroenterol 2008; 14(3): 401-407.

CHAPTER 2

ESSENTIAL FATTY ACID DEFICIENCY IN MICE IMPAIRS LACTOSE

DIGESTION

S. Lukovac1,2

, E.L. Los1,2,3

, F. Stellaard1, E.H.H.M. Rings

1, H.J. Verkade

1

(1) Pediatric Gastroenterology, Department of Pediatrics, Beatrix Children’s Hospital,

Groningen University Institute for Drug Exploration (GUIDE), Center for Liver, Digestive

and Metabolic Diseases, University of Groningen, University Medical Center Groningen,

Groningen, The Netherlands

(2) Both authors contributed equally to the study.

(3) Current address: Department of Cell Physiology, Section Osmoregulation, Radboud

University Nijmegen Medical Centre, Nijmegen, The Netherlands.

Am J Physiol Gastrointest Liver Physiol 2008; 295: G505-G513

CHAPTER 2

30

ABSTRACT

Essential fatty acid (EFA) deficiency in mice induces fat malabsorption. We previously

reported indications that the underlying mechanism is located at the level of the intestinal

mucosa. We have investigated the effects of EFA deficiency on small intestinal

morphology and function.

Mice were fed an EFA-deficient or control diet for 8 weeks. A 72h fat balance, the EFA

status, and small intestinal histology were determined. Carbohydrate absorptive and

digestive capacities were assessed by stable isotope methodology after administration of

U-13

C-glucose and 1-13

C-lactose. The mRNA expression and enzyme activity of lactase

and concentrations of the EFA linoleic acid (LA) were measured in small intestinal

mucosa.

Mice fed the EFA-deficient diet were markedly EFA-deficient with a profound fat

malabsorption. EFA deficiency did not affect the histology or proliferative capacity of the

small intestine. Blood 13

C6-glucose appearance and disappearance were similar in both

groups, indicating unaffected monosaccharide absorption. In contrast, blood appearance

of 13

C-glucose, originating from 1-13

C-lactose, was delayed in EFA-deficient mice. EFA

deficiency profoundly reduced the lactase activity (-58%, p<0.01) and mRNA expression

(-55%, p<0.01) in mid small intestine. Both lactase activity and its mRNA expression

strongly correlated with mucosal LA concentrations (r=0.77 and 0.79, resp., p<0.01).

EFA deficiency in mice inhibits the capacity to digest lactose, but does not affect small

intestinal histology. These data underscore the observation that EFA deficiency

functionally impairs the small intestine, which in part may be mediated by low LA levels

in the enterocytes.

EFA DEFICIENCY AND INTESTINAL FUNCTION

31

2

CH

AP

TE

R

INTRODUCTION

Essential fatty acid (EFA) deficiency can occur in cholestatic liver diseases as a

consequence of fat malabsorption.1,2

However, EFA deficiency itself also induces fat

malabsorption.3,4

The underlying mechanism of EFA deficiency induced fat

malabsorption remains unclear. Absorption of fat involves lipolysis, solubilization,

intestinal translocation from the lumen into the mucosa, chylomicron assembly and

transport into the lymph.5,6

Previous studies in EFA-deficient mice have indicated that

impaired lipolysis or bile formation do not cause the fat malabsorption in EFA deficiency.4

However, studies in rats show that EFA deficiency alters both the intraluminal and

intracellular phases of fat absorption. 3 This implies that the effects of EFA deficiency on

mucosal phases of fat absorption could be species-dependent.

Recently we reported data to suggest that EFA deficiency in mice affects fat absorption

at the level of the small intestinal mucosa.4 However, it has not been proven that EFA

deficiency impairs the mucosal phase of fat absorption. Based on previous findings, we

hypothesize that EFA deficiency functionally impairs the small intestine.

In contrast to fat absorption, the absorption of di- and monosaccharide carbohydrates

exclusively depends on mucosal function.7 The monosaccharide glucose is actively

transported across the brush border membrane in the small intestine by the brush-border

transporter Sodium-dependent glucose transporter (SGLT1).7,8

The disaccharide lactose

is first hydrolyzed by the mucosal membrane anchored lactase-phlorizin hydrolase

(lactase, LPH) into glucose and galactose, prior to their active transport across the brush

border by SGLT1.8,9

Besides being an important enzyme in lactose hydrolysis, lactase is

a marker of enterocyte differentiation.10

Throughout development total intestinal lactase

activity remains similar to that found in newborns.11

If EFA deficiency affects lactase

expression and activity in the small intestine, even slight changes should easily be

detectible in adult EFA-deficient mice. For this reasons, lactase is a good marker for

functional assessment of the small intestine in adult animal.

Essential fatty acids (EFAs) are structural components of membrane phospholipids.

Enterocyte membrane phospholipids are particularly rich in linoleic acid (LA, C18:2ω-

6),12

which is necessary for modulations of a wide variety of biological functions and for

physiochemical adaptations of the membrane lipid matrix to alterations in membrane

fluidity.13

The lipid matrix influences the conformation and function of proteins embedded

in the inner and/or outer leaflet of the membrane.14

Recently, an additional role of EFAs

in alterations of bilayer elastic properties and lipid composition in lipid rafts have been

reported.15,16

Through activation of peroxisome proliferator-activated receptors (PPARs),

EFAs can regulate transcriptional activity of several genes, including of those involved in

fatty acid transport and metabolism.17,18

In the present study we characterized the effects of EFA deficiency on small intestinal

morphology and function in mice. Korotkova et al. have shown that EFA deficiency

affects the fatty acid composition in the phospholipids of the rat small intestinal mucosa

by decreasing jejunal concentrations of linoleic acid.12

However, no studies have been

performed on the effect of EFA deficiency on the small intestinal function concerning

carbohydrate absorption. We assessed the absorption of glucose, a major source of

metabolic energy for mammalian cells,19

and the expression and activity of the lactase

CHAPTER 2

32

enzyme, as appropriate functional markers of the small intestine, in a previously

developed and characterized murine model of EFA deficiency.4 We applied stable

isotope methodology,20

since this approach allows extension to similar studies in patients

with EFA deficiency, cholestasis, or other forms of malabsorption.21,22,23

U-13

C-labeled

glucose and 1-13

C-labeled lactose were administered to EFA-deficient and control mice.

Blood appearance of labels derived from administered glucose (13

C6-glucose) and

lactose (13

C-glucose) into the blood glucose fraction was subsequently quantified. We

also determined the activity and expression of lactase, as well as the concentration of

LA, in the mucosa along the proximal-to-distal axis of the small intestine. To determine

whether EFA deficiency specifically affects lactase activity or disaccharide activity in

general, we in addition measured the activity of another disaccharide, sucrase.

Our data show that EFA deficiency is associated with impaired lactose digestion in mice.

This functional observation is specific for lactase and corresponds with lower lactase

mRNA expression and enzyme activity in the mid small intestine of EFA-deficient mice.

All together, these findings support the idea that EFA deficiency functionally impairs the

small intestine.

MATERIAL AND METHODS

Mice and housing

Wild type mice with a free virus breed background were obtained from Harlan (Horst, the

Netherlands). Male mice (25-35 g) were housed in a light-controlled (lights on 6 AM-6

PM) and temperature-controlled facility and were allowed tap water and chow (AB diets,

Woerden, the Netherlands) ad libitum. The experimental protocol was approved by the

Ethics Committee for Animal Experiments, Faculty of Medical Sciences, University of

Groningen, the Netherlands.

Material

U-13

C-glucose and 1-13

C-lactose were obtained from Isotec Inc. (Miamisburg, Ohio,

USA) with isotopic enrichments of 99%. Unlabeled lactose was obtained from Fluka

(Buchs, Switzerland).

Experimental diets

Similar to previous studies, we used high-fat EFA-deficient and EFA-sufficient (control)

diets (16 wt% and 34 energy% fat), in order to mimic more closely the human diet

composition.4 The diets were custom synthesized by Arie Blok BV (Woerden, the

Netherlands, diet codes EFA-deficient #4141.08 and EFA-sufficient #4141.07). The EFA-

deficient diet contained 64 mol% palmitic acid (C16:0), 18 mol% stearic acid (C18:0), 13

mol% oleic acid (C18:1ω-9) and 5 mol% linoleic acid (C18:2ω-6). The isocaloric EFA-

sufficient diet contained 36 mol% C16:0, 5 mol% C18:0, 31 mol% C18:1ω-9 and 29

mol% C18:2ω-6. Fatty acid contents of the diets were analyzed by extracting,

hydrolyzing and methylating total dietary fatty acids as described by Muskiet et al. and

subsequent separation and quantification of fatty acid methyl esters was performed by

gas chromatography as described previously.4,24

EFA DEFICIENCY AND INTESTINAL FUNCTION

33

2

CH

AP

TE

R

Experimental procedures

Induction of EFA deficiency

Mice were fed standard laboratory chow containing 6 weight% fat from weaning, were

switched to EFA-deficient or control diet at eight weeks of age. After eight weeks of EFA-

deficient or control diet, fat absorption was assessed by measuring food intake and

collecting feces for 72h. Mice underwent a glucose/lactose absorption test with U-13

C-

glucose and 1-13

C-lactose (details see below).25,26

After the test mice were anesthetized

and sacrificed by obtaining a large blood sample through cardiac puncture for

determination of erythrocyte EFA-status by the triene/tetraene (C20:3ω-9/C20:4ω-6)

ratio.4 The small intestine was excised, flushed with ice-cold PBS and divided into a

proximal, mid and distal segment of similar size. Smaller parts from the middle of each

small intestinal segment were harvested for histology and gene expression. Another part

of the proximal, mid and distal small intestine was opened lengthwise and the mucosa

was removed by scrapping the luminal surface with a glass coverslip. Mucosa was

homogenized in buffer (see below for details) and used for the determination of enzyme

activity, proteins and LA concentrations in mucosal phospholipids.

Glucose/lactose absorption

Glucose absorption and lactose digestion were determined by a combined U-13

C-

glucose/1-13

C-lactose absorption test. After an overnight fast, mice received 0.5 mg U-13

C-glucose, 5 mg 1-13

C-lactose and 5 mg naturally enriched lactose in 300 L PBS via

gastric gavage. Before and at time points 7.5, 15, 30, 45, 60, 90, 120 and 180 min. after

administration, blood samples were obtained by blood spot technique from the tail for

determination of blood concentrations of (total) glucose, 13

C6-glucose (glucose

originating from U-13

C-glucose) and 13

C-glucose (originating from 1-13

C-lactose).25

For

reasons of clarity, we will address “blood” 13

C6-glucose and 13

C-glucose as “plasma” in

the Results and Discussion sections.

Analytical methods

Lipid absorption, triene/tetraene ratio, erythrocyte fatty acid concentrations, blood

glucose and serum insulin concentrations

Lipid absorption, erythrocyte fatty acid concentrations, and triene/tetraene ratio were

determined as described previously.4,27

Blood glucose levels were measured with a

Lifescan EuroFlash glucose meter (Lifescan Benelux, Beerse, Belgium). Insulin was

measured in a solid phase two-site enzyme immunoassay in which two monoclonal

antibodies are directed against separate antigenic determinants on the insulin molecule

(Ultrasensitive Mouse Insulin kit; Mercodia, Uppsala, Sweden).

Histology and villus length along the small intestinal axis

Morphology of proximal, mid and distal small intestine was assessed by

hematoxylin/eosin staining of formalin-fixated material. Proliferating cells were detected

by staining of nuclear Ki-67 antigen. Morphometrical analysis of small intestinal samples

was performed as described by evaluation of approximately 5 vertically oriented villi per

intestinal segment of 4 to 6 animals per group. The digitized images were evaluated at

10x magnification using the calibrated image analysis system (Leica Quantimet 570 C;

CHAPTER 2

34

Leica Qwin Pro V 2.8). The epithelial surface lining was demarcated and measured as a

parametrical length, whereby 1 pixel was equal to 0.544 µm.

Glucose/lactose absorption

The analysis of 13

C6-glucose and 13

C-glucose concentrations from blood spots was

performed according to Van Dijk et al. by gas chromatography-mass spectrometry

(SSQ700, ThermoFischer B.V., Breda, The Netherlands).25

Disaccharidase activity assay in mucosal homogenates

A portion of small intestinal mucosa (proximal, mid and distal) was homogenized with

PBS buffer containing protease inhibitors (Roche, Indianapolis, USA) in order to make

4% homogenates for use in enzyme activity assay. Enzyme activity level of lactase and

sucrase were measured in freshly scraped intestinal mucosa as described previously by

Dahlqvist.28

Activity was normalized to protein levels, measured by the BCA method as

described by the manufacturer (Pierce, Rockford, IL).

Measurement of mRNA expression by real-time PCR (Taqman)

mRNA expression of lactase and sucrase isomaltase was measured in proximal, mid

and distal small intestine by real-time PCR as described previously.29

In addition, mRNA

expression levels of intestine-specific transcription factors (Cdx-2, Gata-4 and Hnf-1α)

were measured by real-time PCR in the mid part of the small intestine. PCR results were

normalized to -actin mRNA levels. The sequences of the primers and probes are listed

in Table 1.

Table 1 The PCR Primers and TaqMan Probes.

LA determination in phospholipids of intestinal mucosa

Thirty mg of intestinal mucosa was homogenized in 200 µl of 0.9% NaCl and lipids were

extracted according to Bligh and Dyer, after the addition of the fatty acid internal

standard (C17:0) and anti-oxidant (BHT).30

Lipid extracts were fractionated into

phospholipids and other lipids using TLC (20 x 20 cm, Silica gel 60 F254; Merck) with

hexane/diethyl ether/acetic acid (80:20:1, v:v:v) as running solvent. Phospholipid spots

were scraped and phospholipids were extracted by methanol/chloroform. Phospholipid

LA ratio was determined according to Muskiet et al. as described previously.24,31

EFA DEFICIENCY AND INTESTINAL FUNCTION

35

2

CH

AP

TE

R

Statistical analysis

Values represent means ± SD for the indicated number of mice per group. Using SPSS

version 12.0.2 statistical software (Chicago, IL, USA), we calculated significance of

differences with the Mann-Whitney U-test and p-values below 0.05 were considered

statistically significant.

Correlations between the linoleic concentrations in the mucosa of the mid small intestine,

and mRNA expression and enzyme activity of lactase and sucrase were determined by

means of linear regression and are expressed as non-parametric Spearman correlation

coefficient (SPSS version 14.0, Chicago, IL, USA). Differences between means were

considered significant at the level of p<0.01.

RESULTS

Body weight and food ingestion were assessed every two weeks and there were no

significant differences in basal or final body weight or in food intake between EFA-

deficient and control mice (data not shown).

Figure 1 (a) Fat absorption of total dietary fat, and of major dietary fatty acids (16:0, 18:0, 18:1ω-9 and 18:2ω-6) in EFA-deficient (white bars) and control (black bars) mice. Feces were collected after a 72h period in which the food intake was monitored by weighing food containers. Absorption was calculated by subtracting fecal excretion of these fatty acids after 72h from their dietary intake in 72h and then multiplying the result by 100. (b) Fatty acid composition of erythrocyte lipids of EFA-deficient (white bars) and control (black bars) mice. Fatty acid concentrations are in mol% of total fatty acids. Data represent means ± SD of 7 mice per group. *p<0.05 for EFA-deficient versus control mice.

Pronounced EFA deficiency of EFA-deficient mice

After eight weeks of treatment, in mice fed the EFA-deficient diet, the triene/tetraene

ratio in red blood cell membranes was strongly increased compared with the control

CHAPTER 2

36

group (0.23±0.06 versus 0.01±0.00; respectively, p<0.01). Fatty acids profile in

erythrocyte lipids is very similar to those obtained in our previous studies in mice. 32

They

revealed a severe decrease in ω-6 family of essential fatty acids, accompanied by an

increase in concentrations of non-essential fatty acids of the ω-7 and ω-9 families

(Figure 1b). Fat balance during 72 hours revealed a decreased total fat absorption in

EFA-deficient compared with control mice (81% versus 99%, respectively, p<0.01;

Figure 1a). The absorption of saturated fatty acids, palmitic (C16:0) and stearic (C18:0)

acids, was affected to a greater extent than that of the unsaturated fatty acids oleic

(C18:1ω-9) and linoleic (C18:2ω-6) acid. Together, these observations indicated that the

mice fed the EFA-deficient diet had profound EFA deficiency after 8 weeks on the

experimental diet.4

EFA deficiency in mice is not associated with alterations in intestinal morphology

Hematoxylin/eosin (data not shown) and Ki67 staining of the three segments of the small

intestine revealed no clear differences in morphology or proliferative capacity between

EFA-deficient and control mice (Figure 2a). The villus lengths were similar in EFA-

deficient and control mice, as determined by morphometrical measurements (Figure 2b).

Figure 2 (a) Ki67 staining of sections of the three parts of the small intestine (proximal, mid and distal) from EFA-deficient and control mice. (b) Morphometry of the villus length in the three segments of the small intestine (proximal, mid and distal) of EFA-deficient and control mice. Data represent means ± SD of 4-6 mice per group. No significant differences were found between the two groups.

EFA deficiency is associated with delayed glucose clearance

Basal blood glucose concentrations were similar in EFA-deficient and control mice. After

intragastric administration of the glucose/lactose bolus, glucose concentrations rapidly

EFA DEFICIENCY AND INTESTINAL FUNCTION

37

2

CH

AP

TE

R

increased in control mice, with a maximum concentration at 30 min. after administration

(Figure 3a). In EFA-deficient mice, the increase in blood glucose levels was similar to

that in control mice up to 30 min., but then continued to increase, reaching a maximum

concentration at 60 min. The glucose concentrations between 60 and 180 min. were

slightly, but significantly higher in EFA-deficient mice compared with controls (+10-15%,

p<0.05). Accordingly, the area under the curve was higher for the EFA-deficient mice

compared with controls (+15%, p<0.05, data not shown). Based on the apparently

delayed glucose clearance, we determined insulin concentrations at the end of the

experiment (at ~180 min.). In EFA-deficient mice insulin concentrations were significantly

higher than in control mice (0.55±0.10 µg/ml versus 0.35±0.02 µg/ml, respectively,

p<0.01).

Similar glucose absorption but delayed lactose digestion in EFA-deficient mice

To assess the competence of monosaccharide absorption in EFA deficiency, we

determined plasma appearance of 13

C6-glucose (Figure 3b). After the administration of

the bolus, plasma 13

C6-glucose concentration rapidly increased with a maximum at 45

min. for both groups. After 45 min., 13

C6-glucose rapidly disappeared until 120 min., after

which the rate of disappearance decreased in both EFA-deficient and control mice. Thus,

the plasma 13

C6-glucose appearance and disappearance was similar in EFA-deficient

and control mice, supporting unaffected monosaccharide absorption in the former.

In order to measure the competence of disaccharide digestion and absorption, we

determined plasma appearance of 13

C-glucose, originating from the administered 1-13

C-

lactose (Figure 3c). 13

C-glucose reached a maximum concentration in control mice at 45

min. after bolus administration. The 13

C-glucose disappeared from the blood within the

next 2 hours, with the slowest disappearance during the last hour after the bolus

administration. Plasma appearance of 13

C-glucose in EFA-deficient mice, however,

increased to a slower extent and reached its maximal concentration at approximately 60

min. after the bolus administration. In addition, the peak of 13

C-glucose absorption in

EFA-deficient mice was lower compared with control mice. Thus, the lactose uptake was

delayed in EFA-deficient compared with control mice.

Specific decrease in mRNA expression and activity of lactase in mid small

intestine of EFA-deficient mice

Measurement of the enzyme activity of lactase along the proximal-to-distal axis of the

small intestine revealed a lower activity in the mucosa of the mid part of the small

intestine of EFA-deficient mice (Figure 4a). The decreased lactase activity corresponded

with lower mRNA levels of lactase, as shown by quantitative PCR (Figure 4b). EFA

deficiency was not associated with decreased activity (3.6±1.6 versus 2.8±0.7 nmol/μg

protein in controls, NS) or mRNA expression (0.9±0.1 versus 0.7±0.1 in controls, NS) of

another disaccharidase in the mid small intestine, sucrase, indicating a distinct specificity

of EFA deficiency on lactase. We determined if reduced lactase mRNA expression levels

were regulated by the transcription factors at the transcriptional level. The mRNA

expression of transcription factors involved in regulation of the lactase mRNA

expression, namely Cdx-2, Gata-4 and Hnf-1α (Figure 4c), was not different between

EFA-deficient and control animals. This indicates that the regulation of lactase is at least

CHAPTER 2

38

not regulated at the transcriptional level of the transcription factors Cdx2, Gata-4 and

Hnf-1α.

Decreased lactase activity and mRNA expression are associated with low LA

concentrations in the mid small intestine

EFAs are involved in regulation of membrane fluidity and alterations in membrane lipid

matrix. Therefore, it has been proposed that EFAs indirectly influence normal

conformation and functioning of the proteins embedded in the membrane.14

For this

reason we tested if lactase activity in the mid segment of the small intestine correlated

with LA levels. LA concentrations were highest in the mid part of the small intestine in

control mice. Interestingly, LA concentration was significantly lower in the mid part of the

small intestinal mucosa of EFA-deficient compared to control mice (26 mol% versus 16

mol%, respectively, p<0.01) (Figure 5a).

Figure 3 (a) Total blood glucose response in EFA-deficient and control mice measured at different time points (7.5, 15, 30, 45, 60, 90, 120 and 180 min.) after the intragastric glucose/lactose bolus. (b) Plasma appearance of

13C6-glucose originating from the administered U-

13C-glucose in EFA-

deficient and control mice after an intragastric administration of glucose/lactose bolus. (c) Plasma appearance of

13C-glucose originating from the administered 1-

13C-lactose in EFA-deficient and

control mice after an intragastric administration of glucose/lactose bolus. Data represent means ± SD of 7 mice per group. *p<0.05 for EFA-deficient versus control mice.

LA concentrations in proximal and distal part were similar in both groups. In the mid

small intestine LA concentrations positively correlated with lactase activity (r=0.77,

p<0.01) and with mRNA expression of lactase (r=0.79, p<0.01) (Figure 5b and 5c,

respectively). However, decreased mRNA levels of lactase clearly indicate that the

intestinal impairment cannot exclusively be the result of alterations in membrane

EFA DEFICIENCY AND INTESTINAL FUNCTION

39

2

CH

AP

TE

R

composition and fluidity. There was no relationship, however, between LA concentrations

in the mid small intestine on the one hand and activity (r=-0.08, NS) or mRNA expression

(r=-0.03, NS) of sucrase on the other hand (Figure 5d and 5e, respectively).

DISCUSSION

Our previous studies suggested that EFA deficiency in mice affects fat absorption at the

level of the small intestinal mucosa.4 We now explored the effects of EFA deficiency in

mice on mucosal histology and physiological function of the small intestine, carbohydrate

digestion and absorption. Our data demonstrate that EFA deficiency is not only

associated with fat malabsorption, but also with impaired lactose digestion in the murine

model of EFA deficiency.

Figure 4 (a) Enzyme activity of lactase in the three segments of the small intestine of EFA-deficient and control mice. Enzyme activity is expressed as glucose production after 1 hour of incubation of intestinal mucosa with the substrate of lactase. Data represent means ± SD of 6 mice per group. **P<0.01 for EFA-deficient versus control mice. (b) mRNA expression levels of lactase gene, normalized to β-actin, in the three segments of the small intestine of EFA-deficient and control mice. Data represent means ± SD of 6 mice per group. **P<0.01 for EFA-deficient versus control mice. (c) Relative mRNA expression of the transcription factors Cdx-2, Gata-4 and Hnf-1α (involved in regulation lactase expression) in the mid part of the small intestine. Data represent means ± SD of 5-6 mice per group. No significant differences were found between EFA-deficient and control mice.

The effect of EFA deficiency is mainly on lactase mRNA transcription or stability, since

the impaired lactose digestion coincided with a ~50% reduced lactase activity and mRNA

expression in mid small intestine of EFA-deficient mice. Interestingly, the reduction in

enzyme activity during EFA deficiency was specific for lactase. In addition, intestinal

lactase activity and mRNA expression strongly correlated with mucosal linoleic acid

concentrations, which were depressed in EFA deficiency, particularly in mid intestine.

Our findings seem to be in concordance with a study of Clark et al. which showed the

a

b

CHAPTER 2

40

most marked delay in fat transport during EFA deficiency in the mid portion of the small

intestine.33

We analyzed whether regional differences in the severity of EFA deficiency could

contribute to the observations. However, the severity of EFA deficiency was similar in the

different parts of the small intestine of EFA-deficient mice (data not shown). Based on

pathophysiology we suggest 2 possibilities for this phenomenon: EFA deficiency in the

mid intestine is associated with either increased (local) oxidation of LA or with increased

turnover/export of LA. However, further studies need to be performed in order to

investigate the specific effect of EFA deficiency on fatty acid composition and lactase

activity in the mid small intestine.

As expected from previous studies, our murine model of EFA deficiency was clearly

deficient, as indicated by elevated triene/tetraene ratios in erythrocytes and fat

malabsorption. In addition, we measured the bile production by the collection of bile via

gallbladder cannulation as described previously.34

Conform previous observations, the

mice fed the EFA-deficient diet have increased bile flow, biliary bile salt and phospholipid

secretion rates (data not shown), as well as higher plasma levels of triene/tetraene ratio

(0.55±0.20 versus 0.01±0.00, p<0.01), compared with control mice.4 Although one would

expect that EFA deficiency associated fat malabsorption would lead to a lower body

weight, we did not observe this in our present or previous studies.4 Therefore, we

assume that prolongation of the EFA-deficient state would indeed be expected to result

in lower body weight.

EFA deficiency in mice did not affect morphology or proliferative capacity of the small

intestine. As far as we know, our study is the first to describe the effects of EFA

deficiency on the intestinal morphology in mice. Christon et al. have shown that low

dietary linoleic acid levels were associated with alterations in villi and crypt sizes in

rats.35

We did not observe differences in villus length between EFA-deficient and control

mice in the proximal-to-distal axis of the small intestine. These results indicate that EFA

deficiency associated malabsorption of fats and disaccharides is not associated with

morphological alterations in small intestine of mice.

To assess small intestinal function in EFA-deficient mice, we studied carbohydrate

absorption, using stable isotope methodology.25

The advantage of stable isotope

methodology is that it can easily be extrapolated to patient studies.21,22

EFA-deficient

mice had higher total blood glucose levels from 60 min. after the administration of the

glucose/lactose bolus. High total blood glucose levels could theoretically be explained by

lower blood glucose clearance (slower postprandial uptake of glucose by the peripheral

tissues), rather than by disturbed intestinal absorption. This finding is associated with

higher insulin concentrations at the end of the experiment in EFA-deficient mice, which is

not the result of an impaired glucose disposal, as this appears to be normal. This

observation is in accordance with previous studies suggesting a relationship between

EFA deficiency and insulin resistance.36

However, it is not clear what the exact reason is

for higher insulin levels in EFA-deficient mice. We cannot exclude that the increased

content of saturated fats in the EFA-deficient diet contributes to this phenomenon,

independently from EFA deficiency.37

Measurement of the absorption of 13

C6-glucose, originating from the administered U-13

C-

glucose, revealed similar appearance and disappearance of the labeled glucose in both

EFA DEFICIENCY AND INTESTINAL FUNCTION

41

2

CH

AP

TE

R

Figure 5 (a) LA (C18:2ω-6) concentration in small intestinal mucosa along the proximal-to-distal axis of EFA-deficient and control mice. Concentrations are indicated in mol% of total fatty acids. Data represent means ± SD of 6 mice per group. **P<0.01 for EFA-deficient vs. control mice. (b) Relationship between the mucosal LA concentration (mol%) and enzyme activity of lactase (nmol/µg protein/h) in the mid part of the small intestine. There is a positive correlation (r=0.77, p<0.01) between the LA concentration and lactase activity in mucosa. (c) Relationship between the mucosal LA concentration (mol%) and relative mRNA expression of lactase in the mid part of the small intestine. There is a positive correlation (r=0.79, p<0.01) between the LA concentration and lactase mRNA expression in mucosa. (d) Relationship between the mucosal LA concentration (mol%) and enzyme activity of sucrase (nmol/µg protein/h) in the mid part of the small intestine. There was no significant correlation (r=-0.08, NS) between the LA concentration and sucrase activity in mucosa. (e) Relationship between the mucosal LA concentration (mol%) and relative mRNA expression of sucrase isomaltase gene in the mid part of the small intestine. There was no significant correlation (r=-0.03, NS) between the LA concentration and mRNA expression of sucrase isomaltase.

CHAPTER 2

42

groups. This observation indicates that EFA deficiency does not affect the

monosaccharide glucose absorption in mice. The plasma appearance of 13

C-glucose

originating from lactose, however, was significantly delayed in EFA-deficient mice. The

discrepancy in the effect of EFA deficiency on glucose and lactose absorption could be

explained by the diverse intestinal fates of these carbohydrates. Unlike glucose, which is

directly transported by the glucose transporters across the brush border membrane of

the enterocyte, lactose first needs to be hydrolyzed by the enzyme lactase.7,8

In order to

investigate whether our functional results corresponded with altered lactase activity or

expression, we measured these parameters in EFA-deficient and control mice. Lactase

is a critical disaccharidase during early postnatal life and a sensitive intestinal marker for

functional changes occurring in the small intestine of the adult animal. Its activity

relatively decreases during weaning to low adult levels, thus the total lactase activity

remains the same during the adulthood.10,38

The delayed lactose digestion corresponded

with an approximate 50% reduction in both lactase activity and mRNA expression in

EFA-deficient mice. The mRNA levels of relevant transcription factors for lactase mRNA

expression were unaffected in EFA-deficient mice. Due to the unaltered mRNA levels of

the transcription factors, we conclude that the regulation of lactase by EFA deficiency is

at least not regulated at the transcriptional level of the transcription factors Cdx2, Gata-4

and HNF-1α. In order to assess the specificity of the reduced enzyme activity associated

with EFA deficiency for lactase, we measured the enzyme activity of another

disaccharidase, sucrase. Enzyme activity of sucrase was not decreased in the three

parts of the small intestine of EFA-deficient mice. These data clearly demonstrated that

EFA deficiency does not generally affect disaccharidase function and activity, but rather

specifically affects the mRNA expression and activity of lactase.

Under physiological conditions phospholipids of the small intestinal mucosa contain

considerable amounts of LA (C18:2ω-6) and its long-chain polyunsaturated fatty acid

metabolite arachidonic acid (AA, C20:4ω-6).39

During EFA deficiency LA levels are

decreased in intestinal mucosa.12,40

We observed LA deficiency in mucosal

phospholipids, particularly in the mid part of the small intestine, which strongly correlated

with reduced lactase activity and mRNA expression. It is tempting to speculate that low

LA levels in phospholipids of cellular membranes lead to structural and physiological

changes in the lipid membrane. A study in pigs suggested that EFA deficiency reduces

membrane fluidity and affects membrane protein behavior in the enterocyte

membranes.41

Theoretically, these structural changes in the cellular membrane of the

enterocytes during EFA deficiency could also be the cause of functional alterations in the

membrane anchoring of lactase. However, since not only lactase activity but also its

mRNA expression was decreased in EFA deficiency and since the lactase hydrolytic

portion of the enzyme is at a considerable distance from the membrane,42

it is likely that

altered membrane fluidity is not the (single) factor involved.

It remains unclear how EFA deficiency specifically affects lactase; several factors could

be involved. Theoretically, EFA deficiency could inhibit the differentiation of the

enterocytes, accompanied by reduced expression of lactase. 10

Lactase and sucrase

genes have different mechanisms of transcriptional regulation, what could lead to

differential transcription during EFA deficiency.43,44

Alternatively, under certain

conditions, for example during malnutrition,45

stability of lactase mRNA is more

EFA DEFICIENCY AND INTESTINAL FUNCTION

43

2

CH

AP

TE

R

profoundly decreased than that of sucrase. Our present results do not allow to

discriminate between these theoretical options.

Our present results indicate that EFA deficiency has functional consequences for small

intestinal function in mice, and it provides indirect support for the hypothesis that reduced

mucosal function is involved in fat malabsorption in EFA deficiency. EFA deficiency in

(pediatric) cholestatic patients seems to be primarily caused by fat malabsorption due to

bile deficiency. Recently, we reported that cholestasis per se does not affect

carbohydrate digestion or absorption in a rat model of short-term cholestasis.46

Our

present study indicates, however, that EFA deficiency aggravates the malabsorption of

fat, and decreases the small intestinal capacity to digest lactose. Decreased levels of LA

in the mid part of the small intestine seem to, at least partially, play a pathophysiological

role in the diminished mucosal function in EFA deficiency. Our findings imply that dietary

interventions for patients encountering EFA deficiency should accommodate the

decreased capacity to absorb fat and the reduced capacity to digest lactose.

ACKNOWLEDGEMENTS

The authors would like to thank Rick Havinga, Ingrid Martini, Juul Baller, Theo Boer,

Henk Wolters and Renze Boverhof for excellent technical assistance and helpful

suggestions.

GRANTS

This study was supported by the Dutch Digestive Foundation (MLDS).

