Essential fatty acid absorption and metabolism in …ESSENTIAL FATTY ACIDS Linoleic acid (LA,...

Essential fatty acidabsorption and metabolism

in hepatic disorders

food for thought

Anniek Werner

Department of PediatricsCenter for Liver, Digestive and Metabolic Diseases

University Medical Center Groningen

proefschrift_def_v010605def.qxp 2-6-2005 1:25 Pagina 1

Rijksuniversiteit Groningen

Essential fatty acidabsorption and metabolism

in hepatic disorders

food for thought

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 6 juli 2005

om 16.15 uur

door

Anniek Wernergeboren op 10 februari 1973

te Nijmegen


Promotores:

prof. dr. H.J. Verkade

prof. dr. F. Kuipers

prof. dr. P.J.J. Sauer

Beoordelingscommissie:

prof. dr. M.J. Slooff

prof. dr. E.J. Duiverman

prof. dr. G. Hornstra


Voor Folkert

Il est très simple: on ne voit bien qu' avec le coeur.

L'essentiel est invisible pour les yeux.

Le petit prince, Antoine de Saint-Exupéry


Paranimfen:

Renate Wachters-Hagedoorn

Robert Bandsma

ISBN 90-3672-298-5

The research described in this thesis was performed at the Department of

Pediatrics, Center for Liver, Digestive and Metabolic Diseases, University Medical

Center Groningen, the Netherlands. This thesis was supported by the Groningen

University Institute for Drug Exploration (GUIDE) and by the Netherlands

Organization for Scientific Research (NWO grant 90462210).

Cover: Docosahexaenoic acid (DHA), photomicrograph.

Courtesy to M.W. Davidson, Florida State University.

Graphic design: Werner Bros. Ltd.

Print: Ponsen & Looijen B.V., Wageningen, the Netherlands.

Copyright 2005 by Anniek Werner.

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

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

holding the copyright of the published articles.


CONTENTS

1 - General introduction

Partially published as: "Fat absorption and lipid metabolism in cholestasis"

A. Werner, F. Kuipers, H.J. VerkadeMolecular Pathogenesis of Cholestasis, 2003 M. Trauner, P.L.M. Jansen (Eds), Landes Bioscience, p321-335

2 - Fat malabsorption in essential fatty acid-deficient mice is not due to

impaired bile formation

A. Werner, D.M. Minich, H. Havinga, H. van Goor, F. Kuipers, H.J. VerkadeAm J Physiol Gastrointest Liver Physiol 2002; 283(4): G900-908

3 - Essential fatty acid deficiency in mice is associated with hepatic

steatosis and secretion of large VLDL particles

A. Werner, H. Havinga, T. Bos, V.W. Bloks, F. Kuipers, H.J. VerkadeAm J Physiol Gastrointest Liver Physiol 2005; 288(6): G1150-1158

4 - Lymphatic chylomicron size is inversely related to biliary

phospholipid secretion in mice

A. Werner, H. Havinga, F. Perton, F. Kuipers, H.J. VerkadeConditionally accepted for publication in Am J Physiol 2005

5 - No indications for altered essential fatty acid metabolism in two

murine models for cystic fibrosis

A. Werner, M.E.J. Bongers, M.J. Bijvelds, H.R. de Jonge, H.J. VerkadeJ Lipid Res. 2004;45(12): 2277-2286

6 - Treatment of essential fatty acid deficiency with dietary triglycerides

or phospholipids in a murine model of extrahepatic cholestasis

A. Werner, H. Havinga, F. Kuipers, H.J. VerkadeAm J Physiol Gastrointest Liver Physiol 2004, 286(5): G822-832

7 - Oral treatment of essential fatty acid deficiency with triglycerides or

phospholipids in children with end stage liver disease

A. Werner, C.M.A. Bijleveld, I.A. Martini, M. van Rijn, J. van der Heiden,

P.J.J. Sauer, H.J. VerkadeSubmitted

8 - General discussion and summary

Nederlandse samenvatting

Dankwoord

Curriculum Vitae

Publicaties

9

37

55

75

91

111

131

147

156

163

166

167


ABBREVIATIONS

AA arachidonic acid (C20:4n-6)

ALA alpha-linolenic acid (C18:3n-3)

ALT alanine aminotransferase

AST aspartate aminotransferase

AP alkaline phosphatase

CF cystic fibrosis

CM chylomicron(s)

CPT carnitine palmitoyl-CoA transferase

DHA docosahexaenoic acid (C22:6n-3)

DPA docosapentaenoic acid (C20:5n-3)

EFA essential fatty acid(s)

EFAD essential fatty acid-deficient

EFAS essential fatty acid-sufficient

EPA eicosapentaenoic acid (C20:5n-3)

FA fatty acid(s)

FABP fatty acid binding protein

FFA free fatty acid(s)

gGT gamma glutamyl transferase

HDL high density lipoprotein(s)

HL hepatic lipase

LA linoleic acid (C18:2n-6)

LCPUFA long-chain polyunsaturated fatty acid(s)

LPL lipoprotein lipase

MCT medium-chain triglycerides

Mdr multidrug resistance protein

MUFA monounstaurated fatty acid(s)

OLT orthotopic liver tranplantation

PC phosphatidyl choline

PL phospholipid(s)

PPARa peroxisome proliferator-activated receptor alpha

PUFA polyunsaturated fatty acid(s)