REFERENCES

1 Socha P, Koletzko B, Swiatkowska E, Pawlovska J, Stolarczyk A, and Socha J. Acta Peadiatr 1998;

87(3): 278-283.

2 Burke PA, Ling PR, Forse RA, Lewis DW, Jenkins R, and Bistrian BR. JPEN J Parenter Enteral Nutr

2001; 25(4): 188-193.

3 Levy E, Garofalo C, Thibault L, Dionne S, Daoust L, Lepage G, and Roy CC. Am J Physiol Gastrointest

Liver Physiol 1992; 262(2): G319-G326.

4 Werner A, Minich DM, Havinga H, Bloks VW, van Goor H, Kuipers F, and Verkade HJ. Am J Physiol

Gastrointest Liver Physiol 2002; 283(4): G900-G908.

5 Minich DM, Vonk RJ, and Verkade HJ. J Lipid Res 1997; 38(9): 1709-1721.

6 Friedman HI and Nyland B. Am J Clin Nutr 1980; 33(5): 1108-1139.

7 Swallow D, Poulter M, and Hollox E. Drug Metab Dispos 2001; 29(4): 513-516

8 Holmes R. J clin Path 1971; 24(5): 10-13.

9 Hollox EJ, Poulter M, Wang Y, Krause A, and Swallow DM. Eur J Hum Genet 1999; 7(7): 791-800.

10 Rings EH, Krasinkski SD, van Beers EH, Moorman AF, Dekker J, Montgomery.R.K., Grand RJ, and

Büller H. Gastroenterology 1994; 106(5): 1223-1232.

11 Buller HA, Van Wassenaer AG, Raghavan S, Montgomery RK, Sybicki MA, and Grand RJ. Am J Physiol

Gastrointest Liver Physiol 1989; 257(4): G616-G623.

12 Korotkova M and Strandvik B. Biochim Biophys Acta 2000; 1487(2-3): 319-325.

13 Kang JX. J Membr Biol 2005; 206(2): 165-172.

14 Wahle KWJ. Biochem Soc Trans 1990; 18 (5): 775-778.

CHAPTER 2

44

15 Bruno M, Koeppe RE, II, and Andersen O. PNAS 2007; 104(23): 9638-9643.

16 Li Q, Tan L, Wang C, Li N, Li Y, Xu G, and Li J. Eur J Nutr 2006; 45(3): 144-151.

17 Duplus E, Glorian M, and Forest C. J Biol Chem 2000; 275(40): 30749-30752.

18 Schoonjans K, Staels B, and Auwerx J. J Lipid Res 1996; 37(5): 907-925.

19 Bell GI, Kayano T, Buse JB, Burant CF, Takeda J, Lin D, Fukumoto H, and Seino S. Diabetes Care 1990;

13(3): 198-208.

20 Stellaard F. Isotopes Environ Health Stud 2005; 41(3): 275-286.

21 Koletzko B, Demmelmair H, Hartl W, Kindermann A, Koletzko S, Sauerwald T, and Szitanyi P. Early Hum

Dev 1998; 53 Suppl: 77-97.

22 Demmelmair H, Sauerwald T, Koletzko B, and Richter T. Eur J Pediatr 1997; 156 Suppl 1: 70-74.

23 Wang J, Cortina G, Wu SV, Tran R, Cho J, Tsai M, Bailey T, Jamrich M, Ament M, Treem W, Hill I,

Vargas J, Gershman G, Farmer D, Reyen L, and Martin M. N Engl J Med 2006; 355(3): 270-280.

24 Muskiet FA, van Doormaal JJ, Martini IA, Wolthers BG, and van der Silk W. J Chromatogr 1983; 278(2):

231-244.

25 van Dijk T, Boer T, Havinga R, Stellaard F, Kuipers F, and Reijngoud D. Anal Biochem 2003; 322(1): 1-

13.

26 Wielinga P, Wachters-Hagedoorn R, Bouter B, van Dijk T, Stellaard F, Nieuwenhuizen A, Verkade H, and

Scheurink A. Am J Physiol Gastrointest Liver Physiol 2005; 288(6): G1144-G1149.

27 Minich D, Havinga R, Stellaard F, Vonk R, Kuipers F, and Verkade H. Am J Physiol Gastrointest Liver

Physiol 2000; 279(6): G1242-G1248.

28 Dahlqvist A. Anal Biochem 1968; 22(1): 99-107.

29 Grefhorst A, Elzinga B, Voshol P, Plosch T, Kok T, Bloks V, van der Sluijs F, Havekes L, Romijn J,

Verkade H, and Kuipers F. J Biol Chem 2002; 277(37): 34182-34190.

30 Bligh EG and Dyer WJ. Can J Biochem Physiol 1959; 37(8): 911-917.

31 Werner A, Bongers M, Bijvelds M, de Jonge H, and Verkade H. J Lipid Res 2004; 45(12): 2277-2286.

32 Werner A, Havinga R, Bos T, Bloks V, Kuipers F, and Verkade H. Am J Physiol Gastrointest Liver Physiol

2005; 288(6): G1150-G1158.

33 Clark S, Ekkers T, Singh A, Balint J, Holt P, and Rodgers J. J Lipid Res 1973; 14(5): 581-588.

34 Hulzebos C, Renfurm L, Bandsma R, Verkade H, Boer T, Boverhof R, Tanaka H, Mierau I, Sauer P,

Kuipers F, and Stellaard F. J Lipid Res 2001; 42(11): 1923-1929.

35 Christon R, Meslin JC, Thévenoux J, Linard A, Léger CL, and Delpal S. Reprod Nutr Dev 1991; 31(6):

691-701.

36 Das UN. Nutrition 2001; 17(14): 337-346.

37 Lee J, Pinnamaneni S, Eo S, Cho I, Pyo J, Kim C, Sinclair A, Febbraio M, and Watt M. J Appl Physiol

2006; 100(5): 1467-1474.

38 Rings EH, de Boer PA, Moorman AF, van Beers EH, Dekker J, Montgomery RK, Grand RJ, and Buller

HA. Gastroenterology 1992; 103(4): 1154-1161.

39 Hjelte L, Melin T, Nilsson A, and Strandvik B. Am J Physiol Gastrointest Liver Physiol 1990; 259(1):

G116-G124.

40 Enser M and Bartley W. Biochemical Journal 1962; 85(3): 607-614.

41 Christon R, Even V, Daveloose D, Leger C, and Viret J. Biochimica et Biophysica Acta (BBA) -

Biomembranes 1989; 980(1): 77-84.

42 Wacker H, Keller P, Falchetto R, Legler G, and Semenza G. J Biol Chem 1992; 267(26): 18744-18752.

43 van Beers EH, Ai RH, Rings EH, Einerhand AW, Dekker J, and Büller HA. Biochem J 1995; 308(Pt 3):

769-775.

44 Tanaka T, Suzuki A, Kuranuki S, Mochizuki K, Suruga K, Takase S, and Goda T. Life Sciences; 83(3-

4):122-127.

45 Nichols BL, Dudley MA, Nichols VN, Putman M, Avery SE, Fraley JK, Quaroni A, Shiner M, and Carrazza

FR. Gastroenterology 1997; 112(3): 742-751.

EFA DEFICIENCY AND INTESTINAL FUNCTION

45

2

CH

AP

TE

R

46 Los EL, Wolters H, Stellaard F, Kuipers F, Verkade H, and Rings E. Am J Physiol Gastrointest Liver

Physiol 2007; 293(3):G615-22.

CHAPTER 3

ESSENTIAL FATTY ACID DEFICIENCY IN MICE IS ASSOCIATED

WITH CHOLESTEROL MALABSORPTION AND INCREASED

JEJUNAL LIPID SYNTHESIS

S. Lukovac1, M.Y.M. van der Wulp

1,2, V.W. Bloks

1, M.V. Boekschoten

2,3, J. Dekker

3, A.K.

Groen1, E.H.H.M. Rings

1, H.J. Verkade

1

(1) Pediatric Gastroenterology, Department of Pediatrics, Beatrix Children’s Hospital,

Groningen University Institute for Drug Exploration (GUIDE), Center for Liver, Digestive

and Metabolic Diseases, University of Groningen, University Medical Center Groningen,

Groningen, The Netherlands

(2) Nutrigenomics Consortium, Top Institute Food and Nutrition, Wageningen, The

Netherlands.

(3) Division of Human Nutrition, Wageningen University, Wageningen, The Netherlands.

Manuscript in preparation

CHAPTER 3

48

ABSTRACT

Essential fatty acid (EFA) deficiency in mice is associated with reduced small intestinal

function, as demonstrated by fat malabsorption, impaired lactose digestion and altered

small intestinal feedback regulation of bile salt synthesis. The effects of EFA deficiency

on cholesterol metabolism in the small intestine remained unknown. Here, we studied

the effects of EFA deficiency on jejunal cholesterol and fatty acid metabolism in a mouse

model of EFA deficiency.

EFA deficiency was induced in mice by feeding an EFA-deficient diet for 8 weeks. EFA-

deficient mice excreted 57% more cholesterol via the feces compared with control mice

(p<0.05). A well-known surrogate marker for cholesterol absorption (plasma plant

sterols/cholesterol ratio) was significantly reduced in EFA-deficient mice (-60%, p<0.05),

accompanied by reduced jejunal mRNA expression of the apical sterol uptake

transporter Npc1l1 (-70%, p<0.05). Plasma concentrations of major plant sterols derived

from dietary sources were significantly lower in EFA-deficient mice (p<0.05). EFA

deficiency did not affect total cholesterol concentrations in jejunal mucosa. Triglyceride

and oleic acid concentrations were significantly increased in jejunum of EFA-deficient

mice and lipid staining revealed lipid droplet accumulation in EFA-deficient jejunal

epithelium. Transcriptional analysis of jejunum of EFA-deficient mice revealed significant

induction of Srebp1 and Srebp2 and its target genes involved in fatty acid and

cholesterol synthesis. An unexpected observation from the transcriptional analysis of

jejunal segments of EFA-deficient mice was induced gene expression of genes of the

proteasome complex, suggestive for proteasome inhibition.

Our data show that EFA deficiency is associated with cholesterol malabsorption and

subsequent increased jejunal lipid synthesis, which might serve as a compensatory

mechanism for cholesterol and fatty acid malabsorption. Our observations suggest that

during EFA deficiency in mice, jejunal proteasome degradation pathway is shut down in

order to maintain the lipid homeostatic response in jejunal epithelium.

EFA DEFICIENCY AND JEJUNAL LIPID METABOLISM

49

3

CH

AP

TE

R

INTRODUCTION

Essential fatty acid (EFA) deficiency is associated with several functional abnormalities

in broad range of tissues and organs in both humans and experimental animals.1,2,3,4

Studies in EFA-deficient mice and rats demonstrated several morphological and

functional abnormalities in the small intestine, although some effects were species

specific.5,6,7,8,9

In mice, we have shown that EFA deficiency is associated with reduced

fat absorption, impaired lactose digestion and altered regulation of the enterohepatic

circulation of bile salts.6,7,10

All together, these data strongly suggest that EFA deficiency

negatively affects small intestinal function, and more specifically at the intracellular

level.6 Very little is known about the effects of EFA deficiency on the cholesterol

absorption, and small intestinal lipid metabolism in general. Christon et al. demonstrated

that there was no effect of low linoleic diet on cholesterol content in enterocyte brush

border membrane in rats.9 However, in the epidermal tissue, increased cholesterol

synthesis in EFA-deficient rats was demonstrated by Proksch et al.2 In both rats and

mice, EFA deficiency leads to development of hepatic steatosis, as indicated by elevated

hepatic and reduced plasma triglyceride concentrations.4,11

Werner et al. previously

excluded reduced VLDL-TG secretion from the liver in EFA-deficient mice as the cause

of hepatic steatosis.4 Moreover, Pparα and Pparα-target genes (Acc1, Cpt1a) were

upregulated in livers of EFA-deficient mice, probably due to induced de novo synthesis of

non-essential fatty acids.4 Therefore, hepatic triglyceride accumulation in EFA-deficient

mice was most likely due to the increased lipogenic activity, increased uptake of

circulating lipids, or a combination of both.

The aim of our study was to determine the effects of EFA deficiency on small intestinal

function regarding cholesterol absorption. We hypothesized that impaired small intestinal

function during EFA deficiency might lead to reduced absorption of cholesterol.

Theoretically, increased lipid synthesis as shown in the liver and epidermis during EFA

deficiency might exist in the small intestinal epithelium. Although the whole small

intestine is capable of cholesterol absorption, the main site of absorption is the

jejunum.12

Furthermore, regarding absorption of other nutrients, studies in EFA-deficient

rats and mice revealed that delay in transport of fatty acids and lactose was most severe

in the jejunal part of the small intestine.7,13

Therefore, in the present study we focused

our attention on the effects of EFA deficiency on jejunum.

First, we demonstrated the physiological consequences of EFA deficiency on cholesterol

metabolism by determining the fecal cholesterol excretion and sterol absorption in EFA-

deficient mice. In addition to cholesterol metabolism, we also analyzed triglyceride and

fatty acid levels in jejunum of EFA-deficient mice. Cholesterol and fatty acid metabolism

are closely associated metabolic pathways and are known to have a cross talk mediated

mainly by the sterol-regulatory element-binding proteins (SREBPs).14,15,16

In order to

characterize the effects of the EFA deficiency on the metabolic pathways in jejunum, we

analyzed the transcriptional response to the EFA deficiency in mice by microarray

analysis.

Our data clearly show that EFA deficiency in mice leads to malabsorption of cholesterol,

accompanied by an induction of the transcriptional cascade involved in cholesterol and

fatty acid synthesis. We suggest that the reduced proteasome activity during EFA

CHAPTER 3

50

deficiency is responsible for the induction of lipogenic gene expression via activation of

the transcription factor SREBPs.

The increased cholesterol and fatty acid synthesis suggest a compensatory mechanism

for reduced cholesterol and fatty acid absorption in EFA-deficient mice.

MATERIAL AND METHODS

Mice and diet

FVB (free virus breed) male mice of eight weeks old were purchased from Harlan (Horst,

the Netherlands) and were housed in a light- and temperature-controlled facility. Tap

water and food were allowed ad libitum. At the age of eight weeks, two groups of mice

were switched to either control (#4141.07) or EFA-deficient (#4141.08) diet for eight

weeks, which were both custom synthesized by Arie Blok BV (Woerden, the

Netherlands). Fatty acid composition of the diets is indicated in Table 1. The

experimental protocol was approved by the Ethics Committee for Animal Experiments,

Faculty of Medical Sciences, University of Groningen, Netherlands.

Table 1 Composition of the major fatty acids in the experimental diets. Concentrations are indicated in mol% of total fatty acid concentrations determined by gas chromatography analysis.

Induction of EFA deficiency, bile cannulations and sample preparations

EFA deficiency was induced in mice by means of an EFA-deficient diet for eight weeks,

as described earlier.6 Eight weeks of diet were sufficient to induce EFA deficiency as

determined by the (biochemical) marker for EFA deficiency, triene/tetraene ratio, in

erythrocytes and plasma. At the end of the dietary period, food intake was measured and

feces was collected during a 72h period. Subsequently, bile was cannulated for 30

minutes as previously described.10

At the end of the experiment, mice were anesthetized

and sacrificed through cardiac puncture. The small intestine was rinsed with phosphate-

buffered saline (PBS) and mid part, corresponding to jejunum, was sliced and

immediately frozen in liquid nitrogen. Subsequently, small intestinal and hepatic tissue

was stored at -80°C for further analysis.

Cholesterol balance

Cholesterol balance was determined by calculating the dietary intake and hepatobiliary

secretion of cholesterol, along with fecal output of neutral sterols per day per 100 g of

body weight, as previously described by van der Velde et al.17

Dietary cholesterol intake

EFA DEFICIENCY AND JEJUNAL LIPID METABOLISM

51

3

CH

AP

TE

R

and fecal neutral sterol excretion were analyzed during the 72 hours of feces collection

and weighing of the food pellets in this period.

Analytical methods

Analysis of biliary, dietary, fecal and plasma sterols

Hepatobiliary secretion of cholesterol was analyzed by lipid extraction from bile,

according to the method of Bligh and Dyer, and subsequent cholesterol measurement

according to Gamble et al.18,19

Food pellets and fecal samples were grind and 50 mg

was prepared for neutral sterol analysis by gas chromatography as described

previously.20

Neutral sterol profile and concentrations in plasma were analyzed by gas

chromatography mass spectrometry as described previously.21

Determination of lipids in total jejunal mucosa homogenate

Thirty milligram of intestinal mucosa was homogenized in 200 μl of 0.9% NaCl, and lipids

were extracted according to Bligh and Dyer.18

Triglyceride and cholesterol

concentrations were determined by means of commercially available kits (Roche

Diagnostics, Mannheim, Germany; DiaSys Diagnostic Systems, Holzheim, Germany).

Determination of fatty acids in total jejunal mucosa

Thirty milligram of intestinal mucosa was homogenized in 200 μl of 0.9% NaCl.

Subsequently, ten microliter of the homogenate was used for fatty acid determination

according to of Muskiet et al., after the addition of the internal standard (C17:0) and

antioxidant butylated hydroxytoluene.22

Total homogenate was used for fatty acid profile

and quantification by gas chromatography.

Histology

Jejunal lipids were examined on frozen jejunal sections after Oil Red O staining for

neutral lipids by standard procedures.

RNA isolation and measurement of RNA expression levels by microarray analysis

For the microarray analysis, the Affymetrix microarray platform was used. After RNA

isolation from jejunal tissue with TRIzol reagent, RNA was used individually and further

purified using RNeasy MinElute micro columns (Qiagen, Venlo, the Netherlands). RNA

integrity was checked on an Agilent 2100 bioanalyzer (Agilent Technologies,

Amsterdam, the Netherlands) using 6000 Nano Chips according to the manufacturer’s

instructions. RNA was judged as suitable for array hybridization only if samples exhibited

intact bands corresponding to the 18 and 28S ribosomal RNA subunits, and displayed no

chromosomal peaks or RNA degradation products (RNA Integrity Number.7.0-8.5). Five

hundred nanograms of RNA were used for one cycle cRNA synthesis (Affymetrix, Santa

Clara, CA). Control and EFA-deficient samples (n=6) were hybridized, washed and

scanned on the Affymetrix Gene chip mouse 1.0 ST arrays, according to standard

Affymetrix protocols. Scans of the Affymetrix arrays were processed using packages

from the Bioconductor project.23

Quality control of microarray data (using simpleaffy and

affyplm packages), normalization, differential expression analysis and Gene Set

Enrichment Analysis were performed through the Management and Analysis Database

CHAPTER 3

52

for MicroArray eXperiments (MADMAX) analysis pipeline, (Wageningen, the

Netherlands). Expression levels of probe sets were calculated using regular

normalization strategies: VSN in jejunum followed by identification of differentially

expressed probe sets using Limma.24

Comparison was made between treated (EFA-

deficient mice) and untreated (control mice) groups (Limma package, applying linear

models and moderated t-statistics that implement empirical Bayes regularization of

standard errors.25

False discovery rate (FDR) of 1 % (p <0.01) was used as a threshold

for significance of differential expression. The limma t-test values of differential

expression between groups where used as the input for the PreRanked scoring method

within the Gene Set Enrichment Analysis (GSEA).26

Normalized enrichment scores

(NES) of significantly enriched pathways and the corresponding FDR p-values are

available upon request. Identification of overrepresented functional categories among

responsive genes and their grouping into functionally related clusters (Biological

Processes:BP-4) was performed using DAVID Functional Annotation Clustering tool.27

All microarray data reported in the manuscript is described in accordance with MIAME

guidelines.28

Quantitative PCR

Total RNA was prepared from mouse intestinal (jejunum) or hepatic tissue using TRIzol

reagent (Invitrogen, Breda, The Netherlands). Subsequently, cDNA synthesis and

quantitative PCR analysis were performed as described by Grefhorst et al.29

PCR results

were normalized to RNA expression of the housekeeping gene 18S. Primer and probe

sequences for the Q-PCR analysis have been published (www.LabPediatricsRug.nl:

Realtime PCR Primers & Probes Database), except for Ehhadh (GeneID 74147,

NM_023737; forward primer: GCCTTT CTGTGCACCAATACC; reverse primer:

GAAGAAGTGGGTGCCAATCAC; probe: CATTGCTTC TTCCACAGATCGCCC).

Statistical analysis

Using SPSS version 16 statistical software (Chicago, IL, USA), we calculated

significance of differences between EFA-deficient and control (FVB) mice with the Mann-

Whitney U-test and p-values below 0.05 were considered statistically significant. All data

represent mean values ± SD. Statistical analysis for the microarray analysis is described

in the section above (RNA isolation and measurement of RNA expression levels by

microarray analysis).

RESULTS

EFA-deficient mice showed increased biomarker of EFA deficiency (triene:tetraene ratio)

in erythrocytes and jejunal tissue, which exceeded the threshold value for EFA deficiency

(>0.2; data not shown). In agreement with previous observations, EFA-deficient mice

showed reduced fat absorption without significant changes in body weight or food intake

(data not shown) after eight weeks of EFA-deficient diet.6

EFA deficiency in mice is associated with cholesterol malabsorption

In order to determine the physiological relevance of EFA deficiency on jejunal cholesterol

EFA DEFICIENCY AND JEJUNAL LIPID METABOLISM

53

3

CH

AP

TE

R

absorption, we measured cholesterol excretion in feces of EFA-deficient mice. Neutral

sterol analysis in fecal samples of EFA-deficient mice after 72h of collection revealed

significantly higher concentrations of cholesterol and dihydroxycholesterol compared with

control mice (Figure 1a, 1b). In addition, we analyzed the plasma sterol concentrations to

determine whether EFA deficiency is associated with reduced concentrations of plant

sterols (phytosterols) in plasma, which can only be derived from the dietary sources, thus

from the intestinal uptake. Cholesterol measured in plasma, on the other hand, can, in

addition to intestinal uptake, be derived from the periphery by means of the reversed

cholesterol transport.30

Our plant sterol analysis in plasma show significantly lower

concentrations of two major plant sterols campesterol and β-sitosterol, and of

cholestanol (Figure 1c).31

Figure 1 Cholesterol concentrations in feces and plasma of EFA-deficient (white bars) and control (black bars) mice. (a) Fecal cholesterol and (b) dihydroxycholesterol concentrations were determined by gas chromatography subsequent to 72h feces collection. (c) Plasma sterol profile was determined by gas chromatography-mass spectrometry; from different sterol concentrations in plasma, (d) marker for cholesterol absorption (as indicated by the plant sterol/cholesterol ratio) and (e) markers for cholesterol synthesis (as indicated by the lathosterol/cholesterol and desmosterol/cholesterol ratio) were calculated. (f)Total cholesterol concentrations were determined in jejunal mucosa. Values are means ±SD for n=5-6. *p<0.05 is the significant difference between the two groups.

Concentrations of lanosterol and desmosterol in plasma were significantly increased in

EFA-deficient mice compared with control mice. Furthermore, the marker of cholesterol

absorption, the ratio of plant sterols (campesterol+β-sitosterol) to cholesterol was

significantly decreased in EFA-deficient mice (Figure 1d).32

Most commonly used marker

for cholesterol synthesis, lathosterol/cholesterol ratio, was not different between EFA-

deficient and control mice (Figure 1e). However, another marker for cholesterol synthesis

(desmosterol/cholesterol ratio) was significantly higher in EFA-deficient mice (Figure 1e).

Total cholesterol concentration in jejunal mucosa was not affected by EFA deficiency in

mice (Figure 1f).

CHAPTER 3

54

Cholesterol balance in EFA-deficient mice

In order to determine the cholesterol intake and excretion, we measured the food intake

and collected feces during 72h. In parallel, bile was collected by means of bile

cannulation to determine the hepatobiliary secretion of cholesterol. Total cholesterol

balance shows that there is no significant difference in cholesterol intake between EFA-

deficient and control mice (Table 2).

Table 2 Cholesterol balance (dietary ingestion, hepatobiliary secretion, fecal output; μmol/day/100g body weight). Data represent means ± SD of 4-6 mice per group. *p<0.05 is the significant difference between EFA-deficient and control mice after eight weeks of EFA-deficient or control diet, respectively.

Biliary secretion of cholesterol was significantly higher in EFA-deficient mice, in

agreement with previous studies in EFA-deficient mice (Table 2).6,10

However, fecal

excretion of cholesterol is extremely increased in EFA-deficient mice to such an extent

that it is even higher than the sum of cholesterol derived from the intake and the

cholesterol of hepatobiliary origin (Table 2). This suggests that the alternative route of

cholesterol excretion from the intestine, transintestinal cholesterol efflux (TICE)-route,

might be responsible for the missing cholesterol excreted in EFA-deficient mice (Table

2).33

However, our indirect measurement of TICE did not reach the significant difference

between EFA-deficient and control mice due to the large variation within the groups

(p=0.07) (Table 2).

EFA deficiency in mice leads to increased oleic acid and triglyceride

concentrations in jejunal mucosa

Molar concentrations of polyunsaturated fatty acids (PUFA) were significantly reduced,

while molar concentrations of monounsaturated fatty acids (MUFA) were increased in

jejunal mucosa of EFA-deficient mice (Figure 2a). EFA deficiency had no effect on the

molar concentrations of the saturated acids (SAFA) in mouse jejunum (Figure 2a).

Decreased molar concentrations of PUFA in jejunum of EFA-deficient mice were mainly

due to lower molar concentrations of linoleic acid (18:2ω-6) and its metabolite

arachidonic acid (20:4ω-6) (Figure 2b). Fatty acid profile revealed significantly increased

molar concentrations of 18:1ω9 and 18:1ω7, while the molar concentration of the

precursor of the synthesis of these two fatty acids, stearic acid (18:0) was decreased in

EFA-deficient mice (Figure 2b). Decrease in molar concentrations of palmitate (18:0)

were accompanied by increased molar concentration of oleic acid (18:1ω-9), leading to

EFA DEFICIENCY AND JEJUNAL LIPID METABOLISM

55

3

CH

AP

TE

R

Figure 2 Fatty acids and triglycerides in jejunal mucosa of EFA-deficient (white bars) and control (black bars) mice. (a) Saturated, monounsaturated and polyunsaturated fatty acid concentrations, as well as (b) the concentrations of the individual fatty acids are indicated as molar percentages of total fatty acid concentrations. (c) Desaturation index is determined by the oleic acid/stearic acid ratio. (d) Total triglyceride concentrations were determined in the freshly scraped jejunal mucosa. (e) Elevated fat accumulation in EFA-deficient jejunum is indicated by the representative picture of the Oil Red O staining (40x magnification). Values are means ±SD for n=4-6. *p<0.05 is the significant difference between the two groups.

CHAPTER 3

56

significantly increased desaturation index (18:1ω-9/18:0) in both plasma and jejunum

(Figure 2c) of EFA-deficient mice. Triglyceride concentrations in jejunal mucosa were

increased by more than six fold (Figure 2d), consistent with accumulation of lipid droplets

in jejunal tissue of EFA-deficient mice, as demonstrated by the Oil Red O staining

(Figure 2e).

EFA deficiency in mice is associated with inhibition of the proteasomal activity

Overall microarray analysis revealed in total 1388 of the jejunal genes which were

differentially regulated during EFA deficiency in mice, when the false discovery rate

(FDR) of <1% was used as a threshold for significance of differential expression. From

these genes, 857 were upregulated and 531 were down regulated in jejunal tissue during

EFA deficiency. DAVID annotation of the microarray analysis pointed the attention on the

proteasome (Table 3, bold fond). The importance of the proteasomal pathway in this

model was confirmed by GSEA (NES=2.66), Metacore GeneGo (P<1e-7) and ErmineJ

(raw score 3.38, P<1.67e-11) (data not shown, available upon request).26,34

Over 20

members of the genes of the proteasome complex were significantly upregulated in

jejunum of EFA-deficient mice (Supplementary Table; p<0.05). Within the top of most

affected pathways in EFA-deficient mice were also biosynthesis of steroids and

polyunsaturated fatty acids, as well as the fatty acid biosynthesis in general (Table 3).

Table 3 Cellular pathways enriched during EFA deficiency in mouse jejunum – microarray analysis (KEGG-Pathway). False discovery rate (FDR) <1%. Proteasome pathway, with the largest number of genes affected, is indicated in bold font.

EFA deficiency in mice increased transcriptional activity of jejunal genes involved

in lipid metabolism

Based on the Gene Ontology categorization, EFA deficiency mainly affects metabolic

processes (Table 4). In top 10 of the processes most significantly enriched during EFA

deficiency in mice, lipid biosynthetic processes and processes involved in steroid and

fatty acid metabolism were listed (Table 4, indicated in bold font). This was in agreement

with the enrichment data of the most affected pathways (Table 3).

EFA deficiency in mice is associated with enhanced transcription of genes

involved in jejunal cholesterol synthesis

Microarray analysis of jejunal tissue revealed that many genes involved in sterol

metabolic processes were enriched in EFA-deficient mice (Table 4). More detailed

analysis revealed that the expression of several genes important for cholesterol

EFA DEFICIENCY AND JEJUNAL LIPID METABOLISM

57

3

CH

AP

TE

R

absorption and efflux was affected by EFA deficiency (Table 5). The mRNA expression

of Niemen-Pick C1-like 1 protein (Npc1l1), critical player in the absorption of intestinal

sterol expressed at the apical surface of enterocytes, was decreased in jejunum of EFA-

deficient mice (Table 5).12

The mRNA expression of the basolateral ATP-binding

cassette (ABC) protein Abca1, which mediates HDL secretion from enterocytes, was

decreased in jejunal tissue of EFA-deficient mice (Table 5).35

The mRNA expression of

two other cholesterol transporters scavenger receptor class B, member 1 (Sr-BI; Scarb1)

and low density lipoprotein receptor (Ldlr), both implied in basolateral cholesterol uptake

into the enterocytes, was increased upon EFA deficiency in mouse jejunum (Table

5).35,36

Table 4 Biological processes enriched during EFA deficiency in mouse jejunum – microarray analysis (GOTERM_BP_4). False discovery rate (FDR) <1%. Processes involved in steroid and fatty acid metabolism are indicated in bold font.

The differences in mRNA expression of the jejunal cholesterol transporters between

EFA-deficient and control mice detected by the microarray analysis were validated by the

quantitative PCR (Q-PCR) analysis. Q-PCR analysis demonstrates decreased mRNA of

Npc1l1 and Ldlr, while no significant difference in the mRNA expression of the Abca1

between EFA-deficient and control mice was detected (Figure 3a). There was no

significant difference in mRNA expression of Abcg5/8 in jejunum of EFA-deficient mice

compared with control mice, as demonstrated by both the microarray (Table 5) and Q-

PCR analysis (data not shown).

Microarray analysis revealed that 34 jejunal genes involved in steroid metabolic process

were enriched in EFA-deficient mice (Table 4). Detailed analysis of the microarray data

shows that the mRNA expression of the rate controlling enzyme in the cholesterol

synthesis, 3-hydroxy-3-methyl-glutaryl-CoA reductase (Hmgcr), was significantly

increased in EFA-deficient mice (Table 5). In agreement, the mRNA expression of other

relevant genes in the cholesterol biosynthesis pathway, soluble 3-hydroxy-3-

methylglutaryl-Coenzyme A synthase 1 (Hmgcs1, soluble) and 3-hydroxy-3-

CHAPTER 3

58

methylglutaryl-Coenzyme A synthase 2 (Hmgcs2, mitochondrial), were increased in

EFA-deficient mouse jejunum (Table 5). When cellular cholesterol concentrations

decrease, SREBP is released from the ER or nuclear membrane and transported to the

nucleus where it binds to the sterol regulatory element on the HMGCR-gene. As the

result, transcription of the HMGCR gene is initiated.36,37

Microarray analysis

demonstrated induction of the Srebp2 mRNA expression in jejunal tissue of EFA-

deficient mice (Table 5). Quantitative PCR analysis confirmed the increased mRNA

expression of Hmgcr, Hmgcs1, Hmgcs2 and Srebp2 (Figure 3b). Moreover, Q-PCR

analysis of the hepatic Hmgcr revealed significantly higher expression in EFA-deficient

mice compared with control mice (Figure 3c). The mRNA expression of the liver X

receptors Lxrα (Nr1h3) and Lxrβ (Nr1h2), other important transcriptional regulators of

cholesterol metabolism, were also significantly higher in jejunum of EFA-deficient mice

(Table 5).38

Table 5 Fold changes of genes involved in intestinal cholesterol transport and synthesis – microarray analysis. Fold changes are indicated by the arrows in the direction of up- (↑) or downregulation (↓) of gene expression. False discovery rate (FDR) <1%. mRNA expression of genes indicated with an asterix (*) has been confirmed by the Q-PCR analysis.

EFA deficiency in mice is associated with enhanced transcription of genes

involved in jejunal lipogenesis and beta oxidation

Previous studies in mouse model of EFA deficiency revealed hepatic steatosis

characterized by triglyceride accumulation in hepatic tissue.4 In parallel to these

physiological changes, EFA deficiency in mice was accompanied by increased hepatic

mRNA expression of genes involved in lipogenesis and beta oxidation.4 Microarray

analysis in present study demonstrated that 41 genes involved in fatty acid metabolic

processes were enriched in jejunum of EFA deficient mice. In EFA-deficient mice, the

mRNA expression of almost all lipogenic genes was increased in jejunum (Table 6).

More specifically, the mRNA expression of Scd1, Acc2 and Fasn was significantly

increased in jejunum of EFA-deficient mice, as determined by both the microarray and

Q-PCR analysis. (Table 6; Figure 4a). Scd1 gene encodes the enzyme responsible for

the conversion of stearic acid (18:0) into its metabolite oleic acid (18:1ω9). Thus, the

EFA DEFICIENCY AND JEJUNAL LIPID METABOLISM

59

3

CH

AP

TE

R

increased mRNA expression of Scd1 in EFA-deficient mouse jejunum was in agreement

with the increased desaturation index (marker for the activity of SCD1) in plasma of EFA-

deficient mice (Figure 2c). In addition, the jejunal mRNA expression of Srebp1,

transcription factor regulating fatty acid synthesis, was also induced by EFA deficiency

(Table 6). The mRNA expression of Gpam, involved in triglyceride synthesis, was more

than three times higher in EFA-deficient mice compared with control mice (Table 6,

Figure 4a).39

This is in agreement with increased triglyceride concentrations in mouse

jejunal mucosa (Figure 2d). Moreover, jejunal mRNA expression of most genes involved

in fatty acid oxidation was significantly increased in EFA-deficient mice, as demonstrated

by the microarray analysis (Table 6). Significantly increased mRNA expression of two

relevant beta oxidation genes, Ehhadh and Cpt1a, was confirmed by means of Q-PCR

analysis (Figure 4b).

Figure 3 Jejunal expression of genes involved in (a) cholesterol transport and (b) synthesis in EFA- deficient (white bars) versus control (black bars) mice. (c) Hepatic mRNA expression of Hmgcr in EFA-deficient (white bars) and control (black bars) mice. Data represent mRNA expression relative to the RNA expression of the housekeeping gene 18S. Values are means ±SD for n=6. *p<0.05 is the significant difference between the two groups.

DISCUSSION

In the present study, we addressed the effects of EFA deficiency on jejunal lipid

metabolism in a mouse model of EFA deficiency. Previous studies in this model

suggested impaired small intestinal function during EFA deficiency. However, the effects

of EFA deficiency on jejunal cholesterol metabolism remained unclear. Our data indicate

CHAPTER 3

60

that EFA deficiency in mice leads to reduced Npc1l1 mRNA expression in jejunal

epithelium and subsequent cholesterol malabsorption. Transcriptional analysis of jejunal

tissue of EFA-deficient mice revealed increased transcription of proteasomal genes,

accompanied by increased Srebp2 mRNA expression. Increased jejunal Srebp2

expression in EFA-deficient mice was accompanied by increased transcription of its

target genes involved in cholesterol synthesis. In addition to alterations in cholesterol

metabolism, EFA deficiency in mice was associated with increased triglyceride and oleic

acid concentrations in jejunal mucosa, supported by increased transcription of lipogenic

genes. We propose that during EFA deficiency jejunal tissue induces compensatory

mechanisms to correct for reduced lipid absorption. By shutting down the proteasomal

complex, Srebp2 activation is increased leading to induction of target genes involved in

cholesterol and fatty acid metabolism. These data underscore previous findings

suggesting that EFA deficiency leads to altered small intestinal function. Previous studies

in mouse model of EFA deficiency revealed negative effects on the small intestine,

Table 6 Fold changes of genes involved in intestinal lipogenesis and β-oxidation – microarray analysis. Fold changes are indicated by the arrows in the direction of up- (↑) or downregulation (↓) of gene expression. False discovery rate (FDR) <1%. mRNA expression of genes indicated with an asterix (*) has been confirmed by the Q-PCR analysis.

including reduced fat absorption and impaired lactose digestion.6,7,10

Therefore, we

hypothesized that EFA deficiency additionally might lead to reduced cholesterol

absorption. In order to test this hypothesis we measured fecal cholesterol excretion in

EFA-deficient mice, which appeared to be 50% higher in EFA-deficient mice compared

with control mice. In agreement, the marker for cholesterol absorption was decreased by

59% in EFA-deficient mice. Besides the reduced absorption, increased dietary intake or

increased biliary secretion of cholesterol to the intestinal lumen might be the cause of the

EFA DEFICIENCY AND JEJUNAL LIPID METABOLISM

61

3

CH

AP

TE

R

increased fecal excretion of cholesterol. Therefore, we determined these parameters by

measuring both the dietary intake and the hepatobiliary secretion of cholesterol.

Although the dietary intake was similar between EFA-deficient and control mice, EFA

deficiency significantly increased hepatobiliary secretion of cholesterol, in agreement

with previous data in EFA-deficient mice.6 However, the increased hepatobiliary

secretion of cholesterol was not sufficient to compensate for the difference in fecal

excretion of cholesterol between EFA-deficient and control mice. This suggested the

presence of an alternative route for cholesterol secretion in EFA-deficient mice, besides

the hepatobiliary route. Theoretically, increased cholesterol concentration in feces of

EFA-deficient mice could be due to the increased shedding of the small intestinal

enterocytes. However, this is rather unlikely since previous studies in EFA-deficient mice

showed by means of Ki67 staining in jejunum that there were no significant differences in

proliferative capacity of jejunal enterocytes between EFA-deficient and control mice.7

Recently, several studies in mice described that small intestine plays a significant role in

removal of cholesterol from the body and that the capacity of the transintestinal

cholesterol efflux (TICE) is sufficient to account for the missing cholesterol in the balance

studies.40,41

It is possible that increased TICE in EFA-deficient mice is responsible for

Figure 4 Jejunal expression of genes involved in (a) fatty acid synthesis and (b) beta oxidation in EFA deficient (white bars) versus control (black bars) mice. Data represent mRNA expression relative to the RNA expression of 18S. Values are means ±SD for n=6. *p<0.05 is the significant difference between the two groups.

increased cholesterol excretion. Interestingly, previous study in mice demonstrated that

increased mRNA expression of Sr-BI (Scarb1) correlated with TICE; in EFA-deficient

mice we show that Sr-BI expression is increased, along with increased TICE.42

However,

direct measurement of TICE by perfusion of isolated jejunal segments of the small

intestine in Sr-BI deficient mice revealed a significant, twofold increase in

TICE

compared with wild-type mice.43

The underlying mechanism of this discrepancy

remained unclear. Future studies with direct TICE measurements by perfusion of the

intestinal segments should reveal whether EFA deficiency increases TICE and in which

part of the small intestine. Jejunal Abca1 mRNA expression was reduced in EFA-

deficient mice, probably as the result of reduced cholesterol absorption and to maintain

CHAPTER 3

62

sufficient cholesterol levels within the enterocytes. However, negative regulation of

Abca1 by unsaturated fatty acids has previously been described.44

Therefore, reduced

expression of Abca1 which might also be the result of increased jejunal oleic acid

concentration. In addition to cholesterol malabsorption in EFA-deficient mice, our data

clearly demonstrate that EFA deficiency in parallel leads to the induction of mRNA

expression of majority of genes involved in cholesterol and fatty acid metabolism. In

addition, microarray analysis demonstrated increased gene expression of the

proteasomal pathway in jejunum during EFA deficiency. Increased gene expression of

proteasomal complex leads to negative feedback regulation of the proteasomal activity,

thus inhibition of the proteasome degradation. Theoretically, this could lead to prolonged

expression and activity of certain cellular proteins and transcription factors. Previous

studies revealed increased Srebp activity upon inhibition of the ubiquitin-proteasome

pathway.45

Furthermore, SREBPs are shown to be possible substrates for the ubiquitin-

proteasome system, which in turn controls the expression of SREBP-responsive

genes.46

Surprisingly, mRNA expression of Srepb was demonstrated to be increased in

jejunal tissue of EFA-deficient mice, leading to increased mRNA expression of its target

genes involved in both cholesterol and fatty acid metabolism. Although we did not

measure direct intestinal cholesterol synthesis, it has been previously shown that mRNA

levels of HMGCR strongly correlate with changes in cholesterol synthesis in several

tissues.47

We additionally demonstrated increased expression of HMGCR in the liver of EFA-

deficient mice, suggesting that increased cholesterol synthesis also takes place in the

hepatic tissue during EFA deficiency. Moreover, Proksch et al. demonstrated increased

cholesterol synthesis in epidermal tissue of EFA-deficient mice. Compensatory increase

of cholesterol synthesis in jejunum could explain maintained jejunal cholesterol

concentrations in EFA-deficient mice, despite the malabsorption of cholesterol. One of

the markers of the cholesterol synthesis, namely plasma ratio of lathosterol to

cholesterol, was not affected by EFA deficiency. However, the ratio of two other

cholesterol precursors, desmosterol and lanosterol, to cholesterol was increased in

plasma of EFA-deficient mice. Plasma markers determined do not discriminate between

the intestinal cholesterol synthesis and that in other (i.e. hepatic) tissues, and are

therefore not suitable as specific jejunal markers of cholesterol synthesis.

In agreement with induced jejunal expression of target genes of Srebp involved in fatty

acid metabolism, we demonstrate that physiological consequence of increased lipogenic

genes during EFA deficiency leads to increased oleic acid and triglyceride

concentrations. Lack of EFA in the tissues is known to be demarcated by an increased

synthesis of the non-essential fatty acid oleate. Increased oleic acid synthesis in jejunal

mucosa of EFA-deficient mice is most likely originating from its precursor palmitic acid.

Preliminary study of Hamel suggested that oleic acid itself might inhibit the proteasomal

pathway.48

Whether EFA deficiency directly inhibits the proteasomal complex gene

expression, or indirectly via increased oleic acid synthesis, remains to be elucidated.

Although jejunal tissue seems to compensate for the reduced absorption of lipids, lipids

accumulation occurs within the jejunum. Previous studies in EFA-deficient mice revealed

smaller chylomicrons in these mice, which could explain reduced jejunal capacity to

secrete the synthesized lipids.49

EFA DEFICIENCY AND JEJUNAL LIPID METABOLISM

63

3

CH

AP

TE

R

Increased fatty acid synthesis and oxidation during EFA deficiency are not specific for

the jejunum. Werner et al. demonstrated earlier that similar pathways and genes are

induced by EFA deficiency in mouse hepatic tissue.4 In both liver and jejunum

triglyceride concentrations are severely increased during EFA deficiency in parallel to

increased mRNA expression of Gpam, essential for triglyceride synthesis. Previous

studies in mice fed high fat diet revealed increased triglyceride accumulation in the liver

which was attributed to the high dietary content of the saturated fatty acids.50

EFA-

deficient diet contains more saturated fatty acids palmitate and stearate, thus the effects

seen in jejunal lipid metabolism could be, at least in part, due to increased dietary intake

of saturated fat. However, since fat balance studies revealed overall malabsorption of

both saturated and unsaturated fatty acids, triglyceride accumulation cannot exclusively

be due to the increased intake of saturated fatty acids.

Altogether, our data show that EFA deficiency leads to cholesterol malabsorption.

Metabolic pathways involved in jejunal lipid synthesis are activated in order to

compensate for the malabsorption of both cholesterol and fatty acids. However, we

suggest that elevated synthesis is not effective in improving the absorption of fatty acids

and cholesterol, since most of the lipids accumulate within the enterocytes. In order to

improve the nutritional status of patients with EFA deficiency, future studies should focus

on the mechanism underlying lipid accumulation during EFA deficiency and improvement

of the intestinal function.

ACKNOWLEDGEMENTS

The authors thank Ingrid Martini, Renze Boverhof, Juul Baller and Gertrud Kortman for

their excellent technical assistance. Mechteld Grootte-Bromhaar is acknowledged for her

excellent technical assistance with microarray analyses. Part of this study was supported

by the Dutch Digestive Foundation and Top Institute Food and Nutrition (TIFN;

Wageningen, the Netherlands).

GRANTS

This study was supported by the Dutch Digestive Foundation (MLDS).

REFERENCES

1 Smit E, Muskiet F and Boersma ER. Prostaglandins, Leukotrienes and Essential Fatty Acids 2004;

71(4):241-250.

2 Proksch E, Feingold KR and Elias PM. J Invest Dermatol 1992; 99(2):216-220.

3 Das UN. Essential fatty acids: biochemistry, physiology and pathology. Biotechnology Journal 2006;

1(2):420-439.

4 Werner A, Havinga R, Bos T, Bloks V, Kuipers F and Verkade HJ. Am J Physiol Gastrointest Liver

Physiol 2005; 288(6):G1150-G1158.

5 Levy E, Garofalo C, Thibault L, Dionne S, Daoust L, Lepage G and Roy CC. Am J Physiol Gastrointest

Liver Physiol 1992; 262(2):G319-G326.

6 Werner A, Minich DM, Havinga H, Bloks VW, van Goor H, Kuipers F and Verkade HJ. Am J Physiol

Gastrointest Liver Physiol 2002; 283(4):G900-G908.

CHAPTER 3

64

7 Lukovac S, Los EL, Stellaard F, Rings EH and Verkade HJ. Am J Physiol Gastrointest Liver Physiol 2008;

295(3):G605-G613.

8 Daveloose D, Linard A, Arfi T, Viret J and Christon R. Biochim Biophys Acta 1993; 1166(2-3):229-237.

9 Christon R, Meslin JC, Thévenoux J, Linard A, Léger CL and Delpal S. Reprod Nutr Dev 1991; 31(6):691-

701.

10 Lukovac S, Los EL, Stellaard F, Rings E.H. and Verkade HJ. Am J Physiol Gastrointest Liver Physiol

2009; 297(3):G520-G531

11 Sinclair AJ, Collins FD. Br J Nutr 1970; 24(4):971-982.

12 Lammert F, Wang DQ. Gastroenterology 2005; 129(2):718-734.

13 Clark S, Ekkers T, Singh A, Balint J, Holt P and Rodgers J. J Lipid Res 1973; 14(5):581-588.

14 Tobin KA, Steineger HH, Alberti S, Spydevold O, Auwerx J, Gustafsson JA and Nebb HI. Mol Endocrinol

2000; 14(5):741-752.

15 Brown MS, Goldstein JL. Cell 1997; 89(3):331-340.

16 Bennett MK, Lopez JM, Sanchez HB and Osborne TF. J Biol Chem 1995; 270(43):25578-25583.

17 van der Velde AE, Vrins CL, van den Oever K, Kunne C, Oude Elferink RP, Kuipers F and Groen AK.

Gastroenterology 2007; 133(3):967-975.

18 Bligh EG, Dyer WJ. Can J Biochem Physiol 1959; 37(8):911-917.

19 Gamble W, Vaughan M, Kruth HS and Avigan J. J Lipid Res 1978; 19(8):1068-1070.

20 Gerhardt KO, Gehrke CW, Rogers IT, Flynn MA and Hentges DJ. J Chromatogr 1977; 135(2):341-349.

21 Plösch T, Bloks VW, Terasawa Y, Berdy S, Siegler K, van der Sluijs F, Kema IP, Groen AK, Shan B,

Kuipers F and Schwartz M. Gastroenterology 2004; 126(1):290-300.

22 Muskiet FA, van Doormaal JJ, Martini IA, Wolthers BG and van der Silk W. J Chromatogr 1983;

278(2):231-244.

23 Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dutoit S, Ellis B, Gautier L, Ge Y, Gentry J,

Hornik K, Hothorn T, Huber W, Iacus S, Irizarry R, Leisch F, Li C, Maechler M, Rossini AJ, Sawitzki G,

Smith C, Smyth G, Tierney L, Yang JY and Zhang J. Genome Biol 2004; 5(10):R80.

24 Irizarry RA, Wu Z and Jaffee HA. Bioinformatics 2006; 22(7):789-794.

25 Smyth GK. Stat Appl Genet Mol Biol 2004; 3(article3)

26 Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL,

Golub TR, Lander ES and Mesirov JP. Proceedings of the National Academy of Sciences of the United

States of America 2005; 102(43):15545-15550.

27 Dennis G, Sherman B, Hosack D, Yang J, Gao W, Lane HC and Lempicki R. Genome Biology 2003;

4(5):3.

28 Brazma A, Hingamp P, Quackenbush J, Sherlock G, Spellman P, Stoeckert C, Aach J, Ansorge W, Ball

CA, Causton HC, Gaasterland T, Glenisson P, Holstege FC, Kim IF, Markowitz V, Matese JC, Parkinson

H, Robinson A, Sarkans U, Stewart J, Taylor R, Vilo J and Vingron M. Nat Genet 2001; 29(4):365-371.

29 Grefhorst A, Elzinga B, Voshol P, Plosch T, Kok T, Bloks V, van der Sluijs F, Havekes L, Romijn J,

Verkade H and Kuipers F. J Biol Chem 2002; 277(37):34182-34190.

30 Groen AK, Oude Elferink RP, Verkade HJ and Kuipers F. Ann Med 2004; 36(2):135-45.

31 Wilund KR, Yu L, Xu F, Vega GL, Grundy SM, Cohen JC and Hobbs HH. Arterioscler Thromb Vasc Biol

2004; 24(12):2326-2332.

32 Nissinen MJ, Gylling H and Miettinnen TA. British Journal of Nutrition 2008; 99(02):370-378.

33 van der Velde AE, Vrins CL, van der Oever K, Kunne C, Oude Elferink RP, Kuipers K and Groen AK.

Gastroenterology 2007; 133(3):967-975.

34 Lee H, Braynen W, Keshav K and Pavlidis P. BMC Bioinformatics 2005; 6(1):269.

35 Hui D, Howles P. Seminars in Cell & Developmental Biology 2005; 16(2):183-192.

36 Brown MS, Goldstein JL. Science 1986; 232(4746):34-47.

37 Brown MS, Goldstein JL. Proceedings of the National Academy of Sciences of the United States of

America 1999; 96(20):11041-11048.

38 Zelcer N, Tontonoz P. J Clin Invest 2006; 116(3):607-614.

EFA DEFICIENCY AND JEJUNAL LIPID METABOLISM

65

3

CH

AP

TE

R

39 Gimeno RE, Cao J. Thematic Review Series: Glycerolipids. Journal of Lipid Research 2008; 49(10):2079-

2088.

40 Kruit JK, Plosch T, Havinga R, Boverhof R, Groot PHE, Groen AK and Kuipers F. Gastroenterology 2005;

128(1):147-156.

41 van der Veen JN, Kruit JK, Havinga R, Baller JFW, Chimini G, Lestavel S, Staels B, Groot PHE, Groen

AK and Kuipers F. Journal of Lipid Research 2005; 46(3):526-534.

42 van der Velde AE, Vrins CLJ, van den Oever K, Seemann I, Oude Elferink RPJ, van Eck M, Kuipers F

and Groen AK. AJP - Gastrointestinal and Liver Physiology 2008; 295(1):G203-G208.

43 van der Velde AE, Vrins CL, van den Oever K, Seemann I, Oude Elferink RP, van Eck M, Kuipers F and

Groen AK. Am J Physiol Gastrointest Liver Physiol 2008; 295(1):G203-G208.

44 Uehara Y, Miura Si, von Eckardstein A, Abe S, Fujii A, Matsuo Y, Rust S, Lorkowski S, Assmann G,

Yamada T and Saku K. Atherosclerosis 2007; 191(1):11-21.

45 Kuhn DJ, Burns AC, Kazi A and Ping Dou Q. Biochimica et Biophysica Acta (BBA) - Molecular and Cell

Biology of Lipids 2004; 1682(1-3):1-10.

46 Hirano Y, Yoshida M, Shimizu M and Sato R. J Biol Chem 2001; 276(39):36431-36437.

47 Jurevics H, Hostettler J, Barrett C, Morell P and Toews AD. Journal of Lipid Research 2000; 41(7):1048-

1054.

48 Hamel FG. Metabolism 2009; 58(8):1047-1049.

49 Werner A, Havinga R, Perton F, Kuipers F and Verkade HJ. Am J Physiol Gastrointest Liver Physiol

2006; 290(6):G1177-G1185.

50 Oosterveer MH, van Dijk TH, Tietge UJ, Boer T, Havinga R, Stellaard F, Groen AK, Kuipers F and

Reijngoud DJ. PLoS One 2009; 4(6):e6066.

CHAPTER 3

66

Supplementary table DAVID software analysis of the KEGG-pathways enriched in EFA-deficient mouse jejunum reveal proteasome complex genes as the most significantly enriched cellular process upon EFA deficiency. Fold change of the expression of these genes is indicated for false discovery rate (FDR) <1%.

CHAPTER 4

EFFECTS OF ESSENTIAL FATTY ACID DEFICIENCY ON

ENTEROHEPATIC CIRCULATION OF BILE SALTS IN MICE

S. Lukovac1, E.L. Los

1,2, F. Stellaard

1, E.H.H.M. Rings

1, H.J. Verkade

1

(1) Pediatric Gastroenterology, Department of Pediatrics, Beatrix Children’s Hospital,

Groningen University Institute for Drug Exploration (GUIDE), Center for Liver, Digestive

and Metabolic Diseases, University of Groningen, University Medical Center Groningen,

Groningen, The Netherlands

(2) Current address: Department of Cell Physiology, Section Osmoregulation, Radboud

University Nijmegen Medical Centre, Nijmegen, The Netherlands.

Am J Physiol Gastrointest Liver Physiol 2009; 297: G520–G531

CHAPTER 4

70

ABSTRACT

Essential fatty acid (EFA) deficiency in mice has been associated with increased bile

production, which is mainly determined by the enterohepatic circulation (EHC) of bile

salts. To establish the mechanism underlying the increased bile production, we

characterized in detail the EHC of bile salts in EFA-deficient mice using stable isotope

technique, without interrupting the normal EHC. Farnesoid X receptor (FXR) has been

proposed as an important regulator of bile salt synthesis and homeostasis. In Fxr-/-

mice

we additionally investigated to what extent alterations in bile production during EFA

deficiency were FXR-dependent. Furthermore, we tested in differentiating Caco-2 cells

the effects of EFA-deficiency on expression of FXR-target genes relevant for feedback

regulation of bile salt synthesis.

EFA deficiency enhanced bile flow and biliary bile salt secretion were associated with

elevated bile salt pool size and synthesis rate (+146% and +42%, respectively, p<0.05),

despite increased ileal bile salt reabsorption (+228%, p<0.05). Cyp7a1 mRNA

expression was unaffected in EFA-deficient mice. However, ileal mRNA expression of

Fgf15 (inhibitor of bile salt synthesis) was significantly reduced, in agreement with absent

inhibition of the hepatic bile salt synthesis. Bile flow and biliary secretion were enhanced

to the same extent in EFA-deficient wild type and Fxr-/-

mice, indicating contribution of

other factors besides FXR in regulation of EHC during EFA deficiency. In vitro

experiments show reduced induction of mRNA expression of relevant genes upon

chenodeoxycholic acid (CDCA) and GW4064 stimulation in EFA-deficient Caco-2 cells.

In conclusion, our data indicate that EFA deficiency is associated with interrupted

negative feedback of bile salt synthesis, possibly due to reduced ileal Fgf15 expression.

EFA DEFICIENCY AND BILE SALT METABOLISM

71

4

CH

AP

TE

R

INTRODUCTION

Essential fatty acid (EFA) deficiency is a frequent condition in patients with cholestasis or

cystic fibrosis1,2,3

and has various effects on bile production and absorption of dietary

fat.4,5,6,7

Bile salts are essential for bile production, secretory processes and efficient

intestinal absorption of dietary fat. In our lab we developed a mouse model to study the

effects of EFA deficiency in vivo.7,8

Previous studies in this mouse model have shown

that EFA deficiency-associated fat malabsorption is not caused by impaired bile

formation, like it has been implied earlier in a rat model for EFA deficiency.6 In EFA-

deficient mice an increase in bile flow and biliary secretion was observed.7 The

physiological importance and the underlying mechanism of elevated bile flow and biliary

secretion of bile salts during EFA deficiency in mice remains unclear. Bile flow and biliary

secretion of bile salts are mainly influenced by the circulation of bile salts from the liver to

the intestine, and their reabsorption back to the liver via the portal circulation. This

enterohepatic circulation (EHC) of bile salts involves many hepatic and intestinal

transporters responsible for the uptake and excretion of bile salts and results in efficient

preservation of bile salts within the body.9 Under physiological conditions, per

enterohepatic cycle less than 5 percent of the total amount of bile salts present in the

body, i.e. the bile salt pool, is lost via the feces. Under steady state conditions this

fraction of bile salts lost is compensated by hepatic bile salt synthesis.9,10

Although

increased bile flow and biliary secretion have been reported in EFA-deficient mice, it

remained unclear how EFA deficiency in mice affects the EHC of bile salts. To address

this question, in the current study we measured different steps of the EHC in vivo in

EFA-deficient mice by stable isotope dilution technique. We compared the outcomes of

the measured bile salt synthesis and pool size by the stable isotope dilution technique

with the classical determination of these parameters, namely by determining the

expression of the gene encoding cholesterol 7α-hydroxylase (Cyp7a1), which is the rate

limiting enzyme in bile salt synthesis.11,12

Shortly, by stable isotope dilution technique,

different parameters of the EHC (synthesis, pool size, fractional turnover rate, ileal

reabsorption and cycling time) were determined in vivo by means of the intravenous

injection of a stably labeled bile salt and subsequent determination of plasma enrichment

of the label. This method, known as the stable isotope dilution technique, has been

developed and validated previously in our lab to measure bile salt kinetics in vivo without

interrupting the normal EHC.9,10

We have chosen to inject the stably labeled cholate

(2H4-cholate), as this is the primary and most abundant bile salt in humans and rodents;

total bile acid pool consists of 30% to 50% and 50% to 80% of cholate (CA) in humans

and rodents, respectively.10

To confirm our in vivo findings, we additionally measured

expression of relevant intestinal genes in the EHC of bile salt synthesis in small intestinal

model for EFA deficiency (post confluent Caco-2 cells) upon stimulation with

chenodeoxycholic acid (CDCA) and GW4064 compound, both potent agonists of FXR.

Elucidating the mechanism behind the elevated bile flow during EFA deficiency, might

help to understanding and treat fat malabsorption during EFA deficiency. Our study

demonstrates that increased secretion of bile salts during EFA deficiency in mice is

associated with enhanced bile salts synthesis, despite increased reabsorption of bile

salts in the intestine. Our results clearly show that increased bile production during EFA

CHAPTER 4

72

deficiency cannot be contributed exclusively to FXR, but that other factors must

contribute as well. In vivo and in vitro data show impaired transcriptional regulation of

genes involved in intestinal regulation of bile salt homeostasis under EFA-deficient

conditions. This suggests an impaired intestinal feedback mechanism of bile salt

synthesis during EFA deficiency.

MATERIAL AND METHODS

Mice and housing

Male wild type mice (~25 g) on a free virus breed (FVB) background were obtained from

Harlan (Horst, the Netherlands) and were housed in a light- and temperature-controlled

facility. Tap water and food were allowed ad libitum. In a separate experiment, where we

aimed to determine whether the effects of EFA deficiency on the EHC of bile salts were

FXR-dependent, we used male homozygous (Fxr-/-

) and wild type (Fxr+/+

) mice on mixed

(C57BL/6Jx129/OlaHsd) background of 25-40 g. These mice were generated previously

by Deltagen, Inc. (Redwood City, CA) and bred at the animal facility of the University of

Groningen.9 Food intake and fecal excretion were monitored during a 72h period at the

end of the experiment. For clarity reasons, mice fed the EFA-deficient or control diet (on

FVB background) will be mentioned as EFA-deficient or control mice. Fxr+/+

and Fxr-/-

mice fed the EFA-deficient or control diet will also be entitled as EFA-deficient or control,

respectively, with the genotype indicated in addition.

The experimental protocol was approved by the Ethics Committee for Animal

Experiments, Faculty of Medical Sciences, University of Groningen, Netherlands.

Experimental diets

As in our previous studies, we used high-fat (humanized) EFA-deficient (#4141.08) and

EFA-sufficient (control, #4141.07) diets (16 wt% and 34 energy% fat), which were

custom synthesized by Arie Blok BV (Woerden, the Netherlands.7,8

Essentially,

unsaturated fatty acids in EFA-sufficient diet were replaced by saturated fatty acid in

EFA-deficient diet; EFA-deficient diet was particularly reduced in linoleic acid (essential

fatty acid) concentration. In detail, EFA-deficient diet contained 64 mol% palmitic acid

(C16:0), 18 mol% stearic acid (C18:0), 13 mol% oleic acid (C18:1ω-9) and 5 mol%

linoleic acid (C18:2ω-6). Isocaloric EFA-sufficient diet contained 36 mol% C16:0, 5 mol%

C18:0, 31 mol% C18:1ω-9 and 29 mol% C18:2ω-6. Fatty acid contents of the diets were

analyzed by extracting, hydrolyzing and methylating total dietary fatty acids. Subsequent

separation and quantification of fatty acid methyl esters was performed by gas

chromatography as described previously.7,13

Material for stable isotope dilution technique and cell culture experiments

Consistent with previous studies with stable isotope dilution technique, we administered 2H4-cholate ([2,2,4,4,-

2H]-cholate) of 98% isotopic purity, which was purchased from

Isotec (Miamisburg, OH). Cholyglycine hydrolase from Clostridium perfringens was

obtained from Sigma Chemicals (St. Louis, MO) and pentafluorobenzylbromide (PFB)

was purchased from Fluka Chemie (Buchs, Neu-Ulm, Switzerland). For the in vitro

EFA DEFICIENCY AND BILE SALT METABOLISM

73

4

CH

AP

TE

R

studies chenodeoxycholic acid was purchased from Calbiochem (CDCA; San Diego, CA,

USA) and GW4064 from Tocris Bioscience (Ellisville, MO, USA).

Methods

Induction of EFA deficiency

Mice were fed the EFA-deficient or control diet for 8 weeks, consistent with previous

studies in EFA-deficient mice.7,8,14

After 8 weeks of EFA-deficient or control diet, mice

underwent a 72h fat balance test, bile cannulation and stable isotope dilution test (details

see below). Afterwards, the mice were anesthetized and sacrificed through cardiac

puncture. The marker of EFA deficiency, triene/tetraene ratio (C20:3ω-9/C20:4ω-6), was

determined in erythrocyte lipids as described previously.7,8

Protocol for induction of EFA

deficiency in Caco-2 cells is described in section “Stimulation of differentiating EFA-

deficient and control Caco-2 cells to CDCA and GW4064”.

Fat absorption

Absorption of major dietary fatty acids was assessed by measuring food intake and

collecting feces for 72h. Net amount of fat absorbed was calculated by subtracting the

fecal excretion of the major fatty acids (stearate, palmitate, oleate and linoleic acid) from

the amount of fatty acids ingested, determined by gas chromatography.7

Stable isotope dilution

The stable isotope dilution test was performed as previously described, slightly

modified.10

Three days prior to the end of the experiment, after the fat absorption test

was completed, 400 μg of 2H4-cholate in a solution of 0.5% NaHCO3 in PBS was slowly

injected intravenously under isoflurane anesthesia. At different time points (12, 24, 36,

48 and 60h) blood samples of 75 μl were collected by orbital punction under isoflurane

anesthesia to determine the isotope enrichment in plasma. Blood, collected in

microhematocrit tubes containing heparin, was centrifuged (4,000 rpm for 10 min) and

plasma was stored at -20°C until further analysis. Samples used for baseline isotope

abundance measurements were obtained by orbital puncture from a separate group of

mice.

Bile collection

After the mice were anesthesized with Hypnorm/Diazepam mixture, the bile duct was

cannulated during 30 minutes and bile flow was determined gravimetrically (1 g/ml).

During the cannulation, body temperature was maintained by placing the mice in a

humidified incubator (37°C).9

Sample collection

The small intestine was excised, flushed with ice-cold PBS and the last part (terminal

ileum) was harvested for gene (mRNA) expression. The livers were excised and

weighed. Subsequently, small pieces were cut out for mRNA and biochemical analysis.

Prior to storage at -80°C, tissues were snap-frozen in liquid nitrogen.

CHAPTER 4

74

Cell culture

The human colon carcinoma cell line Caco-2 was obtained from the American Type

Tissue Culture Collection (Manassas, VA, USA). Before the experiment, the cells were

maintained in DMEM (Gibco BRL, USA) supplemented with 10% FBS, 100 IU/ml

penicillin, 100 mg/ml streptomycin, 1% nonessential amino acids and 0.25% human

transferrin in a humidified 5% CO2 atmosphere under standard conditions.

Stimulation of differentiating EFA-deficient and control Caco-2 cells to CDCA and

GW4064

For the experiment, cells (between passage 21 and 40) were seeded at 0.5 × 107

cells/well. Cells were made EFA-deficient according to the adapted protocol of Spalinger

et al.15

Shortly, medium was replaced by DMEM supplemented with dialyzed FBS

(control cells) or with delipidated FBS (EFA-deficient cells) one day after seeding.

Delipidation of FBS was performed by means of di-iso-propylether and 80 ml butan-1-ol

extraction. Seven days after complete confluence, the cells were exposed to serum-free

DMEM containing chenodeoxycholic acid (CDCA; 50 or 250 µM), GW4064 (1 µM) or

vehicle for 24 hours. Afterwards, cells were harvested for fatty acid profile determination

and quantitative PCR. All experiments were performed at least in triplicate.