RBC red blood cell(s)

RDI recommended daily intake

SAFA saturated fatty acid(s)

SREBP sterol regulatory element binding protein

TG triglyceride(s)

VLDL very low density lipoprotein(s)


General introduction1

A. WernerF. Kuipers

H.J. Verkade

Part of this chapter was published under the title

"Fat absorption and lipid metabolism in cholestasis"

in: Molecular Pathogenesis of Cholestasis, 2003

M. Trauner, P.L.M. Jansen (Eds.), Landes Bioscience, p321-335.


ESSENTIAL FATTY ACIDS

Linoleic acid (LA, C18:2n-6) and alpha-linolenic acid (ALA, C18:3n-3) are poly-

unsaturated fatty acids (PUFA) that cannot be synthesized de novo by human or

animal cells. Nevertheless, these fatty acids are indispensable for normal

development and function and therefore they must be provided by the diet. Their

importance was first discovered in the 1920s by Mildred and George Burr, who found

that rat pups fed a lipid-free diet developed growth retardation, infertility, steatosis,

skin lesions and hair loss(1). Gradual reintroduction of lipid to the rats' diets failed to

alleviate these symptoms, and only when linoleic acid and alpha-linolenic acid were

supplied, symptoms disappeared. The importance of essential fatty acids for human

nutrition became apparent in the 1970s, with the introduction of total parenteral

nutrition (TPN) and infant formulas, that initially did not contain LA or ALA(2-5). In the

first part of this chapter, the structure of EFA will be described, their conversion into

LCPUFA and their functions and dietary requirements for the body. The second part

will focus on the processes involved in dietary lipid absorption and metabolism, and

the role of the liver herein, under physiological and cholestatic conditions.

EFA nomenclature and biosynthetic pathwaysThe structural formulas of LA and ALA are depicted below.

From the two "parent" essential fatty acids LA and ALA, two series of long-chain

polyunsaturated fatty acid (LCPUFA) metabolites are formed; the omega-6 or n-6

series, which is synthesized from LA, and the omega-3 or n-3 series which has ALA

as its precursor.

Formation of LCPUFA from EFA involves a series of alternating desaturation (inser-

tion of a double bond) and elongation (addition of two carbon atoms) reactions,

which occur predominantly in the endoplasmic reticulum of the liver(6;7).

10

Chapter 1

Essential fatty acids (EFA)

linoleic acid (LA) alpha-linolenic acid (ALA)C18:2n-6 C18:3n-3

H3C COOHH3C COOH

Essential fatty acids (EFA)

linoleic acid (LA) alpha-linolenic acid (ALA)C18:2n-6 C18:3n-3

H3C COOHH3C COOHH3C COOHH3C COOH

Figure 1


The omega- or n-number refers to the position of the terminal double bond, i.e., the

number of carbon atoms between the methyl-end of the fatty acid and the last dou-

ble bond. The number preceding the omega refers to the number of double bonds,

and the C-number indicates the number of carbon atoms in the fatty acid chain.

The different PUFA series compete for the same desaturation and elongation

enzymes, which have a preferential substrate affinity for n-3 fatty acids over n-6 fatty

acids over n-9 fatty acids. During scarcity of n-3 and n-6 fatty acids, long-chain n-9

fatty acids are synthesized. The concentration of n-9 LCPUFA in body compartments

is used to demarcate EFA deficiency(8). Desaturase enzymes are not only competed

for by the different fatty acid series, also feedback regulation occurs, mediated by the

concentrations of desaturase and elongase substrates and end-products.

In addition, hormonal and dietary factors regulate LCPUFA biosynthesis: desaturase

activities are enhanced by insulin, thyroid hormones and high dietary protein intake

and inhibited by glucagon, catecholamines, corticosteroids, and deficiencies of zinc,

iron, calcium, selenium and vitamins B6, C and E(9-11).

11

General introduction

Unsaturated fatty acids

EFA non-EFA

n-9 familyn-3 family n-6 family n-7 family

d6 desaturation

elongation

d5 desaturation

elongation

d6 desaturation

elongation

d5 desaturation

elongation

elongation

d6 desaturation

beta-oxidation

C18:3n-3 ALA

C18:4n-3

C20:4n-3

C20:5n-3 EPA

C22:5n-3

C24:5n-3

C24:6n-3

C22:6n-3 DHA

C18:2n-6 LA

C18:3n-6

C20:3n-6

C20:4n-6 AA

C22:4n-6

C24:4n-6

C24:5n-6

C22:5n-6

endo

plas

mic

retic

ulum

pero

xiso

mes

beta-oxidation

acylation C18:1n-9

C18:2n-9

C20:2n-9

C20:3n-9

C22:3n-9

C18:0

C16:1n-7

C16:0d9 desaturation


EFA non-EFA



EFA non-EFA


d6 desaturation

elongation

d5 desaturation

elongation

d6 desaturation

elongation

d5 desaturation

elongation

elongation

d6 desaturation

beta-oxidation

d6 desaturation

elongation

d5 desaturation

elongation

d6 desaturation

elongation

d5 desaturation

elongation

d6 desaturation

elongation

d5 desaturation

elongation

elongation

d6 desaturation

beta-oxidation

d6 desaturation

elongation

d5 desaturation

elongation

elongation

d6 desaturation

beta-oxidation

C18:3n-3 ALA

C18:4n-3

C20:4n-3

C20:5n-3 EPA

C22:5n-3

C24:5n-3

C24:6n-3

C22:6n-3 DHA

C18:2n-6 LA

C18:3n-6

C20:3n-6

C20:4n-6 AA

C22:4n-6

C24:4n-6

C24:5n-6

C22:5n-6

endo

plas

mic

retic

ulum

pero

xiso

mes

beta-oxidation

acylation C18:1n-9

C18:2n-9

C18:1n-9

C18:2n-9

C20:2n-9

C20:3n-9

C22:3n-9

C18:0

C16:1n-7

C16:0d9 desaturation

Figure 2


EFA functionsEFA and their long-chain metabolites fulfill a wide array of physiological functions in

the body; they are structural membrane components, precursors of eisosanoids and

ligands for nuclear receptors involved in lipid homeostasis. Similar to other fattyacids, EFA are also an important source of energy for the body; approximately 75%

of EFA is not converted into LCPUFA but is oxidized(12). Oxidation occurs in a

preferential rank order of ALA > LA > oleic acid > palmitic acid > stearic acid.LCPUFA are oxidized at a very low rate (13).

Membrane PLEFA and LCPUFA are structurally incorporated into cell membrane phospholipids

(PL). The degree of PL acyl chain unsaturation modulates fluidity of the membrane

lipid matrix, thus altering membrane permeability and function of membraneenzymes, receptors and transporters. Cellular membranes are typically constituted of

bilayers of PL molecules (Figure 4), in which cholesterol and membrane proteins are

inserted. Membrane PL are oriented with their hydrophobic acyl chains towards eachother, and their hydrophilic headgroups towards the aqueous intra- or extracellular

environment. Due to their amphiphilic nature, PL in membrane bilayers enable inter-

actions between water-soluble and lipid-soluble substances, while simultaneously

12

Chapter 1

hexadecanoic acid palmitic acid

octadecanoic acid stearic acid

eicosanoic acid arachidic acid

docosanoic acid behenic acid

tetracosanoic acid lignoceric acid

hexosanoic acid cerotic acid

9-hexadecenoic acid palmitoleic acid

11-octadecenoic acid vaccenic acid

9-octadecenoic acid oleic acid (OA)

5,8,11-eicosatrienoic acid Mead acid

9,12,15-octadecatrienoic acid alpha-linolenic acid (ALA)

5,8,11,14,17-eicosapentaenoic acid timnodonic acid (EPA)

4,7,10,13,16,19-docosahexaenoic acid cervonic acid (DHA)

5,8,11,14-eicosatetraenoic acid arachidonic acid (AA)

7,10,13,16,19-docosapentaenoic acid clupanodonic acid (DPA)

9,12-octadecadienoic acid linoleic acid (LA)

6,9,12-octadecatrienoic acid gamma-linolenic acid (GLA)

8,11,14-eicosatrienoic acid dihomo-gamma-linolenic acid (DGLA)

4,7,10,13,16-docosapentaenoic acid (DPA)

7,10,13,16-docosatetraenoic acid adrenic acid

C16:0

C18:0

C20:0

C22:0

C24:0

C26:0

C16:1n-7

C18:1n-7

C18:1n-9

C20:3n-9

C18:3n-3

C20:5n-3

C22:6n-3

C20:4n-6

C22:5n-3

C18:2n-6

C18:3n-6

C20:3n-6

C22:5n-6

C22:4n-6

hexadecanoic acid palmitic acid

octadecanoic acid stearic acid

eicosanoic acid arachidic acid

docosanoic acid behenic acid

tetracosanoic acid lignoceric acid

hexosanoic acid cerotic acid

9-hexadecenoic acid palmitoleic acid

11-octadecenoic acid vaccenic acid

9-octadecenoic acid oleic acid (OA)

5,8,11-eicosatrienoic acid Mead acid

9,12,15-octadecatrienoic acid alpha-linolenic acid (ALA)

5,8,11,14,17-eicosapentaenoic acid timnodonic acid (EPA)

4,7,10,13,16,19-docosahexaenoic acid cervonic acid (DHA)

5,8,11,14-eicosatetraenoic acid arachidonic acid (AA)

7,10,13,16,19-docosapentaenoic acid clupanodonic acid (DPA)

9,12-octadecadienoic acid linoleic acid (LA)

6,9,12-octadecatrienoic acid gamma-linolenic acid (GLA)

8,11,14-eicosatrienoic acid dihomo-gamma-linolenic acid (DGLA)

4,7,10,13,16-docosapentaenoic acid (DPA)

7,10,13,16-docosatetraenoic acid adrenic acid

C16:0

C18:0

C20:0

C22:0

C24:0

C26:0

C16:1n-7

C18:1n-7

C18:1n-9

C20:3n-9

C18:3n-3

C20:5n-3

C22:6n-3

C20:4n-6

C22:5n-3

C18:2n-6

C18:3n-6

C20:3n-6

C22:5n-6

C22:4n-6

C16:0

C18:0

C20:0

C22:0

C24:0

C26:0

C16:1n-7

C18:1n-7

C18:1n-9

C20:3n-9

C18:3n-3

C20:5n-3

C22:6n-3

C20:4n-6

C22:5n-3

C18:2n-6

C18:3n-6

C20:3n-6

C22:5n-6

C22:4n-6

Figure 3 : Nomenclature of polyunsaturated fatty acids

proefschrift_def_v0870605.qxp 7-6-2005 22:42 Pagina 12

allowing (sub)cellular compartmentalization. The PL bilayer provides an adaptable

matrix for insertion of membrane proteins, that function as receptors, membrane-

bound enzymes or transmembrane transporters. The nature of PL acyl chains can

significantly affect membrane properties. Saturated fatty acids (SAFA) form a more

rigid configuration, whereas double bonds make the membrane more flexible and

reactive.