Analytical methods

Biliary bile salts and lipids

Bile salt concentration in bile was determined by an enzymatic fluorimetric assay.16

Biliary phospholipids and cholesterol were determined as described by Kuipers et al.17

Gas chromatography

Fatty acids in erythrocytes, food and feces of the mice, and fatty acids in Caco-2 cells

were analyzed by extracting, hydrolyzing and methylating total dietary fatty acids as

described by Muskiet et al.13

Subsequent separation and quantification of fatty acid

methyl esters was performed by gas chromatography.7 Bile salt composition of bile

samples was determined by capillary gas chromatography.9

Preparation of plasma samples for isotope analysis and gas-liquid chromatography-

electron capture negative chemical ionization mass spectrometry (GLC-MS)

Plasma and bile samples were prepared for GLC-MS analysis on a Finnigan SSQ7000

Quadrupole GC-MS machine as described previously by Stellaard et al.18

Isotope

dilution technique has been described in detail by Hulzebos et al.10

Shortly, enrichment

of 2H4-cholate in plasma was determined as the increase of the M4-/M0-cholate relative to

baseline measurements and is expressed as the natural logarithm of atom percent

excess (ln APE). From the decay curve of ln APE (calculated by linear regression

analysis), daily fractional turnover rate (FTR; equals the slope of the regression line) and

pool size ((administered amount of label x isotopic purity x 100) / eintercept of the y-axis of the ln

APE curve) of cholate were calculated. Subsequently, cholate synthesis rate was calculated

by multiplying pool size and FTR. In addition, the amount of cholate reabsorbed per day,

the cycling time and biliary secretion rate of cholate were calculated as described

previously.9,10

Cholate was the most abundant bile acid in the bile salt pool of both EFA-

EFA DEFICIENCY AND BILE SALT METABOLISM

75

4

CH

AP

TE

R

Table 1 Primer and probe sequences used for the quantitative PCR. Genes indicated in capital font are of human origin used for Q-PCR analysis in Caco-2 cells. Remaining primers and probes were used for quantitative PCR analysis in mouse tissues.

CHAPTER 4

76

deficient (~58%) and control mice (~58%). Therefore, we assumed that the parameters

calculated for cholate were representative for the complete bile salt pool.

Measurement of mRNA expression by quantitative PCR (Taqman)

mRNA expression of hepatic genes involved in bile salt synthesis, hepatic transporter

genes (for bile salts, cholesterol and phospholipids), ileal genes implicated in

enterohepatic circulation of bile salts and FXR-target genes in Caco-2 cells were

determined by quantitative PCR as described previously.19

PCR results in the liver were

normalized to the RNA expression of the housekeeping gene 18s, in ileum to the

housekeeping gene -actin and in Caco-2 cells to GAPDH. The sequences of the

primers and probes are listed in Table 1.

Heuman index

The Heuman index is a numeric representation of hydrophilic-hydrophobic balance,

corresponding with the ability of bile acids to solubilize dietary fats. For determination of

the hydrophobicity of the bile salt pool, we calculated the Heuman index after

quantification of major biliary bile salts by gas chromatography.20

Statistical analysis

Using SPSS version 12.0.2 statistical software (Chicago, IL, USA), we calculated

significance of differences between EFA-deficient and control (FVB) mice with the Mann-

Whitney U-test.

For the experiment with the Fxr-/-

mice and their wild type littermates on either EFA-

deficient or control diet, statistical analysis was assessed by One-Way ANOVA test

followed by a post hoc Bonferroni correction.

For in vitro experiments, data were statistically analyzed using Student’s two-tailed t test.

For all experiments, p-values below 0.05 were considered statistically significant.

Table 2 Animal characteristics of FVB mice fed EFA-deficient or control diet for 8 weeks. Values are represented as means ± SD (n=6 mice per group). *p<0.05 is the significant difference between EFA-deficient and control FVB mice.

RESULTS

Body weight and food intake were not different between EFA-deficient and control FVB

mice (Table 2). Consistent to our previous findings in EFA-deficient mice, liver weight

was higher in EFA-deficient mice compared with controls (Table 2), most probably due to

EFA DEFICIENCY AND BILE SALT METABOLISM

77

4

CH

AP

TE

R

fat accumulation as indicated by a trend towards an increased concentration of

triglycerides (117.4±57.4 nmol.mg-1

protein in control mice versus 176.5±57.7 nmol.mg-1

protein in EFA-deficient mice, NS) and significantly increased cholesterol (57.6±10.1

nmol.mg-1

protein in control mice versus 83.1±8.6 nmol.mg-1

protein in EFA-deficient

mice, p<0.05) in the livers of EFA-deficient mice (data not shown).7 After 8 weeks of

EFA-deficient diet-feeding, the triene/tetraene ratio (biochemical marker of EFA

deficiency) was increased in erythrocytes of EFA-deficient mice (0.23±0.06 versus

0.01±0.00 in control mice, p<0.05) (Table 2). The induction of EFA deficiency in mice

decreased fat absorption by approximately 20% (Table 2).

In our second experiment we assessed the possible contribution of FXR to EFA-deficient

phenotype of the EHC in Fxr-/-

mice and their wild type littermates with or without EFA

deficiency. During the last decade, the nuclear farnesoid X receptor (FXR) has been

identified as an important regulator of the bile salt metabolism.21,22

Table 3 Animal characteristics of Fxr-/-

mice and their wild type littermates (C57BL/6Jx129/OlaHsd background) on EFA-deficient or control diet for 8 weeks. *p<0.05 EFA-deficient mice versus control with the same genotype, #p<0.05 Fxr

-/- versus Fxr

+/+ on the same diet. Values are represented as

means ± SD (n=5-7 mice per group).

Bile salts are the natural ligands of FXR and can activate the FXR in the ileum. In ileum,

FXR activated by the bile salts induces release of fibroblast growth factor 15 (Fgf15;

homologue to human FGF19), which is transported to the liver to inhibit the bile salts

synthesis.23

Fgf15 has been characterized as important component of the gut-liver

signaling pathway that regulates the bile salt synthesis. Similar to EFA-deficient and

control FVB mice, after 8 weeks of diet there was no difference in body weight or food

intake between EFA-deficient Fxr-/-

or Fxr+/+

mice compared with mice of the same

genotype on control diet (Table 3). Liver weight was higher in EFA-deficient Fxr-/-

mice

compared with the same mice on control diet (Table 3). In Fxr+/+

mice this difference did

not reach the significant value (Table 3). Upon EFA-deficient diet-feeding, both Fxr-/-

and

Fxr+/+

mice became EFA-deficient compared with mice of the same genotype on control

diet, as indicated by increased triene/tetraene ratio (0.15±0.07 in EFA-deficient Fxr+/+

mice and 0.17±0.08 in EFA-deficient Fxr-/-

mice versus 0.01±0.00 in control mice of both

genotypes, p<0.05) and by fat malabsorption (Table 3). Interestingly, fat malabsorption

was less profound in EFA-deficient Fxr-/-

mice than in their EFA-deficient wild type

littermates (-22% in EFA-deficient Fxr-/-

mice compared with Fxr-/-

mice on control diet,

and -29% in EFA-deficient Fxr+/+

mice compared with Fxr+/+

mice on control diet, p<0.05).

CHAPTER 4

78

EFA deficiency is associated with increased bile flow, but has no effect on biliary

bile salt composition

We determined bile flow and biliary secretion rates by bile cannulations. In accordance

with previous findings EFA deficiency significantly increased bile flow (+78% in EFA-

deficient versus control mice, p<0.05) (Table 4).7

Table 4 Bile flow and biliary secretion rates in FVB mice fed EFA-deficient or control diet for 8 weeks. Values are represented as means ± SD (n=6 mice per group). *p<0.05 is the significant difference between EFA-deficient and control FVB mice.

Biliary secretion rates of bile salts and phospholipids were two- to threefold higher in

EFA-deficient compared with the control FVB mice (each p<0.05) (Table 4). To

determine if EFA deficiency results in altered composition of bile salt pool, we

determined biliary bile salt composition and calculated its hydrophobicity, using the

Heuman index number.20

EFA deficiency did not result in major changes in bile salt

composition (Figure 1) or in the Heuman index (-0.20 ± -0.06 in EFA-deficient versus -

0.21 ± -0.03 in control mice, NS).

In order to investigate if enhanced bile flow in EFA-deficient mice was dependent of

biliary bile salt secretion, we plotted the bile flow (y-axis) against the biliary bile salt

output (x-axis) (data not shown).

Figure 1 Biliary bile salt composition in EFA-deficient (white bars) and control (black bars) mice.

More than 90% of all bile salts are represented and values are expressed as percentage of total bile

salts. Data represent means of 5-6 mice per group. Data are means ± SD of n=6 mice per group.

*p<0.05 is the significant difference between the two groups.

EFA DEFICIENCY AND BILE SALT METABOLISM

79

4

CH

AP

TE

R

Figure 2 (a) Decay curve of the intravenously injected

2H4-cholate in EFA-deficient and control

mice. Enrichment of the administrated 2H4-cholate was determined in plasma during 60 hours after

the administration of the label. From the curve (b) pool size (y-intercept), (c) synthesis, (d) FTR (slope), (e) reabsorption and (f) cycling time of cholate were determined for individual mice. Data are means ± SD of 6 mice per group. *p<0.05 is the significant difference between EFA-deficient (white bars) and control (black bars) mice.

CHAPTER 4

80

There is a classical linear relationship between these two parameters, indicating that

EFA deficiency in mice does not affect the normal bile formation. Increased bile flow

during EFA deficiency is most likely caused by the increased bile salt output.

Increased bile salt secretion during EFA deficiency corresponds with enhanced

synthesis, pool size and reabsorption of bile salts

To evaluate the effects of EFA deficiency on different parameters of bile salt

homeostasis without interrupting the normal EHC, we performed the stable isotope

dilution technique by intravenous injection of 2H4-cholate. Decay curve of plasma

enrichment of 2H4-cholate over time indicates different kinetics of cholate in EFA-

deficient mice compared with control mice (Figure 2a). EFA deficiency was associated

with higher pool size (Figure 2b: 40 ± 1 μmol.100g-1

in EFA-deficient mice versus 16 ± 7

μmol.100g-1

of in control mice, p<0.05), higher synthesis rate (Figure 2c: 11 ± 2

μmol.100g-1

day-1

in EFA-deficient mice versus 8 ± 2 μmol.100g-1

day-1

in control mice,

p<0.05) and lower fractional turnover rate (Figure 2d: 0.6 ± 0.2 per day versus 0.3 ± 0.01

per day in control mice, p<0.05) of cholate. Reabsorption of cholate in ileum was

enhanced in EFA-deficient mice (Figure 2e: 512 ± 275 μmol.100g-1

day-1

versus 156 ± 29

μmol.100g-1

day-1

in control mice, p<0.05), while the cycling time of cholate was not

affected by EFA deficiency (Figure 2f: 2.3 ± 1.1 hours in EFA-deficient versus 2.4 ± 1.1

hours in control mice, NS).

mRNA expression of genes involved in EHC in EFA-deficient mice

By means of quantitative PCR we determined the mRNA expression of hepatic and ileal

genes implicated to be important in bile salt homeostasis or bile flow (Figure 3). EFA

deficiency did not have a major effect on the mRNA expression of the gene encoding the

rate-limiting enzyme in bile salt synthesis, namely Cyp7a1 (0.62 ± 0.31 versus 1.24 ±

0.59 in control mice, NS) (Figure 3a). The mRNA expression of Cyp8b1, which encodes

the gene of the enzyme catalyzing the cholic acid synthesis in the liver and is feedback-

inhibited by bile salts, was significantly increased in EFA-deficient mice (2.23 ± 0.8

versus 1.2 ± 0.3 in control mice, p<0.05) (Figure 3a). However, EFA deficiency did not

have a major effect on the mRNA expression of other genes involved in hepatic bile salt

synthesis (Fxr, Shp, Fgfr4 and Cyp27a1) (Figure 3a). When activated by bile salts in the

intestine, FXR activates the mRNA expression and release of Fgf15, which subsequently

travels to the liver via the portal circulation in order to inhibit the hepatic bile salt

synthesis.23

We observed a decrease in mRNA expression of Fgf15 gene in the terminal

ileum of EFA-deficient (FVB) mice, despite an increased bile salt synthesis and secretion

(Figure 3b). These data are confirmed by measurement of Fgf15 in Fxr+/+

mice on mixed

background (C57BL/6Jx129/OlaHsd); in EFA-deficient Fxr+/+

mice, Fgf15 mRNA

expression was significantly lower than in Fxr+/+

mice on control diet (Figure 4b). These

data indicate an effect of the EFA deficiency on Fgf15 mRNA expression, independent of

the genetic background of the mice. mRNA of Fgf15 was almost absent in Fxr-/-

mice on

both EFA-deficient and control diet (Figure 4b), demonstrating that Fgf15 mRNA

expression is regulated by the FXR, as reported previously.23

The mRNA expression of

FXR itself and genes encoding the ileal bile salt-transporters (Ostα, Ostβ, Ibabp and

Asbt) was not significantly affected by EFA deficiency in mice (Figure 3b).

EFA DEFICIENCY AND BILE SALT METABOLISM

81

4

CH

AP

TE

R

In order to evaluate whether the expression of hepatic transporters of bile salts,

phospholipids and cholesterol correlates with the observed increase in bile flow in EFA-

deficient mice, we measured the mRNA expression of genes encoding the hepatic

transporters (Figure 3c). However, EFA deficiency did not affect the mRNA expression of

the majority of the genes, except for a decrease in Mrp3 (basolateral organic anion

transporter) (Figure 3c).24

Figure 3 mRNA expression of genes involved in (a) hepatic bile salt synthesis, (b) ileal bile salt transport and (c) hepatic transport of bile salts (BS), phospholipids (PL) and cholesterol (CL). Data represented means ± SD of n=6 mice per group. *p<0.05 is the significant difference between EFA-deficient (white bars) and control (black bars) mice.

EFA deficiency associated increase in bile flow is FXR-independent

To study to what extent alterations in bile production upon EFA deficiency were FXR-

dependent, we determined the bile flow by bile cannulation in Fxr-/-

mice and their wild

CHAPTER 4

82

type littermates on EFA-deficient and control diet. Bile flow was enhanced by EFA

deficiency in Fxr+/+

and Fxr-/-

mice (+97% and +112%, respectively, compared with mice

of the same genotype on control diet, p<0.05) (Figure 4a; Table 5). This was

accompanied by an increase in biliary secretion of bile salts in EFA-deficient Fxr+/+

and

Fxr-/-

mice (+36% in EFA-deficient Fxr+/+

mice compared with control Fxr+/+

mice, NS, and

+105% in EFA-deficient Fxr-/-

mice compared with control Fxr-/-

mice, p=0.054). Although

the differences did not reach significant values, there is a trend towards a higher biliary

secretion of bile salts in EFA-deficient mice with both genotypes. Secretion of

phospholipids showed a trend towards an increased secretion in the EFA-deficient mice,

regardless of the genotype (+91% in EFA-deficient Fxr+/+

mice compared with control

Fxr+/+

mice, NS, and +179% in EFA-deficient Fxr-/-

mice compared with control Fxr-/-

mice, p<0.05). These data indicate that the increase in biliary secretion of bile salts and

phospholipids is independent of FXR.

Figure 4 The role of FXR in EFA-deficiency induced alteration in bile production. (a) Bile flow was determined in Fxr

-/- mice and their wild type littermates on EFA-deficient and control diet. (b) mRNA

expression of Fgf15 in Fxr-/-

mice and their wild type littermates on EFA-deficient and control diet. Data represented means ± SD of n=5- mice per group. *p<0.05 is the significant difference between the EFA-deficient (white bars) versus control (black bars) mice of the same genotype. #p<0.05 is the significant difference between Fxr

+/+ and Fxr

-/- mice fed the same diet.

CDCA and GW4064 stimulation of EFA-deficient Caco-2 cells

To study direct effects of EFA deficiency on FXR activation, in vitro experiments were

performed in post confluent Caco-2 cells treated with CDCA and GW4064, both very

potent (and the latter one highly specific) FXR agonists. Upon confluence Caco-2 cells

spontaneously differentiate and develop small intestinal features, as indicated by the

expression of enterocyte markers lactase and sucrose-isomaltase.25

As expected, after

8-10 days in EFA-deficient medium Caco-2 cells showed clear signs of EFA deficiency

as indicated by significantly lower levels of linoleic acid (Figure 5a) and ω-6 family of

fatty acids (Figure 5b) in EFA-deficient compared with control Caco-2 cells.

Triene/tetraene ratio, biochemical marker of EFA-deficiency, was clearly increased in

EFA-deficient Caco-2 cells (0.23±0.06 in EFA-deficient and 0.08±0.05 in control cells;

p<0.05; data not shown). This clearly shows that EFA deficiency in Caco-2 cells

resembles the situation in humans and mice during EFA deficiency and shows that this is

a valid intestinal in vitro model of EFA deficiency. Cellular mRNA expression of intestinal

differentiation marker lactase was significantly lower in EFA-deficient cells, compared

with control Caco-2 cells, without affected morphology of the EFA-deficient cells

EFA DEFICIENCY AND BILE SALT METABOLISM

83

4

CH

AP

TE

R

(unpublished data). These data are in agreement with our previous in vivo observations

in EFA-deficient mice having reduced mRNA expression and enzyme activity of lactase

in mid intestine of EFA-deficient mice.8 Upon treatment with a physiological

concentration of CDCA (50 µM) FGF19 mRNA expression was slightly increased in

control, and to a lesser extent in EFA-deficient Caco-2 cells, although the values did not

reach the significant difference (Figure 5c). FGF19 mRNA expression was further

increased in control Caco-2 cells upon stimulation with higher CDCA concentration of

250 µM (resembling the concentrations bile salts reabsorbed during EFA deficiency in

mice) (Figure 5c). This effect was also seen in EFA-deficient cells after treatment with

250 µM CDCA (Figure 5c). Treatment with GW4064 (1 µM) did not increase mRNA

expression of FGF19 in either control or EFA-deficient Caco-2 cells (Figure 5c). Although

there was no significant difference in FGF19 mRNA expression between EFA-deficient

and control cells, EFA-deficient cells seemed to have slightly lower mRNA expression of

FGF19 compared to control cells in all conditions.

Table 5 Bile flow and biliary secretion rates in Fxr-/-

mice and their wild type littermates (C57BL/6Jx129/OlaHsd background) on EFA-deficient or control diet for 8 weeks. Values are represented as means ± SD (n=5-7 mice per group).

Both CDCA (lower and higher concentrations) and GW4064 (1 µM) significantly induced

the mRNA expression of FXR target-gene IBABP in control Caco-2 cells (Figure 5d).

This effect was almost completely absent in EFA-deficient Caco-2 cells (Figure 5c), as

indicated by absent induction IBABP expression upon 250 µM CDCA and GW4064

treatment in EFA-deficient cells and significantly lower expression of IBABP in EFA-

deficient compared with control cells. Only upon a concentration of 50 µM of CDCA there

was a slight increase in IBABP mRNA expression in EFA-deficient cells compared with

the expression in unstimulated EFA-deficient cells. mRNA concentration of IBABP was

significantly lower in EFA-deficient cells compared with control cells in all conditions.

DISCUSSION

We determined the effects of EFA deficiency on bile flow and kinetic parameters of the

EHC of bile salts in mice, since we previously observed an increased bile flow and biliary

bile salt secretion in combination with fat malabsorption. The mechanism of altered bile

production during EFA deficiency in mice had remained unclear. Insight into this

mechanism will hopefully allow us to design new strategies to interpret and treat fat

malabsorption, specifically in EFA deficiency. Our data show that EFA deficiency in mice

CHAPTER 4

84

increases bile salt synthesis and intestinal reabsorption, resulting in a profoundly

increased bile salt pool size, biliary bile salt secretion and bile flow. In mice in which FXR

was genetically inactivated, EFA deficiency resulted in similar changes as in wild type

mice, regarding biliary secretion of bile salts and bile flow. This finding indicates that the

changes in bile production during EFA deficiency were essentially FXR-independent.

Finally, our data suggest that the increased bile salt synthesis may be related to down-

regulation of ileal Fgf15 gene expression, interfering with the negative feedback

regulation of hepatic bile salt synthesis.

Consistent with our previous findings, after 8 weeks of EFA-deficient diet, mice were

clearly EFA-deficient. Although previous studies in EFA-deficient mice reported slightly

increased cholate and slightly decreased β-muricholate fraction in bile, in present study

EFA deficiency in mice was not associated with major changes in biliary bile salt

composition.7 These data underscore our previous findings that neither a decrease in

biliary flow, nor major alterations in bile composition are the cause of fat malabsorption

in EFA-deficient mice. Theoretically, increased biliary bile salt secretion during EFA

deficiency could act as a compensatory mechanism for reduced absorption of fat. This

is, however, conflicting with studies in rats where fat absorption is similar to that in mice

(80-84%), despite a decreased biliary secretion.6,26

We cannot exclude that the

proposed compensatory mechanism during EFA deficiency is differentially regulated

among the different species.

Enhanced bile flow was associated with increased biliary output of bile salts and

phospholipids. The ratio between bile salts and lipids was similar between EFA-deficient

and control mice (data not shown), suggesting unaltered coupling of bile salts to lipids in

bile upon EFA deficiency. Since the mRNA expression of the hepatic genes encoding

transporters for cholesterol (Abcg5/Abcg8) and phospholipids (Mdr2) was not

significantly changed upon EFA deficiency in mice, it is tempting to assume that the

increased output of lipids in EFA deficiency was entirely based on the increased bile

flow. Increased bile salt secretion was not associated with altered expression of several

genes encoding hepatic bile salts transporters (Bsep, Ntcp, Mrp2 and Oatp1), possibly

due to the fact that the expression of the transporters is not rate limiting factor for the

increased secretion rate of bile salts.27

Moreover, bile salt transporters are localized

along the hepatic acinus, while the bile salt transport is localized mainly at the periportal

zone.28

This implies that the number of hepatocytes, rather than the gene expression of

transporters, is the rate limiting factor during the altered bile salt secretion rate in mice.

Whole body kinetics of cholate,10

clearly demonstrated that the increased biliary

secretion of bile salts is associated with increased bile salt synthesis. The mRNA

expression of Cyp7a1 gene was not affected by EFA deficiency. However, studies on

bile salt formation under different conditions have shown that altered synthesis of bile

salts is not always correlated to changes in Cyp7a1 mRNA expression.29,30

Previous

studies in EFA-deficient mice revealed unaltered mRNA expression of Cyp7a1.7 Since

Cyp7a1 has been shown to have a remarkable circadian mRNA expression, with highest

levels during the night, we cannot exclude that the differences in expression between

EFA-deficient and control mice would have been different during the night.31

Increased

bile salt synthesis was accompanied by the increased mRNA expression of Cyp8b1

gene, encoding the enzyme responsible for the synthesis of cholic acid and control over

EFA DEFICIENCY AND BILE SALT METABOLISM

85

4

CH

AP

TE

R

Figure 5 The effects of EFA deficiency on mRNA expression of FXR-target genes in Caco-2 cells after CDCA and GW4064 stimulation. (a) Linoleic acid concentrations (mol%) in control (black bars) and EFA-deficient (white bars) Caco-2 cells 7 days after confluence. (b) Fatty acid families of essential fatty acids (ω-3 and ω-6) and nonessential fatty acids (ω-7 and ω-8). Control and EFA-deficient cells were treated with vehicle, 50 or 250 µM CDCA or 1 µM GW4064 for 24 hours. Subsequently, cells were harvested for RNA isolation and quantitative PCR analysis of relative mRNA expression of FXR-target genes (c) FGF19 and (d) IBABP was performed. Data of fatty acid determination represent ±SEM of four independent experiments. Quantitative PCR data represent ±SEM of at least three cell experiments. *p<0.05 is the significant difference between the EFA-deficient versus control Caco-2 cells with the same treatment status. #p<0.05 is the significant difference between treated and non-treated Caco-2 cells of the same phenotype.

CHAPTER 4

86

the ratio of cholic acid over chenodeoxycholic acid in the bile.32

However, as stated

previously, our data on biliary composition in EFA-deficient mice do not show major

differences in the concentrations cholic acid and chenodeoxycholic acid in bile compared

with control animals. This suggests that the increased Cyp8b1 mRNA gene expression

does not lead to major physiological changes in EFA-deficient mice. Unlike Cyp7a1,

Cyp8b1 was shown to have the highest mRNA expression during the day; this could be

an explanation for the detected difference in Cyp8b1, but not in Cyp7a1, mRNA

expression between EFA-deficient and control mice.33

Our findings on mRNA gene

expression measurements of Cyp7a1 and Cyp8b1 underscore the importance of

physiological measurements, along with mRNA expression of genes in order to properly

study the EHC of bile salts in vivo. The expression of other relevant genes involved in

bile salt synthesis (Shp, Cyp27a1) remained similar in EFA-deficient compared with

control mice, while the bile salt synthesis was increased.

Stable isotope dilution study revealed an enlargement of the bile salt pool in EFA-

deficient mice, suggesting an impaired feedback inhibition of the hepatic bile salt

synthesis. The FTR, representing the fraction of the pool renewed each day, was

decreased in EFA-deficient mice, while the reabsorption of the bile salts in the intestine

was increased. Normally, the enhanced bile salt reabsorption is expected to activate the

FXR in the ileum and thereby induce the release of Fgf15 into the portal circulation,

which in turn eventually inhibits the bile salts synthesis in the liver.23,34

To our surprise,

we found a decreased mRNA expression of Fgf15 gene in EFA-deficient mice, indicating

a disruption in the intestinal feedback regulation of bile salt synthesis. We realize that the

experimental setting we performed in this study does not allow for direct evidence

demonstrating that the increased bile salts synthesis in EFA-deficient mice is directly

related to, or the result of, the lower plasma concentration of Fgf15. So far, it has not

been possible to determine the concentration of Fgf15 in the (portal) plasma of EFA-

deficient mice and all of the studies performed so far on the role of Fgf15 in bile salt

metabolism in mice are based on the ileal mRNA expression of this gene.23,34,35,36,37,38

For this reason, we performed in vitro experiments in EFA-deficient Caco-2 cells.

Although there might be a trend in lower FGF19 mRNA expression in EFA-deficient

compared with control Caco-2 cells, we do not see a significant difference after

stimulation with CDCA. Unfortunately, we were not able to measure reliable

concentrations of FGF19 secreted in medium of stimulated EFA-deficient and control

cells. Song et al. recently reported successful measurements of secreted FGF19 in

media of cultured human hepatocytes after CDCA treatment for 24 hours.39

However, we

cannot exclude differential regulation of FGF19 secretion in hepatic and intestinal cell

lines after CDCA stimulation. In EFA-deficient mice the mRNA expression of Asbt gene

and the gene encoding intestinal heteromeric basolateral transporter Ostα/β did not

correlate with increased bile salt reabsorption. This suggests a limited role for these

transporters in enhanced reabsorption during EFA deficiency in mice and is in

agreement with several previous findings.9,30,40

mRNA expression of the cytosolic

protein IBABP was not significantly decreased in EFA-deficient mice. However, our in

vitro experiments revealed impaired induction of IBABP mRNA gene expression in EFA-

deficient Caco-2 cells after CDCA and GW4064 stimulation. Discrepancy between in

vivo and in vitro data requires further research. Taken together, in vivo and in vitro data

EFA DEFICIENCY AND BILE SALT METABOLISM

87

4

CH

AP

TE

R

indicate that besides FXR other factors contribute to the altered EHC of bile salts during

EFA deficiency.

In order to determine to what extent the effects of EFA deficiency on bile production

were FXR-dependent, we repeated key experiments in mice lacking FXR and their wild

type littermates. The rationale behind this was based on two findings; first, during the

past decade FXR has been shown to play a crucial role in controlling bile acid

homeostasis,41,42

and second, our data on bile salt kinetics during EFA deficiency in

mice corresponded to a certain extent with changes in bile salt kinetics upon FXR-

inactivation observed by Kok et al.9 Similar to the situation in Fxr

-/- mice, EFA-deficient

mice showed enhanced pool size, increased synthesis and enhanced bile salt

reabsorption of bile salts and similar cycling time compared with control mice. Despite

the similarities in the separate effects of EFA deficiency and FXR deficiency in mice on

bile salt kinetics, we showed that when combined, the effects of EFA deficiency with

additional FXR-inactivation on bile flow and biliary secretion are similar to the effects of

Figure 6 Proposed mechanism of altered bile salt homeostasis during EFA deficiency in mice. Despite an increased secretion of bile salts from the liver to the intestine, bile salt synthesis and pool are increased, while the circulating time of bile salts is unaltered in EFA deficient mice. We suggest that the reduced Fgf15 mRNA expression might be responsible for the lack of the inhibition of hepatic synthesis of bile salts.

EFA deficiency alone. This indicates that when combined, the effects of EFA deficiency

on bile flow and biliary secretion rate of bile salts are independent of FXR. Kok et al.

proposed that the defective negative feedback inhibition of hepatic cholate synthesis

was the consequence of the absence of FXR in vivo. The underlying mechanism,

however, remained unclear.9 In our study we show that the defective negative feedback

inhibition of bile salt synthesis in EFA-deficient mice is probably due to reduced Fgf15

expression. In Figure 6, we propose the mechanism responsible for altered bile salt

kinetics during EFA deficiency in mice. The increased bile salt secretion is consistent

CHAPTER 4

88

with increased synthesis, larger pool size and enhanced ileal reabsorption, without

affecting the cycling time of bile salts. We suggest that the preserved bile salt synthesis

is due to an intestinal (intracellular) defect leading to a decreased, instead of increased,

expression of Fgf15. However, the exact intracellular effects of EFA deficiency on bile

salt activation of FXR and subsequent regulation of the Fgf15 gene expression and its

secretion remain to be elucidated. EFA deficiency can possibly lead to increased cellular

permeability and reduced membrane integrity in the enterocytes, resulting in impaired

uptake of bile salts in the ileal enterocytes. To our knowledge, studies on cellular

permeability in the enterocytes during EFA deficiency have not been performed yet.

Further studies of the effects of EFA deficiency on cellular function will hopefully help us

understand how this correlates to the intestinal regulation of the negative feedback

synthesis of bile salts.

In conclusion, our study demonstrates that EFA deficiency in mice clearly affects bile salt

metabolism at several steps during the EHC of bile salts. We show, indirectly, that

reduced intestinal function is at least partly involved in EFA deficiency associated

alterations in the gut-liver signaling during the bile salt homeostasis. Further studies are

required to determine if adaptations in bile homeostasis can allow for improved fat

absorption during EFA deficiency.

ACKNOWLEDGEMENTS

The authors would like to thank Rick Havinga, Ingrid Martini, Juul Baller, Renze

Boverhof and Theo Boer for excellent technical assistance.

GRANTS

This study was supported by the Dutch Digestive Foundation (MLDS).

REFERENCES

1 Biggemann B, Laryea MD, Schuster A, Griese M, Reinhardt D, and Bremer HJ. Scand J Gastroenterol

Suppl 1988; 143: 135-141.

2 Socha P, Koletzko B, Swiatkowska E, Pawlovska J, Stolarczyk A, and Socha J. Acta Peadiatr 1998;

87(3): 278-283.

3 Strandvik B. Acta Paediatr Scand Suppl 1989; 363: 58-63.

4 Levy E, Garofalo C, Rouleau T, Gavino V, and Bendayan M. Hepatology 1996; 23(4): 848-857.

5 Levy E, Thibault L, Garofalo C, Messier M, Lepage G, Ronco N, and Roy CC. J Lipid Res 1990; 31(11):

2009-2017.

6 Levy E, Garofalo C, Thibault L, Dionne S, Daoust L, Lepage G, and Roy CC. Am J Physiol Gastrointest

Liver Physiol 1992; 262(2): G319-G326.

7 Werner A, Minich DM, Havinga H, Bloks VW, van Goor H, Kuipers F, and Verkade HJ. Am J Physiol

Gastrointest Liver Physiol 2002; 283(4): G900-G908.

8 Lukovac S, Los EL, Stellaard F, Rings EHHM, and Verkade HJ. Am J Physiol Gastrointest Liver Physiol

2008; 295(3): G605-G613.

9 Kok T, Hulzebos C, Wolters H, Havinga R, Agellon L, Stellaard F, Shan B, Schwarz M, and Kuipers F. J

Biol Chem 2003; 278(43): 41930-41937.

EFA DEFICIENCY AND BILE SALT METABOLISM

89

4

CH

AP

TE

R

10 Hulzebos C, Renfurm L, Bandsma R, Verkade H, Boer T, Boverhof R, Tanaka H, Mierau I, Sauer P,

Kuipers F, and Stellaard F. J Lipid Res 2001; 42(11): 1923-1929.

11 Gilardi F, Mitro N, Godio C, Scotti E, Caruso D, Crestani M, and De Fabiani E. Pharmacology &

Therapeutics 2007; 116(3): 449-472.

12 Bergstrom S and Danielsson H. Acta Physiol Scand 1958; 43(1): 1-7.

13 Muskiet FA, van Doormaal JJ, Martini IA, Wolthers BG, and van der Silk W. J Chromatogr 1983; 278(2):

231-244.