CNS membrane PL

Flexibility and reactivity is particularly important in the highly excitable membranes of

the central nervous system (CNS), which are exceptionally rich in DHA (docosa-

hexaenoic acid, C22:6n-3) and AA (arachidonic acid, C20:4n-6). In fact, lipids con-

stitute 60% of brain dry weight, and 50% of PL acyl chains is DHA and AA(14). The sig-

nificant contribution of these fatty acids to CNS phospholipids has induced great

scientific interest in EFA and LCPUFA contents of infant formulas as compared to

breast milk. Human milk is a rich source of both n-3 and n-6 LCPUFA; it provides up

to 0.5% of fatty acids as DHA and AA, which are incorporated preferentially into brain

PL as compared to DHA and AA that originate from conversion of ingested precursor

EFA(15-17). Until recently, infant formulas only contained EFA and no LCPUFA. Both

preterm and term infants can convert EFA into DHA and AA(18;19), but it is questionable

whether this is sufficient to provide the high amounts of LCPUFA required by the

rapidly developing brain during its growth spurt, from 3 months before to 18 months

after birth. In utero, the placenta (which also has desaturase capacity) provides a

selective LCPUFA transfer to the fetus by an as yet unidentified transport mechanism,

resulting in up to 400-fold higher concentrations of DHA and AA in fetal compared to

maternal blood(20). A considerable deposition of LCPUFA in the brain occurs during

late gestation. Breast milk supplies DHA and AA in amounts believed to equal intra-

uterine accretion rates, whereas feeding infant formulas devoid of preformed LC-

PUFA has been associated with decreased brain DHA and AA contents and with

transiently impaired neurological maturation(21). Whether LCPUFA supplementation of

infant formulas, and in which amounts, has long-term beneficial effects on visual or

cognitive development in term infants is still a matter of debate.

Enterocyte and biliary PL

Whereas CNS membranes contain high amounts of DHA and AA, enterocyte mem-

branes and bile PL are particularly rich in LA and AA(22;23). The intestinal mucosa is a

dynamic structure that continuously undergoes biochemical and morphological

modifications during enterocyte differentiation and maturation. Their short life span

(~5 days in humans, ~2 days in rodents(24;25)) makes enterocytes highly sensitive to

13



changes in dietary lipids. A constant intestinal supply of EFA may be required for cell

renewal and maintenance of membrane integrity and function(26;27). Since bile phos-

pholipids contain up to 40% of acyl chains as EFA or LCPUFA, bile is a quanti-

tatively substantial supply of EFA for structural and functional needs of the

intestine(28). Enterocyte membranes are markedly EFA-depleted during EFA

deficiency. This might play a role in the well-recognized phenomenon that EFA

deficiency is not only a consequence of dietary lipid malabsorption, but can in itself

also cause intestinal lipid malabsorption(29-32). Intraluminal events involved in fat

absorption (lipolysis by lipases, solubilization of lipolytic products by bile and uptake

by enterocytes) seem undisturbed during EFA deficiency(32-35), yet qualitative changes

in enterocyte membranes due to LA and AA depletion could impair intracellular

processing of dietary lipid.

Eicosanoids

EFA derivatives are direct precursors for eicosanoids(36), which are involved in a wide

variety of inflammatory and vascular processes. Eicosanoids are synthesized from

C20 LCPUFA by cyclo-oxygenase and lipoxygenase enzymes, yielding

prostaglandins, thromboxanes, leukotrienes and lipoxins. Arachidonic acid (AA) is

the principal eicosanoid precursor in humans, next to dihomo-gamma-linolenic acid

(DGLA, C20:3n-6) and eicosapentaenoic acid (EPA, C20:5n-3). As autocrine and

paracrine hormones, eicosanoids mediate processes such as constriction or relax-

ation of endothelial cells, platelet aggregation, leucocyte activation and chemotaxis.

Eicosanoids derived from n-6 fatty acids generally have pro-inflammatory effects

whereas those with n-3 precursors are anti-inflammatory(37;38). The latter has led to

increased interest in EFA status, and particularly in n-3 to n-6 fatty acid balance, in

patients with cystic fibrosis(39) and autoimmune diseases.

Nuclear receptors, lipid homeostasis

In recent years, EFA and LCPUFA have gained interest as regulators of genes

involved in lipid homeostasis, by direct or indirect interactions with peroxisome pro-

liferator-activated receptor alpha (PPARa), sterol regulatory element binding protein

(SREBP), farnesoid X rexeptor (FXR) and liver X receptor (LXR). These nuclear recep-

tors regulate expression of genes involved in lipogenesis (fatty acid synthase (FAS),

acyl CoA carboxylase (ACC), stearyl CoA dehydrogenase (SCD)), lipid oxidation

(carnitine palmitoyl transferase (CPT), acyl CoA oxidase) and lipoprotein metabolism

(apoC2, C3, A1, E, scavenger receptor B1 (SR-B1), lipoprotein lipase (LPL), hepatic

lipase (HL), phospholipid transfer protein (PLTP), cholesterol ester transfer protein

(CETP), lecithin cholesterol acyl transferase (LCAT))(40-48). Fatty acids differ in their

14

Chapter 1


effects on plasma lipid and lipoprotein levels; generally, saturated fatty acids (SAFA)

increase plasma TG and cholesterol concentrations, whereas PUFA lower plasma

TG, yet species differences between animal models are considerable(49).

Enhancement of hepatic lipogenesis in EFA deficiency presumably results from

removal of inhibitory PUFA, but how EFA and LCPUFA, as core (TG) and surface (PL)

components of lipoproteins, affect physiological lipoprotein metabolism is not fully

clarified. An overview of EFA and LCPUFA functions is depicted in Figure 4.

EFA sources and requirementsLA can be derived from the diet or, in adults, to a limited extent from mobilization from

fat tissue. Healthy adults have approximately 1 kg of LA stored in adipose tissue, yet

adipocytes contain hardly any ALA or LCPUFA. Infants, and particularly preterm

neonates, have a limited adipose reserve, and are therefore especially dependent on

continuous EFA intake from dietary sources. LA is found in plant seed oils such as

corn oil and sunflower oil. ALA is present in green leafy vegetables, in nuts and in

soybean-, linseed- canola- and blackcurrant seed-oils. The n-3 LCPUFA DHA and

EPA are found in high concentrations in fatty fish such as mackerel, herring, salmon,

tuna and trout. Both DHA and AA are present in egg yolk(50), and AA is also found in

substantial concentrations in meat. As mentioned above, human milk is a rich source

of both n-6 and n-3 EFA and LCPUFA.

Recommendations regarding adequate dietary intake of EFA and LCPUFA are

highly variable between countries, and obviously vary with age, i.e., with adipose

tissue stores and with growth rate. For adults, the minimal daily requirement for LA

and ALA has been estimated at approximately 2 and 0.3 en%, respectively(51;52). For

healthy children, a daily intake of 1-5 en% of LA and 0.5 en% for ALA is

15


1 - energy

2 - membrane phospholipids

3 - eicosanoid precursors -prostaglandins-thromboxanes-leukotrienes

4 - lipoprotein metabolism

5 - ligands for transcription factors

1 - energy

2 - membrane phospholipids

3 - eicosanoid precursors -prostaglandins-thromboxanes-leukotrienes

4 - lipoprotein metabolism

5 - ligands for transcription factors

Figure 4: EFA functions


recommended(51;53). For preterm babies, recommendations for infant formulas are

currently 8-10% for LA, 1.5-1.75% for ALA, 0.5% for AA and 0.35% for DHA(54;55).

A deficiency of EFA can develop either when dietary intake is insufficient, when intes-

tinal absorption is impaired or when body requirements or metabolism of absorbed

EFA are (temporarily) increased. Thus, patients particularly prone for development of

EFA deficiency include (preterm) babies, patients with impaired dietary fat absorp-

tion, such as cholestatic patients or patients with chronic intestinal disease, and

patients with cystic fibrosis, in whom both absorption of EFA may be impaired and

metabolism may be increased. An overview of the clinical symptoms associated with

a deficiency of EFA is shown in Figure 5.

LIPID ABSORPTION AND METABOLISM IN CHOLESTASIS

In the various aspects of lipid absorption and metabolism, the liver has a central role.

Primarily, the liver produces bile, constituents of which are required for efficient

intestinal fat absorption. Additionally, biliary secretion of cholesterol (as such, or after

metabolism into bile salts) and phospholipids from the liver into the intestine is of

major importance in body lipid homeostasis. The liver is the major source of plasma

lipoproteins: it synthesizes apoproteins (i.e., apoA-I, apoB, apoE) that regulate meta-

bolic interconversions between lipoprotein classes and lipoprotein lipid constituents

as cholesterol, TG and PL. The liver is also the major site of clearance of circulating

lipoproteins, which are subsequently catabolized in the hepatocytes. Additionally, the

liver synthesizes enzymes such as LCAT, CETP, PLTP and LPL, which are involved in

lipoprotein metabolism in the plasma compartment. Finally, the liver is the site of

active synthesis, metabolism and/or oxidation of various lipid classes, including EFA

and LCPUFA.

16

Chapter 1

Growth retardation

Infertility

Increased perinatal mortality

Increased bleeding tendency

Impaired psychomotor development

Impaired cognitive development

Impaired visual development

Skin scalinesss, hair loss

Increased skin permeability to water

Steatosis

Dietary lipid malabsorption

Growth retardation

Infertility








Steatosis


Growth retardation

Infertility








Steatosis


Figure 5: Symptoms of EFA deficiency


In view of this multitude of essential functions that are often strongly interrelated, it is

evident that disturbances in bile formation in cholestatic liver disease will have a

strong impact on various aspects of lipid metabolism in the body. Consequences of

cholestasis, which is functionally defined as decreased or absent bile flow from the

liver into the intestine, may be related to:

- absence of bile components at their sites of action, particularly in the intestine

- disruption of the continuous flux of lipids from the liver into bile and intestine,

resulting in accumulation of toxic and non-toxic bile components in the body, most

notably in hepatocytes, concomitantly altering hepatocyte or enterocyte function.