14 Werner A, Havinga R, Bos T, Bloks V, Kuipers F, and Verkade H. Am J Physiol Gastrointest Liver Physiol

2005; 288(6): G1150-G1158.

15 Spalinger J, Seidman E, Lepage G, Menard D, Gavino V, and Levy E. Am J Physiol Gastrointest Liver

Physiol 1998; 275(4): G652-G659.

16 Mashige F, Imai K, and Osuga T. Clin Chim Acta 1976; 70: 79-86.

17 Kuipers F, Havinga R, Bosschieter H, Toorop GP, Hindriks FR, and Vonk RJ. Gastroenterology 1985; 88:

403-411.

18 Stellaard F, Langelaar SA, Kok RM, and Jakobs C. J Lipid Res 1989; 30(10): 1647-1652.

19 Grefhorst A, Elzinga B, Voshol P, Plosch T, Kok T, Bloks V, van der Sluijs F, Havekes L, Romijn J,

Verkade H, and Kuipers F. J Biol Chem 2002; 277(37): 34182-34190.

20 Heuman DM. J Lipid Res 1989; 30(5): 719-730.

21 Goodwin B, Jones SA, Price RR, Watson MA, McKee DD, Moore LB, Galardi C, Wilson JG, Lewis MC,

Roth ME, Maloney PR, Willson TM, and Kliewer SA. Molecular Cell 2000; 6(3): 517-526.

22 Makishima M, Okamoto AY, Repa JJ, Tu H, Learned RM, Luk A, Hull MV, Lustig KD, Mangelsdorf DJ,

and Shan B. Science 1999; 284(5418): 1362-1365.

23 Inagaki T, Choi M, Moschetta A, Peng L, Cummins CL, McDonald JG, Luo G, Jones SA, Goodwin B,

Richardson JA, Gerard RD, Repa JJ, Mangelsdorf DJ, and Kliewer SA. Cell Metabolism 2005; 2(4): 217-

225.

24 Hirohashi T, Suzuki H, and Sugiyama Y. J Biol Chem 1999; 274(21): 15181-15185.

25 Chantret I, Barbat A, Dussaulx E, Brattain MG, and Zweibaum A. Cancer Res 1988; 48(7): 1936-1942.

26 Nilsson A and Melin T. Am J Physiol Gastrointest Liver Physiol 1988; 255(5): G612-G618.

27 Wolters H, Elzinga BM, Baller JFW, Boverhof R, Schwarz M, Stieger B, Verkade HJ, and Kuipers F.

Journal of Hepatology 2002; 37(5): 556-563.

28 Groothuis GM, Hardonk MJ, Keulemans KP, Nieuwenhuis P, and Meijer DK. Am J Physiol Gastrointest

Liver Physiol 1982; 243(6): G455-G462.

29 Liu Y, Havinga R, Van der Leij FR, Boverhof R, Sauer PJ, and Stellaard F. Pediatr Res 2008; 63(4): 375-

381.

30 Hulzebos C, Wolters H, Plosch T, Kramer W, Stengelin S, Stellaard F, Sauer P, Verkade H, and Kuipers

F. J Pharmacol Exp Ther 2003; 304(1): 356-363.

31 Noshiro M, Usui E, Kawamoto T, Kubo H, Fujimoto K, Furukawa M, Honma S, Makishima M, Honma Ki,

and Kato Y. J Biol Rhythms 2007; 22(4): 299-311.

32 Zhang M and Chiang JY. J Biol Chem 2001; 276(45):41690-41699.

33 Ishida H, Yamashita C, Kuruta Y, Yoshida Y, and Noshiro M. J Biochem 2000; 127(1): 57-64.

34 Jung D, Inagaki T, Gerard RD, Dawson PA, Kliewer SA, Mangelsdorf DJ, and Moschetta A. J Lipid Res

2007; 48(12): 2693-2700.

35 Gutierrez A, Ratliff EP, Andres AM, Huang X, McKeehan WL, and Davis RA. Arterioscler Thromb Vasc

Biol 2006; 26(2): 301-306.

36 Ricketts ML, Boekschoten MV, Kreeft AJ, Hooiveld GJEJ, Moen CJA, Muller M, Frants RR,

Kasanmoentalib S, Post SM, Princen HMG, Porter JG, Katan MB, Hofker MH, and Moore DD. Mol

Endocrinol 2007; 21(7): 1603-1616.

37 Kim I, Ahn SH, Inagaki T, Choi M, Ito S, Guo GL, Kliewer SA, and Gonzalez FJ. J Lipid Res 2007; 48(12):

2664-2672.

CHAPTER 4

90

38 Ballatori N, Fang F, Christian WV, Li N, and Hammond CL. Am J Physiol Gastrointest Liver Physiol 2008;

295(1): G179-G186.

39 Song KH, Li T, Owsley E, Strom S, and Chiang JY. Hepatology 2009; 49(1): 297-305.

40 Schwarz M, Russell DW, Dietschy JM, and Turley SD. J Lipid Res 1998; 39(9): 1833-1843.

41 Eloranta JJ and Kullak-ublick GA. Physiology 2008; 23(5): 286-295.

42 Kalaany NY and Mangelsdorf DJ. Annual Review of Physiology 2006; 68(1): 159-191.

CHAPTER 5

FUNCTIONAL CHARACTERIZATION OF AN IN VITRO MODEL OF

ESSENTIAL FATTY ACID DEFICIENCY IN INTESTINAL EPITHELIAL

CELLS

S. Lukovac1, E.H.H.M. Rings

1, H.J. Verkade

1

(1) Pediatric Gastroenterology, Department of Pediatrics, Beatrix Children’s

Hospital, Groningen University Institute for Drug Exploration (GUIDE), Center for

Liver, Digestive and Metabolic Diseases, University of Groningen, University

Medical Center Groningen, Groningen, The Netherlands.

Manuscript in preparation

CHAPTER 5

94

ABSTRACT

In vivo studies in mice have indicated that essential fatty acid (EFA) deficiency negatively

affects small intestinal function. It has remained difficult, however, to unravel the

molecular mechanism(s), during EFA deficiency, which affect(s) small intestinal

epithelium in vivo. The aim of the present study was to develop and characterize an in

vitro model for EFA deficiency in small intestinal cells, which could allow unraveling the

(intra)cellular molecular mechanism(s) with respect to the effect on small intestinal

function.

Intestinal epithelial cells (Caco-2) were cultured in medium containing normal or

delipidated fetal calf serum (FCS) for at least 1 week post-confluence. We characterized

fatty acid profiles, morphology and mRNA expression of relevant small intestinal markers

lactase and sucrase isomaltase in EFA-deficient and control Caco-2 cells. Cellular

permeability was assessed by transepithelial electrical resistance (TER) and by

determination of the mRNA expression of relevant tight junction components and

localization of ZO-1. To study the reversibility and specificity of the effects of EFA

deficiency, we cultured the EFA-deficient cells in medium supplemented with linoleic acid

(LA).

The culturing of Caco-2 cells with delipidated FCS decreased the concentration of LA by

81% (p<0.01) and increased the triene:tetraene ratio (+187.5%; p<0.01), compared with

control cells and represents EFA deficiency in these cells. The morphology of the cells

was not severely affected, although mRNA expression of relevant differentiation markers

of the brush border membrane of the small intestine was severely reduced in EFA-

deficient Caco-2 cells. The cellular permeability was significantly increased as assessed

by TER. However, the mRNA expression of tight junction components and localization of

ZO-1 were not affected in EFA-deficient Caco-2 cells. Although the LA supplementation

to EFA-deficient Caco-2 cells was incorporated into cellular phospholipids to a similar

extent as in the control cells, it did not restore reduced mRNA expression of small

intestinal differentiation markers or the increased permeability.

Caco-2 cells exposed to delipidated FCS rapidly demonstrate EFA deficiency with

characteristics mimicking EFA deficiency in the small intestine in vivo. EFA deficiency

severely reduced expression of relevant brush border markers of the small intestine, and

impaired cellular permeability. Interestingly, these effects were not immediately

reversible by LA supplementation to EFA-deficient Caco-2 cells.

IN VITRO MODEL OF EFA DEFICIENCY

95

5

CH

AP

TE

R

INTRODUCTION

Essential fatty acid (EFA) deficiency is a frequent complication of diseases in which fat

malabsorption occurs, mainly during cholestatic liver diseases in pediatric patients with

limited fat storage. Interestingly, EFA deficiency by itself can also induce fat

malabsorption and impair small intestinal function, probably by intracellular

mechanism(s), in rats and mice.1,2,3

Studies on the intracellular consequences of EFA

deficiency in small intestinal enterocytes have been scarce.1 Spalinger et al.

demonstrated in EFA-deficient Caco-2 cells in vitro that supplementation with structured

triglycerides could increases cellular concentrations of linoleic acid (LA, C18:2ω-6) and

its metabolite arachidonic acid (AA, C20:4ω-6), resulting in the correction of the

biochemical marker of EFA deficiency (triene:tetraene ratio).1 However, the functional

consequences of EFA deficiency or those of structural triglyceride supplementation to

EFA-deficient cells were not explored in this in vitro model. EFA deficiency in mice and

rats has been associated with reduced fat and disaccharide absorption,2,3

reduced

expression and enzyme activity of the small intestinal differentiation marker (lactase) and

impaired negative feedback regulation of bile salt synthesis.4 To understand the

mechanism(s) by which EFA deficiency influences small intestinal function, we reasoned

that further characterization of an in vitro model is essential. Therefore, in present study

we further characterized an in vitro model of EFA deficiency in Caco-2 cells, adapted

from the protocol of Spalinger et al.1 We validated our model by analyzing the fatty acid

profiles and triene:tetraene ratio in EFA-deficient Caco-2 cells. Furthermore, we

compared relevant functional parameters in the in vitro model with our previous in vivo

findings in EFA-deficient mice. It has been debated whether EFA deficiency affects small

intestinal permeability.5,6

Therefore, in our functional characterization, we included the

analysis of the permeability of small intestinal cells. Cellular permeability was assessed

by determination of the transepithelial electrical resistance (TER) and by analysis of tight

junction components. Our previous studies in EFA-deficient mice pointed out a strong

correlation between LA concentrations in mucosal phospholipids and lactase mRNA

expression and enzyme activity. To analyze the robustness and specificity of the EFA

deficiency in our in vitro model, we re-supplied LA to EFA-deficient cells. We determined

whether LA supplementation leads to uptake and incorporation of LA into phospholipids,

and whether it can reverse the effects of EFA-deficient phenotype.

We clearly show that EFA-deficient Caco-2 cells are a useful model to study in more

detail the effects of EFA deficiency. Our in vitro data indicate that EFA deficiency

negatively affects the cellular permeability and the presence of typical differentiation

markers of small intestinal cells, and that these effects are not rapidly reversible by LA

supplementation.

MATERIAL AND METHODS

Delipidation of FCS

FCS was delipidated by means of di-iso-propylether and butanol extraction, according to

the protocol of Cham and Knowles.7 This protocol is known to eliminate over 90% of all

fatty acids in a solution. Subsequent to delipidation, both delipidated and control FCS

CHAPTER 5

96

were dialyzed for 72 hours in 0.9% NaCl at 4°C by means of the Spectra/Por3 molecular

porous membrane tubing (MW 3500; Spectrumlabs, Rancho Dominguez, CA, USA).

Coomassie blue staining

Protein concentration was determined according to the manufacturer’s protocol (BCA kit;

Pierce Biotechnology Inc., Rockford, Ill, USA). 10 μg of protein from the FCS was

analyzed on gel by means of the standard protocol for Coomassie Blue staining to

determine the amount and possible selectivity of protein loss by the delipidation

protocol.8

Cell culture

Human Caco-2, an immortalized line of heterogeneous human epithelial colorectal

adenocarcinoma, cells from the American Type Tissue Culture Collection (Manassas,

VA) were cultured in Dulbecco's modified Eagle's medium containing 10%, penicillin

(100 units/mL)/streptomycin (100 μg/mL), 1% non-essential amino acids and 0.25%

human transferrin in an atmosphere of 5% CO2–95% air at 37 °C. Cells were subcultured

at 90% confluence (approximately every 5 days) by trypsin. For the experiments, cells

between passages 20 and 40 were used. Caco-2 cell line was used in the experiments

because this is the only immortalized cell line which differentiates in vitro into small-

intestinal enterocyte-like cells, expressing the hydrolases lactase and sucrase-

isomaltase.9,10

These characteristics make Caco-2 cells a valid model to study the

function of the small intestinal enterocytes.

Induction of EFA deficiency in Caco-2 cells

Cells were made EFA-deficient according to the adapted protocol of Spalinger et al.

1

Shortly, medium was replaced by DMEM supplemented with dialyzed FCS (control cells)

or with delipidated FCS (EFA-deficient cells) 1 day after seeding. Seven

days after

reaching complete confluence, the cells were harvested for several analytic procedures

described below (Morphology and Immunofluorescent stainings, Thin layer

chromatography-TLC, Fatty acid methylation and Gas chromatography, RNA isolation

and Quantitative PCR). By this time, cells were cultured in EFA-deficient medium for ten

days.

LA supplementation to EFA-deficient Caco-2 cells

In the first experiment, at day 7 after plating control and EFA-deficient cells were

supplemented with 50 µM LA and cells were harvested at day 0, 2 and 6 after

supplementation for linoleic acid analysis in total cell lysates or thin layer

chromatography (TLC) was performed for fatty acid analysis in different lipid classes

(phospholipids and triglycerides; see below section Thin layer chromatography for

details).

In the second experiment, at day 7 after plating EFA-deficient Caco-2 cells were

randomly separated in two groups. One group continued in culture with EFA-deficient

medium, while the other group received DMEM supplemented with 50 µM LA (Sigma

Chemical Co., St. Louis, MO, USA). Cells were harvested at day zero and every two

days afterwards for 6 days. Subsequently, harvested cells were used for RNA isolation.

IN VITRO MODEL OF EFA DEFICIENCY

97

5

CH

AP

TE

R

In a separate, similar experiment, control and EFA-deficient cells were separated and

half of control cells was treated with 50 µM LA and half of the EFA-deficient cells was

treated with 50 µM LA. The other half of the cells received regular (control) or EFA-

deficient medium. Cells were harvested at day 0, 4 and 6 after LA supplementation. After

lipid extraction, TLC was performed for fatty acid analysis in different lipid classes

(phospholipids and triglycerides).

Transelectrical epithelial resistance (TER)

For TER measurements, cells were cultured in sterile polystyrene transwell plates

(Costar, Corning, NY, USA), for a period of 2 weeks. TER was measured every second

day after by means of the EVOM meter (World Precision Instruments, Sarasota, Fl,

USA). For the experiments with LA supplementation, TER was measured for a period of

one week. Control cells (cultured in DMEM with normal FCS) were set at 100%.

Analytical methods

Fatty acid methylation and Gas chromatography

Cells were washed twice with PBS and harvested for fatty acid extraction, hydrolysis and

methylation according to the protocol of Muskiet et al.11

Subsequent separation and

quantification of fatty

acid methyl esters was performed by gas chromatography.

Heptadecaenoic acid (C17:0) was added to all samples as an internal standard before

extraction and methylation procedures, and butylated hydroxytoluene was added as an

antioxidant.

Thin layer chromatography (TLC)

Total triglycerides and phospholipids were separated in harvested cells, after lipid

extraction,12

by means of TLC, as described previously.13

Subsequent to TLC, lipid

fractions were scraped and fatty acid extraction, hydrolysis and methylation was

performed as described above, followed by fatty acid analysis by gas chromatography.13

For the separation of different phospholipid classes, chloroform/methanol/acetic

acid/water was used as the running solvent. For the separation of different lipid classes

(triglycerides and phospholipids) hexane/diethyl ether/acetic acid was used as the

running solvent.

Morphology and Immunofluorescent stainings

Morphology of Caco-2 cells was assessed by hematoxylin and eosin staining of formalin-

fixated cells grown in 6-well cell culture plates.

Immunocytochemical staining for tight junction component ZO-1 was performed in Caco-

2 cells. Rabbit rabbit anti-ZO-1 antibody was from Zymed Laboratories Inc. (South San

Francisco, CA, USA). Cells were seeded and grown on cover slips in DMEM with control

or delipidated (EFA-deficient) serum. At the end of the experiment, Caco-2 cell

monolayers were washed in PBS and fixed in acetone at -20°C for 15 min. Cell

monolayers were washed with PBS afterwards on room temperature for 15 min. on a

shaker. Subsequently, the cells were incubated with primary antibodies for 2h at room

temperature (1:100 dilution in 1% BSA/PBS) followed by three washes with PBS. The

incubation with the secondary antibody (FITC-conjugated goat anti-rabbit, 1:400 dilution

CHAPTER 5

98

c

in 1% BSA/PBS) was performed at room temperature for 30 min. Cells were mounted on

a slide and the fluorescence was examined using a fluorescent microscope (Zeiss

Hal100; Carl Zeiss BV, Sliedrecht, the Netherlands) Images were stacked using the

software, Zeiss Axio Vision (Release 4.6.3; Carl Zeiss BV, Sliedrecht, the Netherlands).

RNA isolation and Quantitative PCR

At the end of the experiment, cells were washed in PBS and RNA isolation was

performed using TRIzol reagent (Invitrogen, Breda, the Netherlands), followed by

quantitative PCR analysis of mRNA expression (TaqMan) as previously described.14

PCR results were normalized to the mRNA expression

of the housekeeping gene

GAPDH. Primer and probe sequences for the Q-PCR analysis have been published

(www.LabPediatricsRug.nl: Realtime PCR Primers & Probes Database).

Figure 1 (a) Total fatty acid concentrations and (b) molar concentrations of fatty acid families in control (black bars) and delipidated FCS (white bars). (c) Coomassie blue staining of the protein content in control (lane 1) and delipidated FCS (lane 2).

Statistical analysis

Data were statistically analyzed by Student's two-tailed t-test. For all experiments, p-

values below 0.05 were considered statistically significant.

RESULTS

Delipidation

Cells were cultured in medium containing delipidated FCS or control FCS. Fatty acids

were removed almost completely by the delipidation of FCS (Figure 1a). There was no

major difference in classes of fatty acids removed; both essential (ω-3 and ω-6) and non-

essential (ω-7 and ω-9) fatty acids were virtually absent in delipidated FCS (Figure 1b).

Protein content was not significantly altered in concentration or composition by the

delipidation, as indicated by the Coomassie blue staining, indicating specificity of the

procedure for removal of fatty acids, rather than for proteins (Figure 1c).

Fatty acid profiles

One week after the full confluence, cultured cells were harvested for fatty acid analysis

by means of gas chromatography. In EFA-deficient cells the biochemical marker for EFA

deficiency, triene:tetraene ratio, was significantly higher compared with control cells

(Figure 2a). Total fatty acid concentration was lower in EFA-deficient cells, but the

IN VITRO MODEL OF EFA DEFICIENCY

99

5

CH

AP

TE

R

difference did not reach the significant difference (Figure 2b). While the molar

concentrations of monounsaturated fatty acids (MUFA) tended to be higher in EFA-

deficient cells, the molar concentrations of polyunsaturated fatty acids (PUFA) tended to

be lower in EFA-deficient cells (Figure 2c). However, both differences did not reach the

statistical significance. The molar concentrations of saturated fatty acids (SAFA) were

similar in both EFA-deficient and control cells (Figure 2c).

Figure 2 (a) Triene:tetraene ratio, (b) total fatty acid concentrations, (c) different fatty acid classes

and (d) phosphatidylcholine/phosphatidylethanolamine concentrations in control (black bars) and

EFA-deficient (white bars) Caco-2 cells. Cells were cultured in control or EFA-deficient medium for

one week after complete confluence. Data represent means ± SEM of four independent

experiments and *p<0.05 is the significant difference between the EFA-deficient and control cells.

Phosphatidylcholine (PC) and phosphatidylethanolamine (PE) are major phospholipids in

mammalian membranes. Li et al. recently proposed the PC/PE ratio as a marker for cell

membrane integrity in hepatic tissue.15

We evaluated whether EFA deficiency in Caco-2

cells leads to reduced phospholipid concentrations and alterations in the PC/PE ratio.

There is no significant difference in PC or PE concentrations (or the ratio of these two,

data not shown) between EFA-deficient and control cells (Figure 2d).

Morphology

Cell morphology was assessed every two days of culture to monitor the growth and

confluence of the cells. There was no significant difference observed in the morphology

of the cells at day 0 and day 4 post-confluence between EFA-deficient and control cells.

CHAPTER 5

100

EFA-deficient cell at day 7 post confluence seemed to have several large structures,

possibly gaps or vacuoles (indicated by the white arrows in Figure 3).

Permeability

Transcellular permeability of the cells was assessed by the TER measurements every

two days after complete confluence, for two weeks. EFA-deficient cells followed the

same increasing trend of TER during the first three days of culture (Figure 4a). At day 3,

EFA-deficient cells had even higher TER values than control cells. However, from day 3

TER values decreased drastically in EFA-deficient cells down to 50% of control values at

day 5-6 (Figure 4a), down to basal levels measured at day 1 (~100 Ω, equal to an empty

transwell). These data indicated increased transcellular permeability in EFA-deficient

Caco-2 cells compared with control cells. The values of control cells were stable from

day 14 and further post-confluence (~300 Ω; data not shown).

In addition to the transcellular route, transport of molecules from the apical to the

basolateral compartment can occur by means of paracellular transport. Paracellullar

permeability across the epithelia is regulated by the tight junctions.16

We analyzed the

mRNA expression three relevant tight junction components, ZO-1, occludin and claudin

1. In addition, localization of ZO-1 was assessed by immunofluorescent staining.

Quantitative PCR analysis revealed similar mRNA expression of all three tight junction

markers (Figure 4b), indicating that EFA deficiency does not affect transcriptional

regulation of these tight junction components in vitro. Immunofluorescent staining was

performed to study the protein expression and localization of ZO-1 (Figure 4c). As

expected, in both control and EFA-deficient Caco-2 cells ZO-1 was expressed at the

intracellular junctions of the monolayers.

Figure 3 Representative picture of the hematoxylin/eosin staining of EFA-deficient (lower panel) and control (upper panel) Caco-2 cells at day 0, 4 and 7 after complete confluence. The staining was performed twice in two independent experiments.

IN VITRO MODEL OF EFA DEFICIENCY

101

5

CH

AP

TE

R

In EFA-deficient Caco-2 cells ZO-1 the expression is not only restricted to the

intracellular junctions, but seemed to be present in the intracellular compartment as well

(Figure 4c).

Small intestinal markers/differentiation

In order to determine if EFA deficiency affects small intestinal differentiation in vitro, we

measured mRNA expression of relevant enterocyte markers lactase and sucrase

isomaltase. Upon confluence, Caco-2 cells normally differentiate from the colonic

towards the small intestinal phenotype. This differentiation process is accompanied by

an increasing mRNA expression of lactase and sucrase isomaltase. Despite the

relatively normal morphology, EFA-deficient Caco-2 cells showed almost absent

expression of lactase and sucrase isomaltase (Figure 5a), while the expression in control

cells was comparable to previous studies in post-confluent Caco-2 cells.10

Lactase and

sucrase isomaltase are expressed at the apical cell compartment, at the brush border

membrane. The mRNA expression of another intestinal marker villin, which is not

exclusively located at the brush border membrane, but also intracellularly, was not

affected by EFA deficiency (Figure 5a). Fatty acids are known as the endogenous

ligands for peroxisome proliferator-activated receptors (PPARs). We determined whether

EFA-deficient Caco-2 cells cultured in medium with delipidated serum expressed lower

Figure 4 (a) Transepithelial electrical resistance (TER) in EFA-deficient Caco-2 cells as a percentage of control cells over a period of 2 weeks of culture (control cells reached the maximum TER values of 350 Ohm, while the EFA-deficient Caco-2 cells had maximum TER values of 250 Ohm). (b) mRNA expression of tight junction components occludin, ZO-1 and claudin in EFA-deficient (white bars) and control cells (black bars). (c) Fluorescence staining for ZO-1 in control (left panel) and EFA-deficient (right panel) cells. Data represent means ± SEM and *p<0.05 is the significant difference between the EFA-deficient and control cells. The experiments were performed at least in duplicate.

CHAPTER 5

102

mRNA expression of PPARs. EFA-deficient Caco-2 cells showed reduced mRNA

expression of PPARα (Figure 5b) compared with control Caco-2 cells, while the

expression of PPARδ and PPARγ was similar. Function of PPARs requires

heterodimerization with the retinoid X receptor (RXR). There was no difference in RXR

expression between EFA-deficient and control Caco-2 cells (Figure 5b).

Effects LA supplementation to EFA-deficient Caco-2 cells

Fatty acid profiles in our in vitro and in vivo studies revealed that EFA deficiency in the

small intestinal epithelium mostly affected the concentrations of the LA (specifically in the

phospholipids). Therefore, we aimed to study the specificity and reversibility of the

effects of EFA deficiency. For this reason we cultured the EFA-deficient cells in medium

supplemented with LA.

Figure 5 (a) mRNA expression of relevant small intestinal enterocyte markers lactase, sucrase isomaltase and villin was measured in EFA-deficient (white bars) and control (black bars). (b) mRNA expression of three types of PPARs (α, δ and γ) and of RXRα in EFA-deficient (white bars) and control (black bars).

Data represent means ± SEM of four independent experiments and

*p<0.05 is the significant difference between the EFA-deficient and control cells.

Supplementation with LA resulted in increasing concentrations of LA in total cell lysates

of both control and EFA-deficient Caco-2 cells (Figure 6a). This would be beneficial if LA

was incorporated into the phospholipids, which are normally retained within the

enterocytes and mobilized towards the membrane. Triglycerides, on the other hand, are

not maintained within the enterocytes, but are mainly secreted at the basolateral

membrane after synthesis. Therefore, we determined the LA concentrations in both

phospholipid and triglyceride fractions of control and EFA-deficient cells. LA

concentrations in phospholipids were significantly lower in EFA-deficient compared with

control Caco-2 cells at day 0 (before the start of the LA supplementation; Figure 6b),

supporting our previous in vivo findings in jejunum of EFA-deficient mice.2 LA

concentrations in total cell lysates and in triglycerides fractions were similar between the

EFA-deficient and control mice at all time points (Figure 6a, 6c). After the LA

supplementation, LA was taken up by both control and EFA-deficient cells to a similar

extent in both total cell lysates as in phospholipids (Figure 6a, 6b). There was no

increase in LA concentration in time after the LA supplementation in EFA-deficient or the

control cells, indicating that LA was not taken up by the triglycerides (Figure 6c).

IN VITRO MODEL OF EFA DEFICIENCY

103

5

CH

AP

TE

R

To functionally study if LA supplementation reverses the effects of EFA deficiency, we

analyzed the mRNA expression of lactase, sucrase isomaltase and PPARα. In addition,

we determined the effects of LA supplementation on TER in EFA-deficient Caco-2 cells.

Preliminary data demonstrate that LA supplementation did not increase the mRNA

expression of lactase, sucrase isomaltase or PPARα in EFA-deficient Caco-2 cells

(Figure 7a, 7b, 7c). TER values were significantly higher in EFA-deficient Caco-2 cells

supplemented with LA on day 2 and 4 during the LA supplementation (Figure 7d).

However, after 4 days there TER values decrease to a similar extent in LA treated and

untreated EFA-deficient cells and EFA-deficient cells supplemented with LA, towards the

baseline values (equal to empty wells; negative controls) (Figure 7d).

Figure 6 Molar concentrations of LA in (a) total cell lysates, (b) phospholipids and (c) triglycerides in EFA-deficient cells (white squares) and control cells (black squares) treated with 50 μM LA. Data represent means ± SEM of two independent experiments and *p<0.05 is the significant difference between the EFA-deficient and control cells.

DISCUSSION

We aimed to develop an in vitro model of EFA deficiency, which would be helpful to

identify the (intra)cellular (molecular) mechanism(s) underlying the negative effects of

CHAPTER 5

104

EFA deficiency on the small intestinal function level.2,17

Our data show that Caco-2 cells

cultured in medium with delipidated FCS rapidly develop biochemical EFA deficiency and

have a phenotype that resembles several aspects of EFA deficiency in the small

intestine in mice in vivo. Caco-2 cells cultured in delipidated medium show an elevated

triene:tetraene ratio in their fatty acid profile, a biochemical marker of EFA deficiency.

The difference in triene:tetraene ratio between the control and EFA-deficient Caco-2

cells after 10 days of culture were not as large as observed previously in the in vitro

model of Spalinger et al. and in EFA deficient mice.1,17

Figure 7 Preliminary data on mRNA expression of (a) lactase, (b) sucrase isomaltase and (c) PPARα in untreated EFA-deficient cells (black squares) and EFA-deficient cells treated with 50 μM LA (white squares) harvested every two days. Data represent means ± SD of three wells per condition. (d) TER was measured every two days in untreated EFA-deficient Caco-2 cells (black squares) and EFA-deficient Caco-2 cells treated with 50 μM LA (white squares) as a percentage of control (untreated) cells over a period of 8 days in culture. Data represent means ± SEM of two independent experiments and *p<0.05 is the significant difference between the EFA-deficient and control cells.

Explanation for the observed difference between our in vitro and in vivo models could be

related to differences in timeframe of the exposure to the EFA-deficient condition. Mice

received an EFA-deficient diet for 8 weeks, whereas Caco-2 cells were exposed to EFA-

deficient medium for approximately ten days. As expected, EFA-deficient Caco-2 cell

had decreased molar concentrations of essential fatty acids in their fatty acid profiles.

Another difference between our in vivo and in vitro model is that Caco-2 cells did not

received almost any of the fatty acids, while the mice received reduced concentrations of

dietary fatty acids. Thus, there is a possibility that, at least, a part of the phenotype we

IN VITRO MODEL OF EFA DEFICIENCY

105

5

CH

AP

TE

R

observed in Caco-2 cells might be due to a complete, rather than reduced, fatty acid

deficiency. However, similar to small intestinal enterocytes, Caco-2 cells are capable of

de novo synthesis of non-essential fatty acids. Therefore, it is not likely that the

phenotype observed in Caco-2 cells is mainly due to the total fatty acid deficiency. This

is supported by the observation that several phenotypic features of the in vivo model of

EFA deficiency are present in our in vitro model in Caco-2 cells.

In agreement with studies in rats and mice, Caco-2 deficient cells show significantly

reduced concentrations of LA in the phospholipids.2,18

Furthermore, the mRNA

expression of lactase was reduced in EFA-deficient Caco-2 cells. We previously showed

in EFA-deficient mice that lactase mRNA and enzyme activity were reduced by more

than 50%, associated with impaired lactose digestion.2 Lactase is an apical disaccharide

and a relevant marker for the differentiation of Caco-2 cells differentiating towards the

small intestinal phenotype.10

The mRNA expression of sucrase isomaltase, another small

intestinal differentiation marker, was also significantly reduced in EFA-deficient Caco-2

cells. Similar to lactase, sucrase isomaltase is localized in the brush border membrane of

the enterocytes.19

Interestingly, the expression of villin, which is an intracellular

enterocyte marker, was not affected by EFA deficiency in Caco-2 cells. Combined, these

data suggest an impaired differentiation of EFA-deficient cells at the level of transcription

of brush border membrane anchored proteins. Possibly, the (ultra)structure of the brush

border membrane is impaired by the EFA deficiency in the small intestinal enterocytes.

This idea is supported by previous studies in EFA-deficient piglets and rats which show

several EFA deficiency associated alterations in the intestinal brush border

membrane.20,21

Whether EFA deficiency in mice and Caco-2 cells leads to

(ultra)structural changes in the brush border membrane, remains to be elucidated.

Electron microscopy analysis seems warranted to determine whether EFA deficiency

affects the brush border membrane in Caco-2 cells and in vivo in EFA-deficient mice.

However, possible negative effects on the brush border membrane by EFA deficiency

cannot explain the decreased mRNA levels of the brush border membrane enzymes.

Possibly, EFA deficiency affects the mRNA synthesis and transport of the brush border

membrane enzymes. The mechanism underlying the specific effects of EFA deficiency

on the expression of lactase and sucrase isomaltase is presently unclear. In a

transcriptome analysis, no differences were detected in the biological processes involved

in RNA synthesis and transport in the microarray analysis of EFA-deficient Caco-2 cells

(unpublished data).

Previous studies revealed possible role of EFA deficiency on the barrier function in

different tissues among different species.5,20,22,23

Our data demonstrate that EFA-

deficient Caco-2 cells are more permeable, as demonstrated by increased TER. The

increased permeability in the EFA-deficient Caco-2 cells was accompanied by unaffected

(mRNA) expression of the tight junction components. Localization of the ZO-1 tight

junction component, furthermore, was not significantly different from control Caco-2 cells.