- characteristic alterations in plasma lipoproteins associated with cholestatic liver

diseases, such as decreased HDL levels and the appearance of lipoprotein X.

Intestinal lipid absorptionDietary lipid classification

Dietary fat comprises a wide array of lipid classes, which have been categorized

according to the nature of their interactions with water into polar and non-polar

lipids(30;56;57). Polar lipids (cholesteryl esters, hydrocarbons and carotene), are insolu-

ble in water and are divided into three subclasses. Firstly the insoluble non-swelling

amphiphiles which form a thin stable monolayer in water; secondly the insoluble

swelling amphiphiles, which form both stable monolayers in water and laminated

17


feces

inte

stin

e

b i le

dietarylip id

cholesterol

bile sa lts

synthesis

ch

ylo

mic

ron

s

enterohepat iccircu latio n

liver

P L

H D L

VLDL

C E T P

PLTP

L C A T

Figure 6


lipid-water structures called liquid crystals; and finally the soluble amphiphiles, which

possess strong polar groups that render these molecules soluble in water at low

concentrations, forming both unstable monolayers and micelles(29;33). Examples of

class 1 polar lipids are triglycerides (TG), diacylglycerols (DG), non-ionized long-

chain fatty acids (LCFA), unesterified cholesterol and the fat soluble vitamins A, D,

E and K. Class 2 insoluble swelling amphiphiles are monoacylglycerols (MG), ionized

fatty acids (FA) and phospholipids (PL). Class 3 soluble amphiphiles are sodium salts

of long-chain fatty acids and bile salts. The absence of bile during cholestasis will

differentially affect solubilization and absorption of lipid classes due to their different

interactions with water.

TG is the major fat in human diet, contributing ~90% of energy provided by dietary

lipid. The majority of luminal phospholipid is phosphatidylcholine (PC), which is

mostly of biliary origin (10-20 g daily in humans), with a dietary contribution of 1-2 g

per day(29;30). EFA are present in the diet mostly as acyl chains of TG (90%) and also

as PL (10%). Biliary PL is an important source of intestinal EFA, since it is highly

EFA- enriched.

The predominant dietary sterol is cholesterol (0.5 g/day)(29;30), which is mostly of

animal origin although small amounts are also present in vegetables. Beta-sitosterol

is the most important plant sterol (which account for 25% of dietary sterols), but it is

virtually not absorbed by humans under physiological conditions due to the

intes-tinal half-transporters ABCG5/G8, which have been postulated to play a major

role in efflux of absorbed sterols from enterocytes into the intestinal lumen, and from

the liver into bile(58;59).

The fat-soluble vitamins A, D, E and K, required in small quantities for maintenance

of normal cell and organ function(60-63), are class 1 polar lipids and depend on micel-

lar solubilization for intestinal uptake. Absorption rates differ between vitamin

18

Chapter 1

OPO O

Ocholine

Dietary EFA

> 90% triglycerides (TG) < 10% phospholipids (PL)

OPO O

Ocholine

OPO O

OOPO

O

Ocholine

Dietary EFA

> 90% triglycerides (TG) < 10% phospholipids (PL)

Figure 7


species, averaging from 50 to 80% for vitamins A, D and K but only ~25% for vitamin

E(64). Additionally, there may be competition between vitamin species for intestinalabsorption- or transport-sites (65), although minimal information is available regarding

the nature and function of these sites.

The processes involved in intestinal lipid absorption can be divided into intraluminal

and intracellular events, which differ for TG and PL.

Intraluminal phase of lipid absorption

Before translocation from the intestinal lumen into enterocytes can occur, dietary

lipids must undergo a number of physicochemical alterations. This is achieved in asequence of events called the intraluminal phase of lipid digestion and absorption,

including:

-emulsification of dietary lipid-lipolysis

-solubilization (micelles, vesicles)

-translocation of lipolytic products across the enterocyte membrane

Emulsification and lipolysis

In humans, the first step in dietary fat digestion starts in the stomach with mechani-cal emulsification and partial TG hydrolysis by gastric lipase, resulting in the lipolytic

products DG and free fatty acids (FFA). Gastric lipase does not hydrolyze PL or

cholesterol ester, but its activity in the stomach accounts for 10-30% of TGlipolysis(30;56;66;67). The remaining part of TG digestion is brought about in the duodenal

lumen by pancreatic lipase, which acts mainly on the sn-1 and sn-3 position of TG

molecules, releasing 2-MG and FFA (66;68). Pancreatic (co-lipase-dependent) lipase ispresent in pancreatic secretions in large excess, in accordance with the clinical

observation that only severe pancreatic insufficiency (as in CF) results in lipid mal-

absorption. In the presence of bile salts, pancreatic lipase requires the cofactorpancreatic co-lipase for adequate TG hydrolysis, since TG droplets covered with bile

salts are not accessible to pancreatic lipase(56; 30). Binding of pancreatic co-lipase to

the TG/water interface facilitates binding of pancreatic lipase.

Digestion of PL occurs entirely in the duodenal lumen, predominantly by pancreatic

phospholipase A2. Phospholipase A2 requires calcium and bile salts for activation,and hydrolyzes PL at the sn-2 position resulting in FFA and lyso-PL.

Dietary cholesterol is present for ~90% as free cholesterol, the remainder as choles-

terol esters. Cholesterol esters are hydrolyzed in the small intestine by pancreatic

19



cholesterol esterase. Human cholesterol esterase (also known as carboxyl ester

lipase, bile salt-stimulated lipase, monoglyceride lipase, pancreatic non-specificlipase or human milk lipase) can also hydrolyze TG (sn-1, -2, -3), PL (sn-1, -2) and

lipidic vitamin esters(69) and its activity is enhanced by the presence of bile salts.

Solubilization of lipolytic products

For diffusion through the unstirred water layer, which separates the brush border

membrane of enterocytes from the liquid luminal contents of the intestine, solubiliza-tion of lipolytic products is required. The most important function of biliary bile salts

and PL in the intestinal lumen appears to be their ability to increase solubility of lipo-

lytic products in the aqueous lumen by formation of mixed micelles. Mixed micelleswere first described by Hoffman and Borgstrom (70) as disc-like aggregates of

amphiphilic biliary and dietary components, oriented with their hydrophobic parts to

the inside of the micelles and their hydrophilic polar headgroups towards the aque-ous outside. This conformation increases solubility of FFA and MG 100-1000 fold(29).

Mixed micelles contain bile salts (class 3 polar lipids), hydrogenated fatty acids

(class 1 polar lipids), fatty acid ions (class 2 polar lipids), MG (class 2 polar lipid), PL(class 2 polar lipids) and cholesterol (class 1 polar lipid), and are ~4 nm in diame-

ter(29;30;33;56;57). Carey(65) described the co-existence of mixed micelles in the intestinal

lumen with unilamellar liquid crystalline vesicles or liposomes. He demonstrated thatonly when intraluminal bile salt concentrations exceed the so-called critical micellar

concentration, mixed bile salt / lipid micelles are formed. However, when bile salt

concentrations are low or decreased by dilution, large (20-60 nm) unilamellar liquidcrystalline vesicles or liposomes predominate(71;72). All classes of lipolytic products

can be incorporated into disc-shaped micelles as well as liquid crystalline vesicles.

Since both phases co-exist, quantifying the relative contribution of the two phasesremains difficult, especially due to continuous exchange of 2-MG and FA between

both structures. Dissociation rates of lipolytic products from vesicles and their

subsequent translocation across the enterocyte membrane are slower thandissociation rates from mixed micelles(73).

The existence of liquid crystalline vesicles is thought to have specific pathophysio-logical consequences for lipid absorption in conditions where intraluminal bile salt

concentrations are diminished, as in cholestasis. It has been demonstrated that fat

uptake can still occur rather efficiently, based on balance studies, but that fat absorp-tion rates are profoundly slower in bile salt-deficient states. Porter et al. reported on

a bile fistula patient who continued to absorb up to 80% of dietary lipid, despite the

obvious bile salt deficiency and the 100-fold decreased FFA concentration in the

20

Chapter 1


aqueous phase of the small intestinal lumen(74). Mansbach et al. found similar results

in patients with bile salt malabsorption, where the strong decrease in solubilized fatty

acid concentration led only to a mild degree of lipid malabsorption(75). Solubilization

of lipolytic products into liquid crystalline vesicles during intestinal bile-salt

deficiency, in combination with the reserve capacity of the length of the small

intestine to absorb fat, could explain the slower but preserved rate of lipid absorption

in cholestasis.

Nishioka et al.(76) studied the importance of PL-cholesterol vesicles for lipid absorption

during bile deficiency. Intraduodenal administration of 13C-labeled linoleic acid (LA) or

palmitic acid (PA) to bile-diverted rats was associated with strongly decreased

plasma concentrations of 13C-LA and 13C-PA. Subsequent intraduodenal supplemen-

tation with PL-cholesterol vesicles reconstituted plasma concentrations of labeled PA.

Yet, there appeared to be a delay in plasma appearance of both lipids, since at 5h

after lipid administration plasma concentrations were still increasing. These obser-

vations are in concordance with the slower dissociation and translocation rates of

lipolytic products from vesicles compared to mixed micelles, as proposed by

Narayan and Storch(73).

It is important to note the lipid-class difference in the dependence on bile for solubi-

lization and consecutive uptake of fats. PUFA are less dependent on bile solubili-

zation due to their lower hydrophobicity, compared to SAFA. Even in complete

absence of intestinal bile salts, absorption of PUFA has been demonstrated to be rel-

atively well preserved (up to 80%) compared to that of saturated long-chain fatty

acids (<30%), although absorption remained significantly lower than in the presence

of bile (~97%, Figure 9)(25).

21


Pla

sma

13C

-pal

miti

c ac

id (%

adm

inis

tere

d do

se/m

l)

0.05

0

0.1

0 1 2 3 4 5 6

intact EHCPC+CH liposomebile deficient

Time after administration (hour)

Pla

sma

13C

-lino

leic

aci

d (%

adm

inis

tere

d do

se/m

l)

Time after administration (hour)

0

0.1

0.2

0.3

0 1 2 3 4 5 6

intact EHCPC+CH liposomebile deficient

Figure 8


In contrast to long-chain lipids, short- and medium-chain lipids do not depend on

luminal presence of bile salts for adequate uptake, and can directly be transferred

from the intestinal lumen through the enterocytes. Medium-chain triglyceride (MCT)-based formulas are therefore widely used as energy providers in conditions where

intestinal solubilization is impaired. For absorption of cholesterol and fat-soluble

vitamins, however, micellar solubilization by bile salts is crucial (77).