Since tight junction components are mainly involved in the paracellular permeability, we

suggest that EFA deficiency mainly affects the transcellular permeability, while the

paracellular permeability is preserved during EFA deficiency. Additional measurements

of cascade blue dextran, used as a marker for paracellular permeability, uptake in Caco-

2 cells, will further help elucidate whether paracellular permeability is affected by the

CHAPTER 5

106

EFA deficiency.24,25

In vivo studies (e.g. lactulose/mannitol permeability test) in models

for EFA deficiency might further help elucidate in vivo whether small intestinal

permeability is increased in EFA-deficient mice. Furthermore, Proksch et al. suggested

that LA may play a direct role in the epidermal permeability barrier and that impaired

epidermal barrier function might be reversed by local LA administration.5 Several

attempts have been made in order to reverse the effects of EFA deficiency, mainly in

pediatric cystic fibrosis, including LA supplementation in form of corn oil, safflower oil and

LA-monoacylglycerides.26,27,28

Although the plasma concentrations of LA were

normalized, the effects of LA supplementation on restoring the phenotype of EFA

deficiency remained poor. To address to what extent the observed phenotype of EFA

deficiency (decreased mRNA expression of the brush border enzymes, increased

permeability) depends on LA concentration, short term LA re-supplementation

experiment was performed. Our results indicate that the EFA-deficient Caco-2 cells take

up the administered LA, based on our measurements of LA concentrations in both total

cell lysates, as well as in phospholipids fractions. Interestingly, however, the re-

supplementation of LA did not restore relevant parameters of the EFA-deficient

phenotype, such as the decreased lactase expression. However, these data are

preliminary since the experiment was performed only once. Therefore, further

confirmation by is required in order to demonstrate if LA supplementation has any

positive effects on mRNA expression of lactase, sucrase isomaltase and PPARα. LA

supplementation had a very short, acute effect as demonstrated by the increasing TER

in EFA-deficient cells during the first 4 days of supplementation. After this time point, the

TER values decreased and were similar to untreated EFA-deficient Caco-2 cells. These

data suggest that different tissues are more or less susceptible to the LA

supplementation, since Proksch et al. demonstrated positive effects of local LA

administration in the epidermis.5 Further studies with longer exposure to, and higher

concentrations of LA are necessary.

In the fatty acid analysis in EFA-deficient Caco-2 cells, we did not measure any relevant

difference in alpha-linolenic acid (ALA) concentrations between EFA-deficient and

control Caco-2 cells (data not shown). However, it would be relevant to determine

whether additional supplementation with an ω-3 fatty acid would help to reduce the EFA-

deficient phenotype in EFA-deficient Caco-2 cells.

Our previous studies in EFA-deficient mice revealed impaired fatty acid absorption and

lactose digestion.2,17

Furthermore, we have showed that EFA deficiency leads to

impaired bile salt metabolism in mice and EFA-deficient Caco-2 cells.4 Future studies

with stable isotope labeled nutrients in transwell system with EFA-deficient Caco-2 cells

will reveal whether nutrient absorption is impaired in this in vitro model.

Overall, we have further characterized an in vitro model of EFA deficient small intestinal

cells. In several aspects the phenotype corresponds with in vivo EFA deficiency of the

small intestinal epithelium in mice. We expect that this model will allow performing more

detailed studies on the underlying mechanism(s) of the EFA-deficient phenotype in the

small intestinal enterocyte. Understanding the mechanism(s) by which EFA deficiency

affects the small intestine will hopefully contribute to develop more rational therapies to

improve the nutritional status of patients with EFA deficiency.

IN VITRO MODEL OF EFA DEFICIENCY

107

5

CH

AP

TE

R

ACKNOWLEDGEMENTS

The authors thank Ingrid Martini and Juul Baller for excellent technical assistance and

helpful suggestion.

GRANTS

This study was supported by the Dutch Digestive Foundation (MLDS).

REFERENCES

1 Spalinger J, Seidman E, Lepage G, Menard D, Gavino V and Levy E. Am J Physiol Gastrointest Liver

Physiol 1998; 275(4):G652-G659.

2 Lukovac S, Los EL, Stellaard F, Rings EH and Verkade HJ. Am J Physiol Gastrointest Liver Physiol 2008;

295(3):G605-G613.

3 Miyano T, Yamashiro Y, Shimizu T, Arai T, Suruga T and Hayasawa H. J Pediatr Surg 1986; 21(3):277-

281.

4 Lukovac S, Los EL, Stellaard F, Rings E.H. and Verkade HJ. Am J Physiol Gastrointest Liver Physiol

2009; 297(3):G520-31.

5 Proksch E, Feingold KR and Elias PM. J Invest Dermatol 1992; 99(2):216-220.

6 Hallberg K, Grzegorczyk A, Larson G and Strandvik B J Pediatr Gastroenterol Nutr 1997; 25(3):290-295.

7 Cham BE, Knowles BR. J Lipid Res 1976; 17(2):176-181.

8 Steinberg TH. Chapter 31 Protein Gel Staining Methods: An Introduction and Overview 2009; Volume

463:541-563.

9 Hauri HP, Sterchi EE, Bienz D, Fransen JA and Marxer A. J Cell Biol 1985; 101(3):838-851.

10 van Beers EH, Ai RH, Rings EH, Einerhand AW, Dekker J and Büller HA. Biochem J 1995; 308(Pt 3):769-

775.

11 Muskiet FA, van Doormaal JJ, Martini IA, Wolthers BG and van der Silk W. J Chromatogr 1983;

278(2):231-244.

12 Bligh EG, Dyer WJ. Can J Biochem Physiol 1959; 37(8):911-917.

13 Werner A, Bongers M, Bijvelds M, de Jonge H and Verkade H. J Lipid Res 2004; 45(12):2277-2286.

14 Grefhorst A, Elzinga B, Voshol P, Plosch T, Kok T, Bloks V, van der Sluijs F, Havekes L, Romijn J,

Verkade H and Kuipers F. J Biol Chem 2002; 277(37):34182-34190.

15 Li Z, Agellon L, Allen T, Umeda M, Jewell L, Mason A and Vance D. Cell Metabolism 2006; 3(5):321-331.

16 Anderson JM, Van Itallie CM. Am J Physiol Gastrointest Liver Physiol 1995; 269(4):G467-G475.

17 Werner A, Minich DM, Havinga H, Bloks VW, van Goor H, Kuipers F and Verkade HJ. Am J Physiol

Gastrointest Liver Physiol 2002; 283(4):G900-G908.

18 Christon R, Meslin JC, Thévenoux J, Linard A, Léger CL and Delpal S. Reprod Nutr Dev 1991; 31(6):691-

701.

19 Thomson AB, Keelan M, Thiesen A, Clandinin MT, Ropeleski M and Wild GE. Digestive Diseases and

Sciences 2001; 46(12):2567-2587.

20 Christon R, Even V, Daveloose D, Leger C and Viret J. Biochimica et Biophysica Acta (BBA) -

Biomembranes 1989; 980(1):77-84.

21 Daveloose D, Linard A, Arfi T, Viret J and Christon R. Biochim Biophys Acta 1993; 1166(2-3):229-237.

22 Wertz PW, Cho ES and Downing DT. Effect of essential fatty acid deficiency on the epidermal

sphingolipids of the rat. Biochim Biophys Acta 1983; 753(3):350-355.

23 Melton JL, Wertz PW, Swartzendruber DC and Downing DT. Biochimica et Biophysica Acta (BBA) -

Lipids and Lipid Metabolism 1987; 921(2):191-197.

CHAPTER 5

108

24 Raimondi F, Santoro P, Barone MV, Pappacoda S, Barretta ML, Nanayakkara M, Apicella C, Capasso L

and Paludetto R. Am J Physiol Gastrointest Liver Physiol 2008; 294(4):G906-G913.

25 Raimondi F, Crivaro V, Caspasso L, Maiuri L, Santoro P, Tucci M, Barone MV, Pappacoda S and

Paludetto R. Pediatr Res 2006; 60(1)

26 Kusoffsky E, Strandvik B and Troell S. J Pediatr Gastroenterol Nutr 1983; 2(3):434-438.

27 Chase HP, Cotton EK and Elliot RB. Pediatrics 1979; 64(2):207-213.

28 Landon C, Kerner JA, Castillo R, Adams L, Whalen R and Lewiston NJ. JPEN J Parenter Enteral Nutr

1981; 5(6):501-504.

CHAPTER 6

GELUCIRE®44/14 IMPROVES FAT ABSORPTION IN RATS WITH

IMPAIRED LIPOLYSIS

S. Lukovac1, K.E.G. Gooijert

1, P.C. Gregory

2, G. Shlieout

2, F. Stellaard

1, E.H.H.M.

Rings1, H.J. Verkade

1

(1) Pediatric Gastroenterology, Department of Pediatrics, Beatrix Children’s Hospital,

Groningen University Institute for Drug Exploration (GUIDE), Center for Liver, Digestive

and Metabolic Diseases, University of Groningen, University Medical Center Groningen,

Groningen, The Netherlands.

(2) Solvay Pharmaceuticals GmbH, Hannover, Germany.

Manuscript conditionally accepted for publication in Biochimica et Biophysica

Acta (BBA) Molecular and Cell biology of lipids

CHAPTER 6

112

ABSTRACT

Clinically relevant fat malabsorption is usually due to impaired intestinal fat digestion

(lipolysis) and/or to impaired solubilization of the lipolytic metabolites. We hypothesized

that Gelucire®44/14 –a semi-solid self-micro-emulsifying excipient– could increase fat

absorption. In relevant rat models for impaired lipolysis or for impaired solubilization we

tested whether administration of Gelucire®44/14

enhanced fat absorption. Rats with

impaired lipolysis (lipase inhibitor Orlistat-diet) and rats with reduced solubilization

(permanent bile diversion) underwent a 72h fat balance test to assess fat absorption.

The absorption kinetics of a stable isotope-labeled fatty acid was assessed in rats with

reduced solubilization, in the presence or absence of Gelucire®44/14. Gelucire

®44/14

improved fat absorption in rats with impaired lipolysis (from 70% to 82%, p<0.001). In

rats with reduced solubilization, Gelucire®44/14 did not increase fat absorption nor did it

reconstitute the absorption kinetics of 13

C-labeled palmitate, compared with control rats

administered buffer without Gelucire®44/14.

The present data show that Gelucire®44/14 might enhance fat absorption under

conditions of impaired lipolysis, but not during impaired solubilization. We speculate that,

due to its self-micro-emulsification properties, Gelucire®44/14 stabilizes and improves

residual lipolytic enzyme activity in vivo, which could be of therapeutic value in clinical

conditions of fat malabsorption due to impaired lipolysis.

GELUCIRE®44/14 AND FAT ABSORPTION

113

6

CH

AP

TE

R

INTRODUCTION

Dietary fat is mainly composed of triglycerides, which undergo several intraluminal

processes before their absorption in the form of fatty acids and monoacylglycerides by

the small intestinal enterocytes.1 The first step of fat absorption involves emulsification

and hydrolysis by gastric lipase, which results in partial hydrolysis of triglycerides into

free fatty acids and diglycerides. The remaining partially digested and undigested,

triglycerides are digested by pancreatic lipases in the small intestine, which leads to

lipolysis into free fatty acids and monoglycerides. Subsequent steps involve micellar

solubilization with bile salts, phospholipids and cholesterol, transport to the enterocytes,

and translocation of the fatty acids and monoacylglycerides across the apical brush

border of the enterocytes. Under physiological conditions the pancreas produces

sufficient amounts of pancreatic lipases. However, under conditions of severe pancreatic

insufficiency, lipolysis may be incomplete and fat malabsorption occurs.2

Cystic fibrosis (CF), a common autosomal recessive disorder, is a condition in which

pancreatic secretory function is frequently affected.3 On top of impaired lipolysis, CF

patients often have solubilization defects, which may be due to changes in bile

production and composition, impaired pancreatic and intestinal bicarbonate secretion,

and/or changes in the intestinal microclimate.4,5,6

The combination of impaired lipolysis

and solubilization in CF patients can lead to severely reduced absorption of dietary fats

and to essential fatty acid deficiency.7,8

Several attempts to correct for low fat absorption,

for example with pancreatic enzyme replacement therapies and linoleic acid

supplementations, have shown variable effects in CF patients.9,10,11,12

Gelucire®44/14 is a semi-solid, self-emulsifying excipient frequently used in the

pharmaceutical industry as an enhancer of absorption of poorly soluble and poorly

bioavailable drugs.13,14

Gelucire®44/14 is composed of surfactants (mono- and diesters

of polyethylene glycol), co-surfactants (monoglycerides), and an oily phase (di- and

triglycerides). In vitro, Gelucire®44/14 has been shown to maintain the activity of

pancreatic enzymes under unfavorable conditions at low pH (Patent WO 2005/092370).

Moreover, Fernandez et al. have demonstrated that Gelucire®44/14 is a good substrate

for digestive enzymes.15

However, it remains unclear whether the efficacy of

Gelucire®44/14 is exclusively related to increasing the acid stability of pancreatic lipase,

and thereby to increasing lipolysis of dietary lipids. Alternatively, the mechanism of fat

malabsorption in CF is not exclusively related to impaired lipolysis. Thus it seems

feasible that Gelucire®44/14 might enhance net fat absorption by increasing the

solubilization of the lipolytic metabolites (free fatty acids and monoglycerides) and

thereby improve present CF therapy. Therefore, studies have been performed to test the

effects of Gelucire®44/14 on fat absorption in animals with induced fat malabsorption. In

order to address the potential, specific roles of Gelucire®44/14 in lipolysis and

solubilization of fat in vivo, we used validated rat models for either impaired lipolysis or

for severely reduced solubilization. Rats fed the lipase-inhibitor Orlistat (Xenical®) have a

selectively inhibited hydrolysis of dietary triglycerides, but unaffected

solubilization.16,17,18,19

On the other hand, rats with permanent bile diversion (BDD rats)

are a well characterized model to assess fatty acid uptake under condition of

(exclusively) reduced solubilization.18,20

CHAPTER 6

114

MATERIAL AND METHODS

Compounds

Taurocholate, phosphatidylcholine and cholesterol were purchased by Sigma Chemical

(St. Louis, USA). 13

C-labeled palmitic acid (C16:0) was purchased from Isotec Inc.

(Matheson, USA). Gelucire®44/14 was a generous gift from Solvay Pharmaceuticals

GmbH (Hannover, Germany). Orlistat (tetrahydrolipstatin, Xenical®) was obtained as

capsules containing 120 mg active compound from Roche Nederland B.V. (Mijndrecht,

The Netherlands).

Animals and diets

Male Wistar rats (Harlan, Zeist, The Netherlands), weighing 300-350 g, were housed in a

light-controlled (lights on 7 AM - 7 PM) and temperature-controlled facility with free

access to food and tap water, and, in the case of bile-diverted rats, saline (0.9% NaCl

w/v). The experimental protocol was approved by the Ethics Committee for Animal

Experiments, Faculty of Medical Sciences, University of Groningen, The Netherlands.

A semi-synthetic high-fat diet containing 16 weight% fat (4141.07) and the same diet

containing Orlistat (4141.07 + 200 mg/kg Orlistat) were produced by Arie Blok BV

(Woerden, The Netherlands). The diet contained 35 energy% fat and 16.2 wt% long-

chain fatty acids (fatty acid composition (in mol%): palmitic acid (C16:0), 39.0%; stearic

acid (C18:0), 4.0%; oleic acid (C18:1n-9), 31.7%; linoleic acid (C18:2n-6), 22.9%).

Gelucire®44/14 (1 wt% or 2 wt%) was mixed into the semi-synthetic high-fat diet or into

the semi-synthetic high-fat diet containing Orlistat.

Infusates and intraduodenal infusions

Infusates and bolus were prepared as described previously.20

Buffer contained 10 mM

HEPES and 135 mM NaCl (negative control). Model bile contained 60 mM taurocholate,

8 mM phosphatidyl choline and 1 mM cholesterol (positive control). Gelucire®44/14-

infusates contained buffer with 0.1% or 0.5% Gelucire®44/14.

Bolus (500 μl) was administered intraduodenally and composed of olive oil (25%),

medium chain triglyceride oil (75%; (composed of extracted coconut oil and synthetic

triglycerides; fatty acid composition: 6:0, 2%; 8:0, 50-65% max.;10:0, 30-45%; 12:0, 3%

max.) and 10 mg of 13

C-labeled palmitic acid (> 99% enriched) per 300 g body.20,18

Fat balance study in rats with impaired lipolysis

After a run-in period of two weeks on the semi-synthetic high-fat diet (4141.07) the fat

absorption was assessed during a 72 hours period in individually housed rats.

Subsequently, the rats fed the Orlistat containing diet for two weeks. At the end of the

two weeks, fat absorption test was performed again. Consistent with previous studies

with Orlistat feeding in rats, two weeks of Orlistat-diet (200 mg/kg) was sufficient to

decrease the net fat absorption.17

Next, one group of rats received the Orlistat-diet with

additional 1 wt% Gelucire®44/14 and another group of rats received the Orlistat-diet with

2 wt% Gelucire®44/14 for one additional week. At the end of the experimental week food

intake was determined and feces were collected for the assessment of the fat absorption

upon Gelucire®44/14 feeding. Net fat absorption was determined by measuring the fatty

GELUCIRE®44/14 AND FAT ABSORPTION

115

6

CH

AP

TE

R

acid intake and excretion by fatty acid methylation as described in section “Analytical

techniques”. The experimental set up is shown in detail in Figure 1a.

Fat balance study in rats with reduced solubilization

Rats were individually housed and were fed a semi-synthetic high-fat diet (4141.07) for a

run-in period of 1 week. At the end of the first week, a 72h fat balance was performed

(intact EHC; control situation). During this 72h period, feces were collected and chow

intake was measured. Subsequently, rats received a permanent bile duct catheter, by

means of a surgical procedure as previously described by Kuipers et al.21

After the

recovery period of three days, the rats were fed the same diet as before the surgery and

after one week of diet 72h fat balance was performed (BDD; interrupted EHC).

Afterwards, rats with BDD were separated into 2 different groups receiving the same diet

as before the surgery supplemented with either 1 wt% or 2 wt% Gelucire®44/14, and

after two weeks of diet fat balance was determined (BDD 1% Gelucire®44/14 and BDD

2% Gelucire®44/14). Subsequently, the rats who first received 1% concentration were

fed 2% concentrations of Gelucire®44/14, and vice versa, for additional two weeks. A

72h fat balance was repeated at the end of these two last weeks of diet. The described

72h fat balances were performed in each rat individually. The experimental set up is

indicated in detail in Figure 1b.

Figure 1 Experimental set up of the experiments performed to determine the effects of Gelucire

®44/14 on fat absorption in vivo. (a) Experimental scheme of the fat balance study in rats

with impaired lipolysis. (b) Experimental scheme of the fat balance study in rats with reduced solubilization. (c) Experimental scheme of the kinetics experiment of fat absorption in rats with reduced solubilization.

CHAPTER 6

116

Kinetics of fat absorption in rats with reduced solubilization

Rats were fed standard chow and received permanent catheters in bile duct and

duodenum as previously described by Kuipers et al.21

After the surgery, bile duct and

duodenum catheters were connected with each other at the skull of the rat for at least

three days to restore the enterohepatic circulation, in order to allow the rats to recover

from surgery. Subsequently, catheters of bile duct and duodenum were chronically

interrupted, resulting in permanent intestinal bile-deficiency. On the day of the

experiment, rats were infused intraduodenally for 7 hours (flow rate 1.5 ml/h) with buffer

(negative control), model bile (positive control), and with Gelucire®44/14-buffer

containing 0.1% or 0.5% Gelucire®44/14. The infusion rate and concentrations of bile

components were selected to reflect the physiological rates of bile flow and of the

intestinal delivery of specific bile components in adult Wistar rats. After starting the

intraduodenal infusion, 500 μl of fat per 300 gram body weight was administered slowly

as a bolus (olive oil, medium chain triglyceride oil and 13

C-labeled palmitic acid) via the

intraduodenal catheter. Medium chain triglyceride oil was included in the bolus in order to

obtain a reliable, reproducible vehicle for the quantitative administration of the labelled

compound, without introducing a profound increase in the intake of long-chain fatty

acids.22

The fat bolus represented approximately 15% of the daily fat intake of the semi-

synthetic high-fat diet. Blood samples (approximately 200 μl) were taken from the tail

vein at base line and every hour for 6 hours after administration of the fat bolus. The

baseline sample was taken prior to the administration of the fat bolus. Plasma and

erythrocytes were separated by centrifugation (2000 rpm, 10 min at 4˚C) and afterwards

stored at -20˚C until further analysis. Rats were used as their own controls during the

experiment, which was performed four times during the two weeks subsequent to the

chronic interruption of the bile duct (with different infusates); the above described

intraduodenal infusion conditions were performed in each rat individually. The

experimental set up is shown in detail in Figure 1c.

Analytical methods

Fatty acid analysis in chow and feces

Feces and chow were freeze-dried and homogenized mechanically. From aliquots of

feces and chow, lipids were extracted, hydrolyzed and methylated according to Muskiet

et al.23

Resulting fatty acid methyl esters were analyzed by gas chromatography to

calculate ingestion and fecal excretion of major fatty acids. Fatty acids were quantified

using heptadecanoic acid (C17:0) as internal standard. Total fecal fat excretion was

calculated from the daily fat intake and the daily fecal fat excretion and expressed as a

percentage of the daily fat intake as indicated in the following formula:

Fat intake (g day-1

) – Fecal fat output (g day-1

)

Percentage of total fat absorption = x100%

Fat intake (g day-1

)

Plasma lipids

Total lipids of plasma samples were extracted, hydrolyzed and methylated for gas-

chromatographic analysis of fatty acid profile as described by Muskiet et al.23

13

C

enrichment of fatty acid methyl esters was determined on a gas chromatography

GELUCIRE®44/14 AND FAT ABSORPTION

117

6

CH

AP

TE

R

combustion isotope ratio mass spectrometer (GC-C-IRMS). The concentration of 13

C

fatty acid in plasma at each time point was calculated from the fatty acid concentration

and 13

C enrichment and expressed as the percentage of the dose administered per ml

plasma (%dose/ml).20

Statistical analysis

Statistical analyses were performed using analysis of variance (One-Way ANOVA)

followed by post-hoc analysis (Bonferroni) using the SPSS version 12.0.2 software

(Chicago, IL, USA). For all experiments, p-values below 0.05 were considered

statistically significant.

RESULTS

After bile diversion, rats transiently lost up to 10% of their body weight. However, body

weights returned to normal within days (data not shown). There was no significant

difference in body weight between rats fed Orlistat and rats fed control high fat diet, or

between rats fed Orlistat and rats fed Orlistat with additional Gelucire®44/14 (data not

shown).

Dietary Gelucire®44/14 increases food intake in rats with reduced lipolysis and

solubilization, but increases fecal fat excretion exclusively in rats with reduced

solubilization

Food ingestion and feces production slightly increased in rats fed Orlistat, which is

consistent with previous findings in Gunn rats fed Orlistat.17

Feces production was

approximately 14% lower in rats with impaired lipolysis fed 2% Gelucire®44/14 diet

compared to rats fed Orlistat diet alone, but the difference did not reach statistical

significance (Figure 2b, NS). Gelucire®44/14 (1% and 2%) significantly increased food

ingestion in rats with impaired lipolysis compared to rats fed Orlistat alone (+23 and

+19%, respectively, each p<0.001, Figure 2a).

As expected, bile diversion significantly enhanced the amount of feces produced per day

by 76% compared to the condition in the same rats before bile diversion (p<0.001,

Figure 2d). Comparable to Orlistat-treated rats, Gelucire®44/14 increased food intake in

bile diverted rats (by 35% and 41%, in 1% and 2% Gelucire®44/14 diet fed rats,

respectively; each p<0.001, Figure 2c). Simultaneously, Gelucire®44/14 (both doses)

increased feces production by ~30%, compared to bile diverted rats fed the control diet

(p< 0.05, Figure 2d).

Dietary Gelucire®44/14 enhances absolute absorption of fat in rats with impaired

lipolysis, and to a lesser extent in rats with permanent bile diversion

The ingestion of fatty acids in absolute terms (mmol/day) was unchanged upon bile

diversion, but excretion was significantly increased, resulting in a lower absolute amount

of fat absorbed (Table 1). Inclusion of 1% or 2% Gelucire®44/14 in the diet increased the

absolute amount of fatty acids excreted in these rats (Table 1). Therefore, the absolute

amount of fat absorbed per day in rats with bile diversion fed Gelucire®44/14 was slightly

increased, but still significantly lower than in rats with intact enterohepatic circulation.

CHAPTER 6

118

Orlistat feeding increased absolute amounts of fatty acids ingested in two weeks (Table

2). Gelucire®44/14 decreased the amounts of fatty acids excreted in the feces of rats

with impaired lipolysis (Table 2). Accordingly, Gelucire®44/14 (both doses) increased the

absolute amounts of fatty acids absorbed daily in rats with impaired lipolysis, even to a

higher level than observed in rats with normal lipolysis (Table 2).

Figure 2 Data are represented as means ± SD of (a, b) 6-18 rats per group for impaired lipolysis study and (c, d) 5-9 rats per group in impaired solubilization study. (a, b) Significant difference in (a) ingestion and (b) excretion between control rats on high-fat diet () versus rats fed Orlistat ( for two weeks) or Orlistat with Gelucire

®44/14 ( 1% and 2%) is indicated as * p<0.05. Significant

difference between rats fed Orlistat diet with additional Gelucire®44/14 ( 1% and 2%) versus

rats fed only Orlistat ( for two weeks) is indicated as #p<0.05. (c, d) Significant difference in ingestion (c) and excretion (d) between control rats with intact EHC () versus bile diverted rats on control diet (), bile diverted rats on 1% Gelucire

®44/14 diet () or 2% Gelucire

®44/14 diet () is

indicated as *p<0.05. Significant difference between bile diverted rats on control diet () versus bile diverted rats on Gelucire

®44/14 diets ( 1% or 2%) is indicated as #p<0.05.

GELUCIRE®44/14 AND FAT ABSORPTION

119

6

CH

AP

TE

R

Gelucire®44/14 (2%) increases net fat absorption in rats with impaired lipolysis,

but has no effect on net fat absorption in rats with permanent bile diversion

We analyzed the relative fat absorption from the fat balance. Orlistat administration significantly decreased net fat absorption in rats already after one week from 88% in control rats to approximately 70%, attributed to reduced lipolytic activity of lipases (data not shown). An additional week of Orlistat feeding did not further reduce net fat absorption (Figure 3a), neither did an additional 2 weeks (data not shown). This was in agreement with previous findings in Gunn rats.

17

Table 1 Absolute dietary fat ingestion, excretion and absorption in rats with intact EHC (control), bile diverted rats on control diet (BDD) and bile diverted rats on diet supplemented with 1% or 2% Gelucire

®44/14 (BDD 1% Gelucire

®44/14 and BDD 2% Gelucire

®44/14, respectively). Data are

means ± SD of 5-9 rats per group. Mean values represent the average of 72h per rat. *p<0.05 versus control rats with intact EHC. #p<0.05 versus BDD rats on control diet.

The 1% dose of Gelucire®44/14 did not significantly affect net fat absorption in rats fed

Orlistat (72%, NS), but 2% Gelucire®44/14 significantly increased the net fat absorption

reaching close to physiological values (82%, p<0.001, Figure 3a). In accordance with

previous observations, bile diversion lowered net fat absorption to 45% (Figure 3b).22

However, in these rats net fat absorption was not significantly altered by either dosage of

Gelucire®44/14 compared to fat absorption in the same rats on control diet without

Gelucire®44/14 (both 52%, NS, Figure 3b).

Table 2 Absolute dietary fat ingestion, excretion and absorption in control rats, rats fed control diet with Orlistat for one week or two weeks, and rats fed control diet with Orlistat for three weeks with additional Gelucire

®44/14 during the last week of treatment. Data are means ± SD of 6-18 rats per

group. Mean values represent the average of 72h per rat. *p<0.05 versus rats on control diet. #p<0.05 versus rats on Orlistat diet for two weeks.

CHAPTER 6

120

Gelucire®44/14 increases absorption of saturated and unsaturated fatty acids in

rats with reduced lipolysis, but does not affect absorption of fatty acids in rats

with impaired solubilization

In order to determine whether the positive effects of Gelucire®44/14 on fat absorption

were selective for specific fatty acid species, we assessed the net fat absorption of the

four major dietary fatty acids (linoleic, oleic, stearic and palmitic acid). Both impaired

lipolysis (Orlistat feeding, Figure 4a, 4b) and impaired solubilization (bile diversion,

Figure 4c, 4d) reduced the net absorption of all of the major fatty acids.

Figure 3 The effect of Gelucire

®44/14 on total net fat absorption in rats with (a) impaired lipolysis

and (b) impaired solubilization. Data are represented as means ± SD of (a) 6-18 rats per group for impaired lipolysis study and (b) 5-9 rats per group in impaired solubilization study. (a) Significant difference between control rats on high-fat diet () versus rats fed Orlistat ( for two weeks), rats fed Orlistat diet with additional Orlistat with 1% Gelucire

®44/14 () or 2% Gelucire

®44/14 () is

indicated as *p<0.05. Significant difference between rats fed Orlistat diet with additional 1% Gelucire

®44/14 () or 2% Gelucire

®44/14 () versus rats fed only Orlistat ( for two weeks) is

indicated as #p<0.05. (b) Significant difference between control rats with intact EHC () versus bile diverted rats on control diet (), bile diverted rats on 1% Gelucire

®44/14 diet () or 2%

Gelucire®44/14 diet () is indicated as *p<0.05. No significant difference was found between bile

diverted rats fed Orlistat diet with additional Gelucire®44/14 and bile diverted rats fed Orlistat diet

without Gelucire®44/14.

In rats with impaired lipolysis, Gelucire

®44/14 dose-dependently improved the uptake of

saturated fatty acids (Figure 4b). The absorption of unsaturated fatty acids was also

increased, but only in rats fed 2% Gelucire®44/14 diet (Figure 4a). In rats with permanent

bile diversion neither of the two Gelucire®44/14 dosages significantly affected the

absorption of unsaturated fatty acids (Figure 4c, 4d).

Intraduodenal administered Gelucire®44/14 does not reconstitute plasma

appearance of 13

C-labeled palmitic acid in rats with permanent bile diversion

Gelucire®44/14 caused a weak but significant increase in absorption of saturated fatty

acids in rats with permanent bile diversion. We therefore determined whether

Gelucire®44/14 affected the kinetics of fat absorption during impaired solubilization. We

assessed the absorption of 13

C-labeled palmitic acid for six hours after its intraduodenal

GELUCIRE®44/14 AND FAT ABSORPTION

121

6

CH

AP

TE

R

administration in rats with permanent bile diversion. Figure 4a shows the time course of

plasma enrichment of 13

C-labeled palmitic acid after intraduodenal administration of a

bolus with buffer (negative control), model bile (positive control) or buffer supplemented

with 0.1 or 0.5% Gelucire®44/14. Bile diverted rats infused with model bile had a

significantly increased plasma concentration of 13

C-labeled palmitic acid at all time points

after the administration (Figure 5a). The plasma appearance of 13

C-labeled palmitic acid

Figure 4 The effects of Gelucire

®44/14 on net fat absorption of the major dietary (a, c) unsaturated

linoleic and oleic fatty acids and (b, d) saturated stearic and palmitic fatty acids in (a, b) rats with impaired lipolysis and (c, d) rats with impaired solubilization. Data are represented as means ± SD of (a, b) 6-18 rats per group for impaired lipolysis study and (c, d) 5-9 rats per group in impaired solubilization study. (a, b) Significant difference between control rats on high-fat diet () versus rats fed Orlistat ( for two weeks), rats fed Orlistat diet with 1% Gelucire

®44/14 () or 2%

Gelucire®44/14 () is indicated as *p<0.05. Significant difference between rats fed Orlistat diet with

additional 1% () or 2% Gelucire®44/14 () versus rats fed only Orlistat ( for two weeks) is

indicated as #p<0.05. (c, d) Significant difference between control rats with intact EHC () versus bile diverted rats on control diet (), bile diverted rats on 1% Gelucire

®44/14 diet () or 2%

Gelucire®44/14 diet () is indicated as *p<0.05. (c) No significant difference was found between bile

diverted rats on control diet versus bile diverted rats fed with additional 1% or 2% Gelucire®44/14.

(d) Significant difference between bile diverted rats on 2% Gelucire®44/14 () versus bile diverted

rats on control diet () is indicated as #p<0.05.

CHAPTER 6

122

did not differ between bile diverted rats infused with buffer only or with buffer

supplemented with 0.1 or 0.5% Gelucire®44/14 (Figure 5a). Figure 5b shows area under

the curves of Figure 5a which reflect the amounts of 13

C-labeled palmitic acid absorbed

during the six hours after administration of the bolus. In accordance with the

observations in Figure 5a, bile diverted rats receiving model bile infusion had a higher

area under the curve than bile diverted rats administered buffer alone (Figure 5b).

Intraduodenal administration of buffer supplemented with either 0.1% or 0.5%

Gelucire®44/14 did not increase the amount of the label absorbed (Figure 5b).

Figure 5 The effect of Gelucire

®44/14 on kinetics of fat absorption in rats with impaired

solubilization. (a) Curves of plasma concentrations of 13

C-palmitic acid during 6 hours after administration and (b) cumulative areas under the curves in rats with permanent bile diversion administered an intraduodenal fat bolus containing

13C-palmitic acid in combination with buffer ()

as negative control, 0.1% Gelucire®44/14 in buffer (), 0.5% Gelucire

®44/14 in buffer () or model

bile as positive control (). Data are represented as means ± SEM of 3-6 rats per group. Significant difference between bile diverted rats administered model bile and bile diverted rats administered buffer, 0.1% Gelucire

®44/14 in buffer or 0.5% Gelucire

®44/14 in buffer is indicated as *p<0.05. No

significant difference was found between BDD rats administered buffer with Gelucire®44/14 and

buffer alone.

DISCUSSION

Novel role for Gelucire®44/14 in fat absorption

Previous studies revealed a role for Gelucire®44/14 as an efficient emulsifier for

improvement of dissolution and absorption of poorly water soluble drugs.15,24,25

Hamid et

al. showed by means of an in vitro diffusion chamber and an in situ closed-loop

technique using rat intestinal tissue, that the use of Gelucire®44/14 can be considered

safe and that it does not lead to any intestinal membrane damage.26

Recently,

Fernandez et al. described in vitro effects of different lipases on Gelucire®44/14 and

showed that in particular gastric lipase was able to lipolyse Gelucire®44/14.

13 However, it

was not clear whether Gelucire®44/14 in turn appears to exert effects on lipases, namely

to enhance the lipolytic activity of lipases. We now show that Gelucire®44/14 might

improve the absorption of fatty acids in vivo under conditions of impaired lipolysis. These

findings could be of value for improving the nutritional status of patients with reduced

activity of pancreatic lipase, e.g. due to cystic fibrosis or chronic pancreatitis. Our data

GELUCIRE®44/14 AND FAT ABSORPTION

123

6

CH

AP

TE

R

show that Gelucire®44/14 might act as an enhancer of lipolysis, rather than a solubilizer

under conditions of impaired fat malabsorption.

Dietary Gelucire®44/14 as enhancer of impaired lipolysis

In accordance with previous studies, our rat models of impaired lipolysis and impaired

solubilization show signs of fat malabsorption.17,18,22

Rats fed Orlistat show more

preserved fat absorption (71% after one week and 70% after two weeks of Orlistat

feeding) compared with rats with reduced solubilization (45% two weeks after the

operation). These percentages of net absorption coefficients are in agreement with

previously published data in similar rat models.17,18,27

However, rats fed Orlistat show

lower levels of inhibition of fat absorption (20% reduction) compared with human subject

who received lower amount of Orlistat (120 mg per meal; 30-40% reduction in fat

absorption).28,29,30

It appears that human subject received 0.59% of Orlistat per grams of

fat ingested, while the rats in our study received 0.89% of Orlistat per gram of fat

ingested. Present and other studies performed in rats fed Orlistat show that rats seem to

compensate the fecal fat loss by increasing their food intake during Orlistat feeding.17,19

To our knowledge human subjects fed Orlistat do not compensate for fecal fat loss by

increased food intake and, therefore, these observations might explain the discrepancy

between the effects of Orlistat fat absorption in rats and humans.

In vitro studies of Subramanian and Wasan suggested that Gelucire®44/14 might inhibit,

rather than improve, lipolytic activity.31

However, these data are not conclusive since no

significant difference was found in lipolytic activity between untreated lipases and lipases

treated with increasing concentrations of Gelucire®44/14.

31

One could have anticipated that Gelucire®44/14 would have had the largest effect on fat

absorption in bile diverted rats in which the fat malabsorption was most severely

affected. However, this would only be the case if the effects of Gelucire®44/14 could

restore (to some extent) the solubilization. Our data show that Gelucire®44/14 improves

total net fat absorption to a much larger extent in rats with impaired lipolysis; indeed, the

coefficient of fat absorption returned to almost normal values after one week of 2%

Gelucire®44/14 diet in Orlistat fed rats. In contrast, there was no effect of Gelucire

®44/14

on the solubilization of unsaturated fatty acids and only a minimal effect on the

solubilization of saturated fatty acids in rats with bile diversion. This suggests that

Gelucire®44/14 has only a slight effect on fat absorption under conditions of fat

malabsorption exclusively due to impaired solubilization.

Both rat models showed increased food ingestion of Gelucire®44/14 diets, while rats with

bile diversion also showed increased feces production upon Gelucire®44/14 feeding. The

underlying mechanism of the increased food intake by Gelucire®44/14 in these ad libitum

fed rats remains to be elucidated. We cannot exclude the possibility that difference in the

texture of the granules (which were manually made for Gelucire®44/14 diets and custom

made for the other diets) contribute to increased food ingestion in rats fed

Gelucire®44/14 diets. It has been described previously by Sako et al. that texture of food

plays an important role in food selection behavior in rats.32

Another possibility would be

that pre-digested lipolysis products (monoglycerides) directly derived from the ingested

Gelucire®44/14 might have a positive effect on food intake in these animals. The

elucidation is relevant for several reasons, including an explanation for the observed

CHAPTER 6

124

increase in net absorption of fatty acids by Gelucire®44/14. Gelucire

®44/14 may directly

enhance the activity of pancreatic lipases on fat digestion in the small intestine or retard

their degradation, for example by pancreatic proteases. However, it is unlikely that an

increase in food ingestion due to softer food pellets is the main cause of increased fat

absorption. Food intake was similarly increased during impaired lipolysis and reduced

solubilization, while the net absorption of fat was only enhanced under conditions of

impaired lipolysis and moreover was not improved in rats fed Orlistat with 1%

Gelucire®44/14. Therefore other factors must be responsible for enhanced absorption of

fat in rats with impaired lipolysis. It is possible that due to the emulsification properties of

Gelucire®44/14 an increase in the specific surface area of the fat could enable the lipase

to be more effective in vivo.

To expand our insights in the (possible) effects of Gelucire®44/14 on the kinetics of fat

absorption, we additionally measured the absorption kinetics of the saturated fatty acid

palmitate. Direct duodenal infusion of Gelucire®44/14 did not result in increased plasma

appearance of 13

C-labelled palmitate, indicating that there is no significant effect of the

compound on the kinetics of palmitate absorption. The ratio of Gelucire®44/14 to the total

amount of fat administered was equivalent to the ratio of Gelucire®44/14 to the total

amount of fat ingested in diet so that it is unlikely that the concentration of

Gelucire®44/14 infused was too low to exert an effect. The possibility remains however,

that by intraduodenal administration of the bolus, preduodenal lipolysis of Gelucire®44/14

is bypassed and that this limited its biological activity.

The exact reason for the higher specificity of Gelucire®44/14 on saturated fat compared

to unsaturated fatty acid absorption remains to be elucidated, but it seems that there

could be an effect on both solubilization and lipolysis with saturated fats but only an

effect on lipolysis for unsaturated fats. We cannot exclude, however, that some of the

absorbed palmitate and stearate in rats fed Orlistat with additional Gelucire®44/14 are

derived directly from Gelucire®44/14 itself. Fatty acids within Gelucire

®44/14 are mainly

incorporated within monoglycerides and are absorbed independently by the lipase

activity. Since absolute absorption rates of palmitic and stearic fatty acids in rats fed

Orlistat and Gelucire®44/14 exceed the amount of fatty acids that is present within

Gelucire®44/14 (data not shown), we expect that the contribution of increased fatty acid

absorption directly by fatty acids derived from Gelucire®44/14 is very low.

Future studies on the effects of Gelucire®44/14 on fat absorption

The main focus of the present study was to determine whether dietary supplementation

of Gelucire®44/14 enhances the absorption of fatty acids in vivo in relevant rat models for

different types of fat malabsorption. The underlying mechanisms might include direct

effect on lipolysis leading to increased fat absorption or indirect improvement of fat

absorption by means of enhanced emulsification properties of fat absorption. Our

findings support the concept that Gelucire®44/14 indeed enhances fat absorption in rats

in vivo. Concerning the improved absorption of saturated palmitic (C16:0) and stearic

(C18:0) acids, it would be interesting to measure the solubility of saturated and

unsaturated fatty acids in Gelucire®44/14. It is possible that saturated fatty acid have a

better solubility in a hydrogenated vegetable oil, which is the excipient in Gelucire®44/14.

Furthermore, it would be interesting to perform similar studies as presented here in a

GELUCIRE®44/14 AND FAT ABSORPTION

125

6

CH

AP

TE

R

mouse model for cystic fibrosis, where both lipolysis and solubilization are impaired and

lead to fat malabsorption. We have shown that there seems to be a dose dependent

effect of Gelucire®44/14 on fat absorption in rats with impaired lipolysis. It would be

interesting to further characterize this dose dependency using different diets and animal

models. Moreover, since unsaturated fatty acids may undergo microbial hydrogenation in

the large intestine,33

future studies would also measure absorption of fatty acids at the

level of the terminal ileum to clarify the differential effects of Gelucire®44/14 on

absorption of saturated and unsaturated fatty acids.

Overall conclusion

Dietary supplementation of Gelucire®44/14 to rats with impaired lipolytic activity corrects

the net total fat absorption. If Gelucire®44/14 would similarly improve fat absorption in

patients with impaired lipolysis, such as in CF patients or patients with chronic

pancreatitis, it could constitute a major improvement in the current therapy. However,

essential fatty acid deficiency during CF may remain prominent, even under dietary

Gelucire®44/14 supplementation, since this compound mainly improves the absorption of

saturated fatty acids. Studies in e.g. a mouse model of CF should reveal if

Gelucire®44/14 can normalize fat absorption and can be used in combination with other

compounds to improve the absorption of saturated, but also of essential fatty acids in

clinical conditions of pancreas insufficiency.

ACKNOWLEDGEMENTS

The authors would like to thank Rick Havinga for his excellent technical assistance

during the studies in bile diverted rats. Furthermore, we would like to thank Theo Boer

for his technical assistance during GC-C-IRMS measurements.

GRANTS

Part of this study was supported by an unrestricted grant of Solvay Pharmaceuticals

GmbH (Hannover, Germany) and by the Dutch Digestive Foundation.

REFERENCES

1 Carey MC, Small DM, and Bliss CM. Annu Rev Physiol 1983; 45: 651-677.

2 Bliss CM. Arch Intern Med 1981; 141(1213): 1215.

3 Quinton PM. Lancet 2008; 372(9636): 415-417.

4 Strandvik B, Einarsson K, Lindblad A, and Angelin B. J Hepatol 1996; 25(1): 43-48.

5 Roy CC, Weber AM, Morin CL, Combes JC, Nussle D, Megevand A, and Lasalle R. N Engl J Med 1977;

297(24): 1301-1305.

6 Shumaker H, Amlal H, Frizzell R, Ulrich CD, and Soleimani M. Am J Physiol 1999; 276(1 Pt 1): C16-C25.

7 Rivers JP and Hassam AG. Lancet 1975; 2(7936): 642-643.

8 Innis SM and Davidson AG. Annu Rev Nutr 2008; 28: 55-72.

9 Symonds EL, Omari TI, Webster JM, Davidson GP, and Butler RN. J Pediatr 2003; 143(6): 772-775.

10 Meyer JH, Elashoff JD, and Lake R. Dig Dis Sci 1999; 44(6): 1076-1082.

11 Kusoffsky E, Strandvik B, and Troell S. J Pediatr Gastroenterol Nutr 1983; 2(3): 434-438.

CHAPTER 6

126

12 Bronstein MN, Sokol RJ, Abman SH, Chatfield BA, Hammond KB, Hambidge KM, Stall CD, and Accurso FJ.

J Pediatr 1992; 120(4 Pt 1): 533-540.

13 Fernandez S, Rodier JD, Ritter N, Mahler B, Demarne F, Carriere F, and Jannin V. Biochim Biophys Acta

2008; 1781(8): 367-375.

14 Yuksel N, Karatas A, Ozkan Y, Savaser A, Ozkan SA, and Baykara T. Eur J Pharm Biopharm 2003; 56(3):

453-459.

15 Fernandez S, Chevrier S, Ritter N, Mahler B, Demarne F, Carriere F, and Jannin V. Pharm Res 2009; 26(8):

1901-1910.

16 Henness S and Perry CM. Drugs 2006; 66(12): 1625-1656.

17 Nishioka T, Hafkamp AM, Havinga R, vn Lierop PP, Velvis H, and Verkade HJ. J Pediatr 2003; 143(3): 327-

334.

18 Kalivianakis M, Minich DM, Havinga R, Kuipers F, Stellaard F, Vonk RJ, and Verkade HJ. Am J Clin Nutr

2000; 72(1): 174-180.

19 Hafkamp AM, Havinga R, Ostrow JD, Tiribelli C, Pascolo L, Sinaasappel M, and Verkade HJ. Pediatr Res

2006; 59(4 Pt 1): 506-512.

20 Nishioka T, Having R, Tazuma S, Stellaard F, Kuipers F, and Verkade HJ. Biochim Biophys Acta 2004;

1636(2-3): 90-98.

21 Kuipers F, Havinga R, Bosschieter H, Toorop GP, Hindriks FR, and Vonk RJ. Gastroenterology 1985; 88(2):

403-411.

22 Minich DM, Kalivianakis M, Havinga R, Van GH, Stellaard F, Vonk RJ, Kuipers F, and Verkade HJ. Biochim

Biophys Acta 1999; 1438(1): 111-119.

23 Muskiet FA, van Doormaal JJ, Martini IA, Wolthers BG, and van der Slik W. J Chromatogr 1983; 278(2):

231-244.

24 Damian F, Blaton N, Naesens L, Balzarini J, Kinget R, Augustijns P, and Van den Mooter G. Eur J Pharm

Sci 2000; 10(4): 311-322.

25 Barker SA, Yap SP, Yuen KH, McCoy CP, Murphy JR, and Craig DQ. J Control Release 2003; 91(3): 477-

488.

26 Hamid KA, Katsumi H, Sakane T, and Yamamoto A. Int J Pharm 2009; 379(1): 100-108.

27 Ferraz RR, Tiselius HG, and Heilberg IP. Kidney Int 2004; 66(2): 676-682.

28 Drent ML, Larsson I, William-Olsson T, Quaade F, Czubayko F, von BK, Strobel W, Sjostrom L, and van der

Veen EA. Int J Obes Relat Metab Disord 1995; 19(4): 221-226.

29 Davidson MH, Hauptman J, DiGirolamo M, Foreyt JP, Halsted CH, Heber D, Heimburger DC, Lucas CP,

Robbins DC, Chung J, and Heymsfield SB. JAMA 1999; 281(3): 235-242.

30 Carriere F, Renou C, Ransac S, Lopez V, De CJ, Ferrato F, De CA, Fleury A, Sanwald-Ducray P, Lengsfeld

H, Beglinger C, Hadvary P, Verger R, and Laugier R. Am J Physiol Gastrointest Liver Physiol 2001; 281(1):

G16-G28.

31 Subramanian R and Wasan KM. Drug Dev Ind Pharm 2003; 29(8): 885-890.

32 Sako N, Okamoto K, Mori T, and Yamamoto T. Behav Brain Res 2002; 133(2): 377-382.

33 Jorgensen H, Gabert VM, Hedemann MS, and Jensen SK. J Nutr 2000; 130(4): 852-857.

CHAPTER 7

SUMMARY AND FUTURE PERSPECTIVES

S. Lukovac

CHAPTER 7

130

SUMMARY CONTENTS

SUMMARY 131

Essential fatty acid deficiency in mice impairs lactose digestion 131

and alters jejunal cholesterol metabolism

Essential fatty acid deficiency in mice leads to enhanced ileal bile 133

salt reabsorption and to persistent hepatic bile salt synthesis in

mice

In vitro model of essential fatty acid deficiency reveals increased 134

permeability and impaired mRNA expression of brush border

enterocyte markers, which are not rapidly reversible by linoleic

acid (LA) supplementation

Gelucire®44/14 improves fat absorption in rats with impaired 135

lipolysis

OVERALL CONCLUSION AND IMPLICATIONS 135

REFERENCES 136

SUMMARY AND FUTURE PERSPECTIVES

131

7 C

HA

PT

ER

SUMMARY

The experiments described in this thesis aimed to characterize and unravel the effects of

essential fatty acid (EFA) deficiency on the function of the small intestine. EFA deficiency

is common in pediatric patients with cholestasis (characterized by decreased or absent

hepatic secretion of bile into the intestine), where it is associated with fat malabsorption

and severely impaired nutritional status. Often, EFA deficiency aggravates the

cholestasis induced failure to thrive (CIFTT) in pediatric patients. In our laboratory, a

mouse model for EFA deficiency has been developed. Previous studies in this model

focused on the absorption and metabolism of EFA in hepatic disorders.1,2

In this mouse

model, as well as in other species, it has become apparent that EFA deficiency likely

affects the small intestine.1,3,4,5

However, detailed information about the effects of EFA

deficiency on the small intestinal morphology and function had remained scarce. In order

to improve the nutritional status of pediatric patients encountering EFA deficiency,

improvement of the intestinal function is essential. Therefore, we studied in detail several

effects of EFA deficiency on the small intestinal function in the mouse model for EFA

deficiency. We analyzed the effects of EFA deficiency on the absorption and digestion of

several nutrients and on the enterohepatic circulation of bile salts by means of a stable

isotope dilution technique. We determined the effects of EFA deficiency on the intestinal

morphology in vivo and in vitro. Finally, we performed experiments in vitro to determine

the intracellular effects of EFA deficiency and to analyze whether some of the symptoms

of EFA deficiency can be reversed. Increased knowledge on the role of the small

intestine in EFA deficiency could lead to a development of improved nutritional therapies

in pediatric patients with CIFFT awaiting liver transplantation.

Essential fatty acid deficiency in mice impairs lactose digestion and alters jejunal

cholesterol metabolism

Previous studies in rat and mouse models of EFA deficiency revealed that EFA

deficiency by itself leads to fat malabsorption, even in absence of cholestasis. However,

the effects of EFA deficiency on digestion and absorption of other (dietary) compounds

remained unclear. Theoretically, if EFA deficiency is associated with overall impaired

small intestinal function, not only the absorption of fat but also that of other nutrients, like

carbohydrates and cholesterol, could be expected to be decreased during EFA

deficiency. A general absorptive defect would have consequences for the nutritional

treatment of these conditions.

First, we assessed the capacity of EFA-deficient mice to digest and absorb

carbohydrates, using stable isotope methodology and administration of the disaccharide

[1-13

C]lactose and the monosaccharide [U-13

C]glucose (chapter 2). While the absorption

of the monosaccharide glucose was unaffected in EFA-deficient mice, digestion of the

disaccharide lactose was significantly delayed. The functional observation corresponded

with severely reduced mRNA expression and enzyme activity of the hydrolyzing enzyme

of lactose, lactase, in the small intestine of EFA-deficient mice. These data underscored

the observation that EFA deficiency functionally impairs the small intestine. Biochemical

analysis suggested that the digestive function correlated with the linoleic acid (LA)

concentration in the phospholipids of the enterocytes; upon EFA deficiency, the LA

CHAPTER 7

132

concentration decreased, simultaneously with a decrease in the lactase expression and

activity. Whether digestion of lactose is impaired in pediatric cholestatic patients, with or

without EFA deficiency, is not known. Studies in (non-EFA deficient) bile duct ligated rats

(cholestatic rat model) revealed no significant differences in absorption of glucose or

sucrose between control and bile duct ligated rats.6 However, lactose digestion was not

studied in these rats, which limits the extrapolation of these results to our mouse model

for EFA deficiency. Future studies should focus on determining whether EFA deficiency

is associated with specifically impaired lactose digestion or with a more general defect in

disaccharide digestion and absorption. In addition, studies should be performed to

determine whether it would be beneficial to increase the dietary intake of

monosaccharides compared to that of disaccharides, in order to improve the nutritional

status of patients with CIFTT.

Cholesterol is quantitatively and metabolically an important lipid that enters the intestine

via the diet and via biliary secretion. We studied cholesterol absorption and metabolism

in jejunal intestinal segments in a mouse model of EFA deficiency (chapter 3). The fecal

cholesterol excretion was 57% higher in EFA-deficient mice compared with control mice,

indicating reduced cholesterol absorption. In accordance with reduced cholesterol

absorption, the marker for cholesterol absorption (plasma plant sterols/cholesterol ratio)

was significantly decreased in EFA-deficient mice. Niemen-Pick C1-like 1 protein

(NPC1L1) is the critical player in the absorption of intestinal sterols expressed at the

apical surface of enterocytes. Npc1l1 mRNA expression was decreased in jejunum of

EFA-deficient mice.7 EFA deficiency had no effect on total cholesterol concentrations in

jejunal mucosa, what could be due to a compensatory increase in jejunal cholesterol

synthesis. Additional analysis of triglyceride and fatty acid metabolism revealed elevated

jejunal triglyceride content, accompanied by increased concentrations of oleic acid.

Interestingly, microarray analysis revealed that the mRNA expression of all the genes

involved in cholesterol synthesis was increased in jejunum EFA-deficient mice.

Cholesterol and fatty acid metabolism are mainly regulated by the sterol regulatory

element binding proteins, SREBP2 and SREBP1C, respectively.8 When cholesterol

concentration in the cell decreases, SREBP2 is cleaved by proteases and transported to

the nucleus where it binds to the DNA to increase the mRNA expression of genes

involved in cholesterol synthesis. SREBP2 can also be activated by increased oleic acid

concentration. In jejunum of EFA-deficient mice Srebp2 mRNA expression was

significantly increased, resulting in increased mRNA expression of its target genes

involved in cholesterol metabolism. Srebp1c mRNA expression was also increased in

jejunum of EFA-deficient mice, leading to the increased induction of mRNA expression of

Srebp1c target genes involved in fatty acid metabolism.

Transcriptional analysis further revealed that the pathway involved in the inhibition of the

proteasome was the most significantly affected cellular process by the EFA deficiency in

jejunum. Reduced proteasome activation leads to prolonged expression and activity of

several cellular proteins. Recently, the activity of several transcription factors, proteins

and enzymes involved in cholesterol synthesis (SREBP, ABCA1, HMGCR) has been

shown to be regulated by the ubiquitin-proteasome pathway in a sterol-dependent

manner.9,10

In addition, preliminary data of Hamel imply fatty acids, mainly oleic acid, as

possible regulators of the proteasome pathway.11

Thus, on one hand increased oleic

SUMMARY AND FUTURE PERSPECTIVES

133

7 C

HA

PT

ER

acid can induce jejunal cholesterol synthesis in EFA-deficient mice and on the other

hand, increased oleic acid inhibits the proteasome pathway leading to enhanced

expression of genes involved in cholesterol synthesis. In summary, our data show that

EFA deficiency in mice is associated with cholesterol malabsorption, and with increased

cholesterol biosynthesis. This increase in cholesterol synthesis during EFA deficiency is

not specific for jejunal tissue, as it occurs in epidermal tissue as well.12

Furthermore, we

demonstrated that the mRNA expression of HMGCR, rate limiting enzyme of cholesterol

synthesis, is increased in livers of EFA-deficient mice (chapter 3). We speculate that

modifications in the jejunal proteasome regulation pathway, which prolong the

expression of relevant proteins of lipid metabolism, may be a compensatory mechanism

for the malabsorption of lipids. It is of interest to study whether cholesterol malabsorption

during EFA deficiency in mice leads to reduced membrane cholesterol concentrations

and structural changes within the enterocyte membrane. If this would be the case, it

would be worthwhile to determine to what extent the (postulated compensatory) increase

in cholesterol synthesis in jejunum is capable to correct for reduced membrane

cholesterol content. In order to improve the intestinal function during EFA deficiency,

further studies on cholesterol content in the membranes of the enterocytes in the

intestinal epithelium are relevant.

Essential fatty acid deficiency in mice leads to enhanced ileal bile salt

reabsorption and to persistent hepatic bile salt synthesis in mice

Our previous studies revealed impaired small intestinal function during EFA deficiency,

mainly located at the level of the mid small intestine, i.e. corresponding to the jejunum

(chapter 2 and 3). In order to determine whether EFA deficiency affects other parts of

the small intestine, we studied bile salt reabsorption, what mainly occurs in the terminal

ileum, i.e. the last part of the small intestine (chapter 4). It has been shown by Werner et

al. that EFA-deficient mice have increased bile production and enhanced biliary secretion

of bile salts.1 Using a stable isotope methodology, we characterized relevant kinetic and

quantitative parameters of the enterohepatic circulation of bile salts in EFA-deficient

mice, without interrupting the normal enterohepatic circulation. EFA deficiency-enhanced

bile flow and biliary bile salt secretion were associated with increased ileal bile salt

reabsorption and unexpectedly elevated bile salt synthesis rate. The persistent hepatic

bile salt synthesis was most likely to be explained by the reduction in ileal mRNA

expression of Fgf15 (inhibitor of bile salt synthesis13

). To confirm our in vivo findings, we

additionally measured expression of relevant intestinal genes in the enterohepatic

circulation of bile salt synthesis in EFA-deficient (post-confluent) Caco-2 cells. The cells

were stimulated with chenodeoxycholic acid and GW4064 compound, both potent

agonists of farnesoid X receptor (FXR), relevant regulator of the bile salt synthesis.

Stimulation of EFA-deficient Caco-2 cells resulted in lower induction of the mRNA

expression of relevant genes (FGF19, IBABP) compared with control Caco-2 cells.

Together, these data clearly show that besides the effects on the jejunum, EFA

deficiency affects the terminal ileum function, with respect to the enterohepatic

circulation of bile salts. Interestingly, EFA deficient mice have fat malabsorption, despite

the observation that mice with EFA deficiency have an increased biliary bile salt

secretion, and thus more specifically increased intestinal availability to bile salts. EFA

CHAPTER 7

134

deficiency has some species specific phenotype: in rats, bile salt secretion is decreased

during EFA deficiency, coinciding with a more prominent fat malabsorption, compared

with mice. It seems thus that mice, at least in part, compensate for severe fat

malabsorption by increasing the bile salt concentrations in the intestine. It remains to be

elucidated whether these high bile salt concentrations in the intestine of EFA-deficient

mice are toxic to intestinal tissue in long terms, leading to more functional problems.

In vitro model of essential fatty acid deficiency reveals increased permeability and

impaired mRNA expression of brush border enterocyte markers, which are not

rapidly reversible by linoleic acid (LA) supplementation

In order to study the intracellular effects of EFA deficiency, we established an in vitro

model of EFA deficiency in post-confluent Caco-2 cells (chapter 5). Upon confluence,

these cells differentiate towards the small intestinal phenotype, as indicated by dome

formation, microvilli, and expression of brush-border enzymes.14

Caco-2 cells were cultured in medium containing normal or delipidated FCS (control and

EFA-deficient cells, respectively) for one week post-confluent. From previous studies this

condition has been shown to be sufficient for reproducible induction of the small

intestinal phenotype and of EFA deficiency.14,15

To study EFA deficiency in an in vitro

model, with a phenotype similar to in vivo situation, we have reproduced and further

characterized an in vitro model originally described by Spalinger et al.15,16,17

EFA

deficiency severely reduced the expression of relevant brush border markers of the small

intestine, and impaired the cellular permeability, as demonstrated by increased

transepithelilal electrical resistance (TER). These effects were not rapidly reversible by

LA supplementation to EFA-deficient Caco-2 cells. Theoretically, persistent effects on

permeability and mRNA expression of lactase and sucrase isomaltase after LA

supplementation might be caused by absent capability of EFA-deficient Caco-2 cells to

incorporate the supplemented LA into the cellular phospholipids. However, we

demonstrated that EFA-deficient Caco-2 cells were capable of LA incorporation into

phospholipids to similar extent as control Caco-2 cells during 6 days of LA

supplementation. Despite similar LA concentrations in the phospholipids of EFA-deficient

and control cells, EFA-deficient cells retained increased TER values and severely

reduced mRNA expression of the brush border enzymes. Another possibility is that EFA

deficiency leads to reduced differentiation of the small intestinal enterocytes, as

demonstrated by the decreased expression of the enterocytic markers. Reduced

differentiation might, at least in part, be caused by impaired transcriptional regulation of

the mRNA expression of certain enterocytic markers and could be difficult to reverse by

LA supplementation. Prolonged treatment with (higher concentration) of LA might restore

the enterocyte function during EFA deficiency. Furthermore, additional supplementation

with α-linolenic acid (C18:3ω-3, ALA) along with LA might be useful in order to reverse

the effects of EFA deficiency. Theoretically, increased permeability would be expected to

influence the transmucosal transport of dietary compounds. Yet, EFA deficiency in mice

is associated with reduced nutrient absorption. Studies in EFA-deficient pigs revealed

decreased lipid fluidity within the brush border membrane of the enterocytes.5 Further in

vivo studies in mouse model of EFA deficiency are warranted to determine whether

increased permeability indeed is a common feature of EFA deficiency in vivo.

SUMMARY AND FUTURE PERSPECTIVES

135

7 C

HA

PT

ER

Nevertheless, we expect that this in vitro model will allow performing more detailed

studies on the underlying (molecular) mechanism(s) of the EFA-deficient phenotype in

the small intestinal enterocyte.

Gelucire®44/14 improves fat absorption in rats with impaired lipolysis

EFA deficiency, by itself or in combination with CIFTT or cystic fibrosis, is associated

with severe fat malabsorption. Apart from improving the small intestinal function

(in)directly, the supplementation with exogenous compounds might enhance the

absorption of dietary fat. Gelucire®44/14 is a semi-solid self-micro-emulsifying excipient,

frequently used as an absorption enhancer of water insoluble drugs.18

In chapter 6 we

show that Gelucire®44/14 can enhance fat absorption under conditions of impaired

lipolysis, but not during impaired solubilization in rats. Possibly, Gelucire®44/14 stabilizes

and improves residual lipolytic enzyme activity in vivo. This could be of therapeutic value

in clinical conditions of fat malabsorption due to impaired lipolysis, but probably not in

case of EFA deficiency, since EFA-deficient mice seem to have normal lipolysis.

However, Levy et al. showed in rats impaired intraluminal steps of fat absorption.3 If

lipolysis indeed would be reduced in (pediatric) patients with EFA deficiency,

Gelucire®44/14 might help reduce the severe fat malabsorption. We reasoned, however,

that experiments with Gelucire®44/14 supplementation to mouse model of cystic fibrosis

could be helpful in this respect, before patient studies would be designed and planned.

OVERALL CONCLUSION AND IMPLICATIONS

Our studies on the effects of EFA deficiency on the small intestine in mice and in

immortalized in vitro model of EFA deficiency clearly show that EFA deficiency leads to a

variety of functional changes in the small intestine. More specifically, lipid malabsorption

and disaccharide digestion are impaired during EFA deficiency in mice. Increased

intestinal reabsorption of bile salts is insufficient to normalize the decreased lipid

absorption, underscoring previous implications in mice that intracellular rather than

intraluminal steps of fat absorption are impaired during EFA deficiency in mice. The

(isolated) supplementation of LA does not seem to reverse the effects of EFA deficiency

on the small intestinal enterocytes as revealed by our in vitro study. Maintenance and/or

improvement of the nutritional status of cholestatic patients with EFA deficiency is

relevant since the number of patients on the waiting list for liver transplantation

increases. In order to improve the nutritional strategies of CIFTT patients with EFA

deficiency, further studies on the effects of EFA deficiency on the intestinal function are

warranted as the follow up to our presently obtained results in mice. Stable isotope

dilution techniques represent an elegant methodology that is applicable in pediatric

patients to assess the absorption of several nutrients.19

Patients studies using stable

isotope-labeled macronutrients, i.e. lipids, carbohydrates and proteins, will further assess

nutritional status of children with CIFTT. We expect that these mechanistic nutritional

studies will help to develop and rationalize nutritional therapies for pediatric patients with

impaired digestion or absorption, including patients with EFA deficiency.

CHAPTER 7

136

REFERENCES

1 Werner A, Minich DM, Havinga H, Bloks VW, van Goor H, Kuipers F and Verkade HJ. Am J Physiol

Gastrointest Liver Physiol 2002; 283(4):G900-G908.

2 Werner A, Havinga R, Bos T, Bloks V, Kuipers F and Verkade HJ. Am J Physiol Gastrointest Liver

Physiol 2005; 288(6):G1150-G1158.

3 Levy E, Garofalo C, Thibault L, Dionne S, Daoust L, Lepage G and Roy CC. Am J Physiol Gastrointest

Liver Physiol 1992; 262(2):G319-G326.

4 Christon R, Meslin JC, Thévenoux J, Linard A, Léger CL and Delpal S. Reprod Nutr Dev 1991; 31(6):691-

701.

5 Christon R, Even V, Daveloose D, Leger C and Viret J. Biochimica et Biophysica Acta (BBA) -

Biomembranes 1989; 980(1):77-84.

6 Leonie Los E, Wolters H, Stellaard F, Kuipers F, Verkade HJ and Rings EHHM. AJP - Gastrointestinal

and Liver Physiology 2007; 293(3):G615-G622.

7 Lammert F, Wang DQ. Gastroenterology 2005; 129(2):718-734.

8 Horton JD, Shimomura I, Brown MS, Hammer RE, Goldstein JL and Shimano H. J Clin Invest 1998;

101(11):2331-2339.

9 Bengoechea-Alonso MT, Ericsson J. Current Opinion in Cell Biology 2007; 19(2):215-222.

10 Tanaka AR, Kano F, Yamamoto A, Ueda K and Murata M. Genes Cells 2008; 13(8):889-904.

11 Hamel FG. Metabolism 2009; 58(8):1047-1049.

12 Proksch E, Feingold KR and Elias PM. J Invest Dermatol 1992; 99(2):216-220.

13 Jung D, Inagaki T, Gerard RD, Dawson PA, Kliewer SA, Mangelsdorf DJ and Moschetta A. J Lipid Res

2007; 48(12):2693-2700.

14 Ding QM, Ko TC and Evers BM. Am J Physiol Cell Physiol 1998; 275(5):C1193-C1200.

15 Spalinger J, Seidman E, Lepage G, Menard D, Gavino V and Levy E. Am J Physiol Gastrointest Liver

Physiol 1998; 275(4):G652-G659.

16 Lukovac S, Los EL, Stellaard F, Rings EH and Verkade HJ. Am J Physiol Gastrointest Liver Physiol 2008;

295(3):G605-G613.

17 Lukovac S, Los EL, Stellaard F, Rings E.H. and Verkade HJ. Am J Physiol Gastrointest Liver Physiol

2009:00091.

18 Chambin O, Jannin V. Drug Development and Industrial Pharmacy 2005; 31(6):527-534.

19 Hulzebos C, Renfurm L, Bandsma R, Verkade H, Boer T, Boverhof R, Tanaka H, Mierau I, Sauer P,

Kuipers F and Stellaard F. J Lipid Res 2001; 42(11):1923-1929.

APPENDICES

S. Lukovac

NEDERLANDSE SAMENVATTING

140

NEDERLANDSE SAMENVATTING

Essentiële vetzuren, linolzuur (C18:2ω-6, LA) and α-linoleen zuur (C18:3ω-3, ALA), zijn

heel belangrijk voor verscheidene biologische functies in mens en dier. Essentiële

vetzuren kunnen niet in het lichaam zelf worden gemaakt en moeten dus door middel

van voeding opgenomen worden. Een tekort aan essentiële vetzuren kan verstoringen

veroorzaken in de ontwikkeling van de hersenen en het immuunsysteem, maar ook in de

celmembranen van verscheidene weefsels, waaronder die van de hersenen, nieren,

lever, huid en retina. Essentiële vetzuur deficiëntie (EVZD) is een aandoening waarbij er

een tekort aan essentiële vetzuren is. Deze aandoening kenmerkt zich door een

verstoring van de cognitieve en motorieke functies, groeistoornis, droge huid, haarverlies

en functionele afname van organen zoals het hart en de lever. EVZD kan het resultaat

zijn van een te lage voedingsinname van essentiële vetzuren, maar ook van een

verlaagde opname of een verhoogd metabolisme van essentiële vetzuren. Vooral

kinderen met ernstige leverziekten die lijden aan cholestase (verminderde

galuitscheiding vanuit de lever naar de darm) kunnen EVZD ontwikkelen. Als gevolg van

de EVZD ontwikkelen deze patiënten ernstige malabsorptie van nutriënten. Dit leidt tot

een zeer slechte voedingsstatus en daarmee samenhangend een zeer slechte prognose

voor deze patiënten. Om de voedingsstatus van kinderen met cholestase en EVZD te

verbeteren is het van belang de functie van de dunne darm, het orgaan waar het

grootste deel van de opname van voeding plaatsvindt, te onderzoeken.

Eerdere studies naar EVZD hebben zich vooral gericht op de consequenties voor de

werking van de lever, de hersenen en het hart. Over de effecten van de EVZD op de

dunne darm functie en fysiologie is niet veel bekend. Om de voedingsstatus van

patiënten met cholestase en EVZD te verbeteren is het onderzoek naar de effecten van

deze aandoening op de dunne darm functie essentieel. Studies in proefdieren met EVZD

hebben in het verleden laten zien dat EVZD leidt tot een verslechterde opname van

vetten in de dunne darm. Naar een verklaring voor de verslechtering van deze opname

werd tot op heden geen onderzoek verricht.

Het doel van dit proefschrift was om de effecten van EVZD op de functie van de dunne

darm in kaart te brengen door middel van studies in een muismodel voor EVZD. Dit

model werd eerder ontwikkeld in ons laboratorium. Verder hebben wij voor de

gedetailleerde bestudering van darmcellen een celkweekmodel ontwikkeld. In deze

modellen zijn de opname en vertering van cholesterol en koolhydraten bestudeerd.

Absorptiestudies hebben wij kunnen uitvoeren met behulp van de “stabiele isotopen

techniek”. Deze methode, in ons laboratorium ontwikkeld, wordt toegepast om de

verschillende stappen van de genoemde metabole processen nauwgezet te kunnen

meten. Een groot voordeel van deze techniek is dat er maar relatief kleine bloedvolumes

nodig zijn voor de uiteindelijke meting. Daarom is deze elegante techniek ook

toepasbaar in mensen, inclusief kleine kinderen. Dit is een belangrijk aspect met het oog

op vervolgstudies in jonge patiënten met cholestase en EVZD.

Verstoorde lactosevertering en veranderingen in het cholesterol metabolisme bij

essentiële vetzuurdeficiëntie

Studies hebben aangetoond dat EVZD in zowel rat- als muismodellen leidt tot een

NEDERLANDSE SAMENVATTING

141

AP

PE

ND

ICE

S

verstoorde opname van vet. Effecten van EVZD op de opname van andere nutriënten in

deze diermodellen werd echter niet onderzocht. In onze hypothese stellen wij dat EVZD

de functie van de dunne darm verstoort. Indien dit waar is, dan zou niet alleen de

opname van vetten, maar ook de opname van andere voedingsstoffen verstoord moeten

zijn. Functieverlies tijdens EVZD met betrekking tot de opname van meerdere soorten

nutriënten, kan belangrijke consequenties hebben voor de voedingssamenstelling bij

deze aandoening.

We hebben ons gericht op de opname van twee belangrijke koolhydraten, namelijk de

monosaccharide glucose en de disaccharide lactose. Laatste bestaat uit twee

componenten, namelijk uit glucose en galactose. Lactose moet gesplitst worden door

middel van het enzym lactase, voordat het kan worden opgenomen door de dunne

darmcel.

In ons muismodel hebben we de stabiele isotopen techniek toegepast waarmee we de

opname van de koolhydraten gemeten hebben na toediening van stabiel gelabelde

varianten van glucose en lactose (hoofdstuk 2). In deze studie laten wij zien dat in het

muismodel voor EVZD de glucose opname niet is verstoord, terwijl er wel sprake is van

vertraagde lactosedigestie. Deze observatie ging gepaard met verlaagde activiteit van

het enzym lactase in de dunne darm. Dit suggereert een verminderde splitsing van dit

molecuul. Biochemische analyses hebben laten zien dat verminderde enzymactiviteit

van lactase sterk gepaard ging met verlaagde LA concentraties in de dunne darm.

Verdere studies zullen nodig zijn om te verifiëren of de lactose vertering ook verstoord is

in cholestatische patiënten. Deze studies zijn van belang om de opnamecapaciteit van

mono- en disacchariden in deze patiënten te bepalen om hierdoor de voedingsstatus te

kunnen optimaliseren.

Een ander belangrijke component in het dieet is cholesterol. Cholesterol kan naast

opname via dieet ook vanuit de lever in gal naar de dunne darm uitgescheiden worden.

In hoofdstuk 3 hebben we het cholesterolmetabolisme in muizen met EVZD

bestudeerd. Verder hebben wij in dit hoofdstuk de transcriptionele regulatie van genen,

betrokken bij de lipide metabolisme in de dunne darm, op mRNA niveau bestudeerd met

behulp van de zogenaamde “microarray” analyse.

Cholesteroluitscheiding in feces van muizen met EVZD was 57% hoger dan in gezonde

muizen. Dit suggereerde verminderde opname van cholesterol tijdens EZVD. Deze

resultaten werden ondersteund door verlaagde markers voor cholesterolopname, zoals

in plasma gemeten. Daarnaast was er een significant verlaagde expressie van het

cholesterol transportmolecuul NPC1L1. Echter de concentratie van cholesterol in

darmmucosa bleef onveranderd in de muizen met EVZD. Dit leidde tot de hypothese dat

er een compensatie mechanisme aanwezig zou moeten zijn dat ervoor zorgt dat de

cholesterol concentraties op peil blijven. Dit zou bijvoorbeeld kunnen door een toename

in de cholesterolsynthese. Onze microarray analyse bevestigde deze hypothese; de

mRNA expressie van alle genen waarvan het bekend is dat ze betrokken zijn bij de

cholesterol synthese was significant verhoogd in muizen met EVZD. Tevens bleek ook

het mRNA van genen die betrokken zijn bij de synthese van triglyceriden en vetzuren

verhoogd te zijn in het jejunum van muizen met EVZD. De toename in mRNA ging

gepaard met een verhoogde triglyceride en oleaat concentratie in het mucosa van het

jejunum. De geïnduceerde cholesterol- en vetsynthese worden sterk gereguleerd door

NEDERLANDSE SAMENVATTING

142

de transcriptiefactoren SREBP1c en SREBP2. Inderdaad bleek uit onze studies ook dat

de expressie van deze transcriptiefactoren significant verhoogd te zijn tijdens EVZD in

jejunum. Dit werd gevolgd door de verhoging van de genen betrokken bij de cholesterol-

en vetsynthese. Verder hebben wij aangetoond met behulp van de microarray analyse

dat ook genen gerelateerd aan het proteasoom een verhoogde expressie hadden tijdens

EVZD. Het proteasoom reguleert de aanwezigheid van eiwitten in de cellen door deze,

indien nodig, af te breken. De transcriptionele veranderingen in de proteasoomgenen

zouden een oorzaak kunnen zijn van de geobserveerde verhoging in genexpressie en de

toename in lipide concentraties in jejunum. In het licht dat cholesterol van belang is voor

het behoud van de functie en morfologie van celmembranen zijn vervolgstudies naar de

effecten van EVZD op het celmembraan van belang. Hierin zou ook onderzocht moeten

worden of de compensatoire inductie van cholesterol synthese tijdens EVDZ afdoende is

om de functie van de celmembranen te waarborgen.

Verhoogde reabsorptie en aanmaak van galzouten bij essentiële vetzuurdeficiëntie

De studies in het muismodel voor EVZD hebben aangetoond dat de verstoringen in de

darm vooral plaatsvinden op het niveau van het jejunum. Om uit te zoeken of EVZD ook

tot verstoringen in andere delen van de dunne darm leidt, hebben we in ons muismodel

ook de galzoutreabsorptie bestudeerd. Galzoutreabsorptie vindt namelijk voornamelijk

plaats in het terminale ileum, het meest distale deel van de dunne darm. De

galzoutsynthese in de lever wordt geblokkeerd als er voldoende galzouten in de darm

worden opgenomen, om de totale galzoutconcentraties in het organisme constant te

houden. Hiernaast is een additioneel mechanisme, gereguleerd door het eiwit Fgf15 in

de muis, aanwezig dat zorgt voor de negatieve inhibitie van galzoutsynthese vanuit het

ileum. Dit mechanisme houdt in dat hoe meer galzouten er in de darm aanwezig zijn,

hoe meer Fgf15 door de darm wordt uitgescheiden naar de bloedbaan. Fgf15 wordt dan

naar de lever getransporteerd om het signaal af te geven dat er minder galzouten

aangemaakt moeten worden door de lever.

De stabiel gelabelde experimenten met het galzout cholaat (2H4-cholaat) stelden ons in

staat om verschillende parameters van de enterohepatische circulatie van galzouten

kwantitatief in vivo te bepalen. Uit de resultaten beschreven in hoofdstuk 4 bleek dat de

verhoogde galzoutsecretie uit de lever naar de darm gepaard ging met een verhoogde

galzoutsynthese, ondanks een verhoogde galzoutopname in de dunne darm. Dit duidt op

de afwezigheid van de normale terugkoppeling van galzoutsynthese in de dunne darm.

Onze data werden verder bevestigd door de verlaagde expressie van Fgf15 in de

terminale ileum. Effecten van EVZD op de dunne darm expressie van genen betrokken

bij het galzoutmetabolisme zijn ook in celkweek modellen bevestigd.

Deze data laten zien dat naast lactose- en lipidemetabolisme ook het

galzoutmetabolisme verstoord is tijdens de EVZD. Het laat ook zien dat niet alleen het

jejunum, maar ook de functie van het ileum verstoord is in muizen met EVZD. De

bevinding dat deze muizen, ondanks hun verhoogde darmopname van galzouten, vet

malabsorptie vertonen hebben wij nog niet kunnen verklaren.

Effecten van de EVZD in een in vitro model

Om de effecten van de EVZD in de cel te kunnen bestuderen, hebben wij een model

NEDERLANDSE SAMENVATTING

143

AP

PE

ND

ICE

S

ontwikkeld waarbij er gebruik werd gemaakt van de Caco-2 cellijn (hoofdstuk 5). Voor

dit doeleinde hebben wij de Caco-2 cellen gebruikt die afkomstig zijn van de humane

dikke darm kankercellen. Deze cellen kunnen lang kunstmatig in kweek gehouden

worden. Caco-2 cellen hechten aan het kweekoppervlak en differentiëren in verloop van

tijd richting dunne darm epitheel cellen. Het dunne darm fenotype van deze cellijn

kenmerkt zich door de expressie van de zogenaamde brush border membraan enzymen.

Deze enzymen komen tot expressie in het apicale membraan van enterocyten en

hebben een essentiële rol in de opname van koolhydraten.

Onze experimenten in de EVZD-Caco-2 cellen laten zien dat EVZD sterk de expressie

van de belangrijke brush border enzymen reduceert. Verder beïnvloedt de EVDZ de

permeabiliteit van de membranen van de cellen. Linolzuur (LA) is een belangrijk

component van deze membranen en in een poging om de effecten veroorzaakt door de

EVZD op de membranen te reduceren, hebben we aan EVZD-Caco-2 cellen het

essentieel vetzuur LA toegediend. Echter, de toediening van LA leidde niet tot een

significante afname in de effecten veroorzaakt door de EVZD.

Verdere studies zijn nodig om uit te zoeken of LA in een andere dosis en langdurige

behandeling met LA de functie van de EVZD-darmcellen kan herstellen. In deze studies

zou het ook raadzaam zijn om het effect van LA in combinatie met het andere essentiële

vetzuur α-linoleen zuur (ALA) te bestuderen. Het zou kunnen zijn dat het combineren

van essentiële vetzuren leidt tot een synergistische afname van de effecten van EVDZ.

Dit in vitro model zal verder kunnen bijdragen aan de uitvoering van meer gedetailleerde

studies naar achterliggende, cellulaire mechanismen van de EVZD in de dunne darm

enterocyten.

Gelucire®44/14 kan de vetabsorptie verhogen in ratten met verstoorde lypolise

Essentiële vetzuur deficiëntie, in combinatie met cholestase of cystic fibrosis (CF,

taaislijnmziekte), gaat bijna altijd gepaard met verminderde vetopname. Behalve het

herstellen van de dunne darm functie, kan de toediening van stoffen die de vetabsorptie

verbeteren een uitkomst zijn. Een mogelijk geschikte kandidaat is Gelucire®44/14 dat

toegediend wordt om de opname van verschillende, moelijk oplosbare, medicijnen te

bevorderen. In hoofdstuk 6 beschrijven wij een studie in ratten waarin we het effect van

Gelucire®44/14 testen op twee verschillende stadia van vetopname. Vetopname wordt in

vier verschillende stappen onderverdeeld. Eerst vindt de emulsificatie plaats waarbij vet

verdeeld wordt in vetdruppels die zo beter toegankelijk worden voor de enzymen. Deze

enzymen zorgen vervolgens dat de tweede stap, de lipolyse van vet, wordt uitgevoerd

waarbij vetten gesplitst worden in vetzuren. Deze vetzuren vormen dan samen met

galzouten en andere lipiden de zogenaamde micellen. Dit proces wordt de solubilizatie

genoemd en zorgt ervoor dat vetten beter opgenomen kunnen worden door de dunne

darm. De laatste stap van de vetopname is het transport over het apicale membraan van

de enterocyten en de uitscheiding in het bloed. Wij hebben Gelucire®44/14 getest in

ratten met een verstoorde lipolyse (model voor de patiënten met CF) en een

verminderde solubilizatie (model voor patiënten met cholestatische leverziektes). We

laten zien dat Gelucire®44/14 de vetopname in ratten met verminderde lipolyse bevordert

doordat het de opname van verzadigde vetzuren stimuleert. Echter, Gelucire®44/14 heeft

geen significant invloed op de vetopname in ratten met een verstoorde solubilizatie. De

NEDERLANDSE SAMENVATTING

144

volgende stap in deze studies zal het testen van Gelucire®44/14 in specifieke

muismodellen voor CF zijn, om zodoende de data verkregen in ratten te valideren.

CONCLUSIE EN TOEPASSINGEN

Studies beschreven in dit proefschrift laten zien dat EVZD gepaard gaat met een aantal

functionele veranderingen van de dunne darm. In het bijzonder, zijn de lipide en lactose

absorptie verstoord tijdens de EVZD in muizen. Verder gaat EVZD in muizen gepaard

met een verhoogde galzoutabsorptie in de dunne darm. Echter, deze is onvoldoende om

de verstoorde vetopname tijdens EVZD volledig te herstellen. Dit suggereert dat de

defecten veroorzaakt door de EVZD niet zozeer in het darmlumen plaatsvinden, maar

voornamelijk in de enterocyten. Dat de effecten van EVZD gecompliceerd zijn wordt

duidelijk gemaakt doordat het simpelweg toedienen van het essentiële vetzuur LA niet

voldoende is om de effecten van EVDZ tegen te gaan.

Het behoud en herstel van de voedingsstatus van kinderen met cholestase en EVZD is

essentieel. Dit is zeker het geval naarmate het aantal patiënten op de wachtlijsten voor

de levertransplantatie stijgt. Om de voedingsstatus van deze groep patiënten beter te

bestuderen en uiteindelijk te verbeteren is het van belang om de studies beschreven in

dit proefschrift uit te breiden met patiëntenstudies. Een belangrijke onderzoeksmethode

hierbij kan de stabiele isotopen techniek zijn. Deze techniek kan de absorptie van

verschillende voedingstoffen helpen monitoren, zonder schadelijke gevolgen voor de

patiënt. Deze studies kunnen een belangrijke bijdrage leveren voor de ontwikkeling van

rationele voedingstherapieën voor kinderen en andere patiënten met EVZD.

DANKWOORD

145

AP

PE

ND

ICE

S

DANKWOORD

Aan al het moois komt een einde. Ik heb genoten van de afgelopen vier jaar en wil graag

alle mensen bedanken die daarbij betrokken waren.

Beste Henkjan, bedankt dat je mij de weg en de schoonheid van de wetenschap hebt

helpen ontdekken. Jouw scherpte en onvermoeibaarheid hebben mij direct vanaf het

begin enorm geïnspireerd en gemotiveerd. Beste Edmond, bij jou kwam ik altijd vandaan

met nieuwe ideeën en inzichten. Jouw relaxedheid en gevoel voor humor zorgden er

voor dat ik alles weer kon relativeren. Onze lunchbesprekingen in ‟t Feithhuis zal ik niet

snel vergeten. Heerlijk was ‟t om aan de dagelijkse werkomgeving te ontsnappen om

belangrijke toekomstplannen en de gang van zaken door te nemen. Ik ben trots dat ik

deel mocht uitmaken van jullie team. Bedankt!

Bert en Folkert, “de grote bazen”, ook al waren jullie niet mijn directe begeleiders, jullie

scherpe oog, dat mij regelmatig in de gaten hield, zorgde ervoor dat ik altijd tot nieuwe

inzichten kwam. Bedankt dat ik deel uit mocht maken van jullie lab. Folkert, ik ben nog

steeds onder de indruk van je operatiekunsten, bedankt dat je tijd kon vrijmaken voor

mijn BDD experimenten. Bert, dank dat je mij in de mysterieuze wereld van TICE hebt

geïntroduceerd…ik ben ervan overtuigd dat we het binnenkort helemaal zullen begrijpen.

Graag wil ik de leden van de leescommissie bedanken bestaande uit Prof. dr. Bos, Prof.

dr. Groen en Prof. dr. Heineman voor het beoordelen van mijn proefschrift.

Lieve Els, soms was jij een beetje streng voor mij….Als ik mij afvroeg of ik een stuk

binnen drie weken af kon ronden, maakte jij een afspraak voor me over twee weken.

“Het kan makkelijk”, zei jij dan….En altijd had je gelijk! Bedankt voor het maken van al

die onmogelijke afspraken met de heren, voor de goede gesprekken, je interesse en de

koekjes!

Graag wil ik een aantal mensen van het AMC bedanken voor het gebruik maken van hun

muizen toen ik mij ineens in PFIC-3 ging verdiepen. Prof. dr. Oude Elferink, bedankt dat

we af en toe jullie gave master classes mochten bijwonen! Suzanne en Cindy, jullie zijn

mijn AMC-heldinnen! Alles was bij jullie binnen no time geregeld. Bedankt voor de goede

samenwerking. Ook al staan de resultaten niet in dit proefschrift, ik weet zeker dat we ze

binnenkort zullen publiceren.

Dear Dr. Gregory, Peter, thank you for giving me the opportunity to study the role of

Gelucire within the field of fat malabsorption. Your critical revisions of our manuscript

were very useful and hopefully our data will be published in BBA soon.

Ik wil graag al mijn collega‟s van de Kindergeneeskunde bedanken voor de

samenwerking en de leuke tijd op het lab en daarbuiten: Frans S (hierna ben ik echt

weg, Frans), Dirk-Jan, Uwe (thanks for your great La Colombe tip, where the real Philly

coffee is), Klary, Barbara, Rebecca, Dolf, Anke (ik wacht op je met Philly cheesesteaks),

DANKWOORD

146

Hilde (ben je al afgestudeerd?), Gemma (billis ganador), Marijke (je lijstje heeft

gewerkt!), Aldo (darmen zijn echt cool), Janine, Marije, Jurre, Maurien, Torsten, Agnes,

Maaike (PUFA-PPAR-oorbellen koningin, bedankt voor alle discussies en

kennisoverdracht, nogmaals: die wandelschoenen konden echt niet), Leonie (bedankt

voor je hulp, gezelligheid en een mooi begin van mijn promotie), Jelske (bier staat koud

in Philly), Niels, Esther (mijn techno-collegaatje, ik zal nooit onze

maandagochtendblikken vergeten), Jaana, Margot, Wytske, Arne, Maxi (keep on rrrrrr),

Annelies (ik mis ons King-Size bed), Jan Freark (dichter), Mariëtte, Robert, Frank,

Anniek K, Jaap, Hester, Thomas, Yen, Frans C (a.k.a. Ray), Karin G, Anja, Marjan,

Willemien, Andrea, Juul (bedankt voor alle hulp met de proeven en de

hardloopmomenten), Wytze, Aicha, Elles, Theo B, Theo van D, Vincent

(wetenschapverslaafde en microarray genie, bedankt), Janny, Nicolette, Renze (ik ben

nog steeds jaloers op jouw Madagaskar avontuur), Henk (blijf zingen!), Ingrid (bedankt

voor alles wat je me geleerd hebt over de vetzuren, GC en het waarderen van die

eindeloze pieken….en voor alle goede gesprekken), Rick (bedankt voor alle hulp- en

leermomenten en natuurlijk de gezelligheid; zouden ze in Philly galcanulaties doen?),

Trijnie, Hilda, Fjodor, Klaas, Pim I en II, Albert, Hermie, Conny, Janneke, Hilde en Geja.

Alle collega‟s van het MDL wil ik bedanken voor alle hulp en gezelligheid: Golnar (thanks

for the great time we had in Boston and good luck, you‟ll do great!), Mark, Manon,

Mariska, Elise, Tjasso, Jannet, Martijn, Jannes, Krystof, Titia, Axel, Sandra, Floris, Fiona,

Lisette, Rebecca, Atta, Anouk, Klaas Nico en Han.

Alex, Flip, Sylvia, Natascha, Hester en de rest van het CDL wil ik bedanken voor alle

goede zorg en planning rondom mijn dierstudies.

Wetenschap is geweldig, maar tijd met vrienden het dierbaarst. Graag wil ik al mijn

vrienden bedanken voor hun interesse in mijn vage studies en de leuke dingen die we

deden om aan de dagelijkse sleur te ontsnappen: Suzanne en Hilmar (nu zijn we alle

vier zeergeleerd…afstand Philly-Kopenhagen is zeker niet te groot), Karin en Guido (Ka,

succes met jouw promotie; ik ben blij dat ik jullie heel snel weer in de VS zie), Floor,

Aleida-Irma-Fleur (ik zal Den Haspel, maar vooral jullie enorm missen…zullen we snel

SATC bij mij in de buurt spelen?), Maan (succes met jouw laatste loodjes en geniet van

jullie nieuwe aanwinst!), Emily en Yvonne (en “de man” van jullie leven), Nardy (niet

weggaan als ik terug ben, we moeten nog naar Stubnitz), Marijn (die Awakenings komt

nog wel), Josien, Robbert (bedankt voor alle tips en tot snel @UPenn), Karin K, Niels

(als ik gek word, kom ik bij jou), Johnny (Vielen Dank für die Gemütlichkeit und der

Schnäpse), Sebastiaan, Esther (bedankt dat ik bij jou mocht oefenen) en Maarten van

der V, Paul en Christine, Laura en Steijn, Sarah, Huygen, Wouter B en Sachi, Maarten

de J, Raphaël en Margriet (newlyweds, jammer dat ik jullie feestje moet missen), Luuk

en Lobke (en de kriebels…), Wouter de J, Dima, Frank en Marieke, Frank S, Tjalling,

B.I.P., Tobias, Robert, Michael, Mark H, Marc (je hebt de helft van de wereldcreativiteit in

je en dat is stoer! Bedankt!).

DANKWOORD

147

AP

PE

ND

ICE

S

Mijn lieve paranimfen, onze historie gaat nog ver voor de promoties. Bedankt dat jullie bij

mij zijn op die belangrijke dag en ik hoop dat ik het jullie niet al te moeilijk heb gemaakt.

Partner, Hilde, afstand Boston-Philadelphia is makkelijk te overzien, dit is essentieel voor

het langsbrengen van prachtig ge-photoshopte kunstwerken! Maak me trots op 28 april!

Lieve Irma, zet „m op die laatste paar maandjes! Ik zou je tekort doen door op deze

pagina op te sommen waar ik je allemaal voor wil bedanken. Je weet het wel. Bedankt!

Eeuwe Jan, mijn lievelingsleraar, bedankt voor je geweldige biologielessen! Door jou ben

ik (medisch) bioloog geworden.

Sve mi skupa: Bilja, Ljilja, Maja, Ogi i Minka, moje najdraže Dobojke! Hvala vam na svim

divnim trenutcima skupa, mada nisu brojni. Kvalitet je vazniji od kvantiteta. Nadam se da

ćemo se uskoro opet sve sastati (“naiškat se i pravit haos“). Majo, kako da se zahvalim

najboljoj drugarici koja je uz mene sve ove godine i koja zna i razumije sve? Hvala na

predivnom prijateljstvu, ovo je samo početak! Kad ćemo više opet živjeti u istom gradu?!

Lieve Harm, Coby, Agnes en Jochem, bedankt dat ik mij altijd zo thuis voel bij jullie.

Jullie zijn lief. Ik hoop (zeker vanaf juli) regelmatig op de hoogte gehouden te worden

van alle spannende, nieuwe ontwikkelingen! Ik zal onze Sinterklaasavonden missen.

Zullen we ze volgend jaar naar Philly verplaatsen?

Majka i djeda, i ostali članovi familije, hvala na podršci i lijepim trenutcima skupa. Nadam

se da će ih biti još više u buduće.

Mama, tata, seka, Jasmin, Damir, bedankt lieverds. Philadelphia is echt niet te ver en we

hebben Skype! Damir, bedankt voor al je kusjes en knuffels. Emina, de enige echte

superwoman, bedankt dat je mijn zusje bent. Ik ben beretrots op je! Lieve paps en

mams, bedankt voor alles. Ooit hoop ik net als jullie te worden. Volim vas do neba i

dalje.

Lieve Bram, Dr. ten Cate, ik heb het allerliefste vriendje van de hele wereld! Bedankt dat

ik bij je mocht afkijken en bedankt dat je er altijd voor me bent. Dankzij jou is het in vier

jaar gelukt. De rest weet je wel…

Op naar een nieuw avontuur!!

Sabina

BIOGRAPHY

148

BIOGRAPHY

Sabina Lukovac was born on August 28, 1980 in Doboj (Bosnia and Herzegovina). In

1992 she moved to the Netherlands, where she graduated in 1999 from high school

(Pieter Jelles College) in Leeuwarden. Afterwards, she continued her education at the

University of Groningen where she studied Biology and in 2005 she received her

Master’s degree in Medical Biology. Research projects performed during her studies was

mainly related to subjects within the field of Molecular Neurobiology. In 2005, Sabina

started her PhD project at the department of Pediatrics at the Universituy Medical Center

Groningen under supervision of Prof. dr. Henkjan Verkade and Prof. dr. Edmond Rings.

Results obtained from her PhD-studies are described in this dissertation and were part of

the MWO 04-38 project “Treatment of Cholestasis-induced failure to thrive (CIFTT) via

improvement of intestinal function” (funded by the Dutch Digestive Foundation). From

April 15, 2010 Sabina is appointed as the postdoctoral fellow at the Department of

Genetics at the University of Pennsylvania under supervision of Prof. dr. Klaus Kaestner.

Her postdoctoral research will focus on molecular mechanisms of the organogenesis and

physiology of the liver, pancreas and the gastrointestinal tract.

BIOGRAFIE

Sabina Lukovac werd op 28 augustus 1980 geboren te Doboj (Bosnië en Hercegovina).

In 1992 is zij naar Nederland verhuisd en heeft in 1999 haar Atheneumdiploma aan de

Pieter Jelles College in Leeuwarden behaald. In dat jaar is zij aan haar studie Biologie

begonnen aan de Rijksuniversiteit Groningen, waar zij in 2005 haar doctorandus titel in

de Medische Biologie behaalde. Haar onderzoeken tijdens de studie Medische Biologie

hadden als hoofdonderwerp moleculaire neurobiologie. In 2005 is Sabina aan haar

promotieonderzoek bij de afdeling Kindergeneeskunde aan het Universitair Medisch

Centrum Groningen begonnen onder begeleiding van Prof. dr. Henkjan Verkade en Prof.

dr. Edmond Rings. De resultaten van haar onderzoek, gefinancierd door de Maag Lever

Darm Stichting (MWO 04-38; project “Treatment of Cholestasis-induced failure to thrive

(CIFTT) via improvement of intestinal function”), worden in dit proefschrift beschreven.

Per 15 april 2010 zal Sabina als postdoctoraal onderzoeker op de afdeling Genetica

gaan werken aan de Universiteit van Pennsylvania onder begeleiding van Prof. dr. Klaus

Kaestner. Daar zal ze zich bezig gaan houden met moleculaire mechanismen van de

ontwikkeling en de fysiologie van de lever, pancreas en darm.

PUBLICATIONS

149

AP

PE

ND

ICE

S

LIST OF PUBLICATIONS

Nutrition for children with cholestatic liver disease

Los EL, Lukovac S, Werner A, Dijkstra T, Verkade HJ, Rings EH.

Nestle Nutr Workshop Ser Pediatr Program. 2007;59:147-57; discussion 157-9. Review

Essential fatty acid deficiency in mice impairs lactose digestion

Lukovac S, Los EL, Stellaard F, Rings EH, Verkade HJ.

Am J Physiol Gastrointest Liver Physiol. 2008 Sep;295(3):G605-13.

Effects of essential fatty acid deficiency on enterohepatic circulation of bile salts

in mice

Lukovac S, Los EL, Stellaard F, Rings EH, Verkade HJ.

Am J Physiol Gastrointest Liver Physiol. 2009 Sep;297(3):G520-31

The role of CXC chemokine ligand (CXCL)12-CXC chemokine receptor (CXCR)4

signalling in the migration of neural stem cells towards a brain tumour

van der Meulen AA, Biber K, Lukovac S, Balasubramaniyan V, den Dunnen WF,

Boddeke HW, Mooij JJ.

Neuropathol Appl Neurobiol. 2009 Dec;35(6):579-91

Gelucire®44/14 improves fat absorption in rats with impaired lipolysis

Lukovac S, Gooijert KEG, Gregory PC, Shlieout G, Stellaard F, Rings EHHM, Verkade

HJ

Manuscript conditionally accepted for publication in Biochimica et Biophysica Acta (BBA)

Molecular and Cell biology of Lipids

Essential fatty acid deficiency in mice is associated with cholesterol

malabsorption and increased jejunal lipid synthesis

Lukovac S, van der Wulp MYM, Blok VW, Boekschoten M, Dekker J, Groen AK, Rings

EHHM, Verkade HJ

Manuscript in preparation

Functional characterization of an in vitro model of essential fatty acid deficiency in

intestinal epithelial cells

Lukovac S, Rings EHHM, Verkade HJ

Manuscript in preparation