Translocation

For many years, translocation of lipolytic products across the unstirred water layerand the enterocyte membrane has been assumed to occur through passive diffusion.

Yet reports from Stremmel and Schulthess et al.(78;79) suggested that an active carrier-

mediated process by fatty acid binding proteins (Fabp) is involved in transport of FAacross the intestinal brush border membrane. Members of the Fabp family appeared

to have both fatty acid transport and esterification capacities(80;81), yet studies in mice

in which intestinal Fabp was genetically inactivated (Fabpi -/- mice) revealed that thisprotein is not essential for dietary fat absorption(82). Stahl et al. addressed FATP4,

abundantly present in apical membranes of enterocytes, as the principal intestinal

transporter of long-chain fatty acids(83). A fatty acid transporter not specific for theenterocyte was identified by Harmon et al.(84), who isolated a 88-kDa membrane

protein termed FAT (fatty acid transporter) which appeared to be the rat homologue

of human CD36. CD36 is expressed in platelets, macrophages and endothelial cellsas well as in intestine, adipose tissue, heart and muscle, where it mediates long-

chain fatty acid uptake. Impaired CD36 function is associated with a large (~70%)

deficit in FA uptake in these tissues(85). Some authors have suggested interactionsbetween different fatty acid transporters to regulate intestinal fatty acid uptake(86),

yet the exact molecular mechanism by which translocation of lipolytic products

occurs is still a matter of debate.

22

Chapter 1

0

50

100

16:0 18:0 18:1 18:2

Fatty acid species

Abs

orpt

ion

effic

ienc

y (%

)

Control rats

Cholestatic rats

**

**

0

50

100

16:0 18:0 18:1 18:2

Fatty acid species

Abs

orpt

ion

effic

ienc

y (%

)

Control rats

Cholestatic rats

**

**

Figure 9


The net uptake of cholesterol is highly specific, since the plant sterol ß-sitosterol is

poorly absorbed under physiological conditions despite its structural similarity to

cholesterol. In healthy individuals, ~60% of dietary cholesterol is taken up, whereas

absorption of plant sterols is less than 1%(29;87). The half-transporters ABCG5/G8,

implied in the autosomal recessive disorder sitosterolemia, are held responsible for

efficient efflux of absorbed dietary sterols from enterocytes into the intestinal lumen,

and from liver into bile(58;59), possibly with different affinities for sterol species.

Sitosterolemia patients, in whom ABCG5/G8 function is genetically impaired, have

30-fold increased plasma plant sterol levels, increased cholesterol absorption and

decreased biliary sterol secretion, resulting in sterol accumulation and athero-

sclerosis(88;89).

Intracellular phase of lipid absorption

After translocation across the apical enterocyte membrane, dietary lipids migrate to

the endoplasmic reticulum. At the cytosolic membrane of the smooth endoplasmic

reticulum (SER), re-esterification of absorbed fatty acids into TG takes place. Two

different biochemical pathways are involved in TG resynthesis, of which the mono-

acyl glycerol (MG) pathway is the most important under physiological conditions. In

the MG pathway, 2-MG is re-acylated to DG and subsequently to TG by monoacyl-

glycerol acyltransferase (MGAT) and diacylglycerol acyltransferase enzymes (DGAT1

and 2)(90;91), respectively. The alternative route of re-esterification is the alpha-

glycerophosphate pathway, which involves conversion of glycerol-3-phosphate via

phosphatidic acid to DG and then to TG, also mediated by DGAT. Under physiologi-

cal conditions there is an abundant supply of 2-MG and FFA during lipid absorption,

and the 2-MG route will predominate over the alpha-glycerophosphate route.

Newly synthesized TG from both pathways are thought to be metabolically distinct:

TG from the 2-MG route is secreted more rapidly across the basolateral membrane

than TG originated from the alpha-glycerophosphate route. It has been suggested

that the DG from each pathway enter into separate intracellular pools. DG from the

alpha-glycerophosphate route is preferentially used for de novo PL synthesis.

Absorbed cholesterol, either from biliary or dietary origin, enters the free cholesterol

pool inside the enterocyte, which also contains cholesterol originating from absorp-

tion of shed intestinal mucosal membranes and from de novo synthesis. Cholesterol

is transported into the lymphatic system mainly as cholesterol ester (CE) in the

neutral lipid core of chylomicrons. It is unclear whether a fraction of cholesterol

escapes the lymphatic route and crosses the basolateral membrane of the entero-

cyte via Abca1. The enzymes involved in cholesterol esterification are the acyl-CoA

23



cholesterol acyltransferases ACAT-1 and -2. Inhibition of ACAT activity decreases

absorption of dietary cholesterol, associated with lymphatic release of aberrant apoB

containing lipoproteins devoid of CE, containing mostly TG in their cores(90;92-94).

Newly synthesized TG and CE form lipid droplets in the SER, where packaging

occurs, mainly into lipoprotein particles called chylomicrons (CM). In the SER,

nascent chylomicrons associate with PL, cholesterol and apoA-I, apoA-IV and

apoB48. Under physiological conditions, surface coat PL of lymph chylomicrons are

of biliary origin rather than of dietary sources whereas chylomicron TG fatty acids

closely correspond with those of dietary TG(95). As fat absorption and TG resynthesis

proceeds, lipoprotein particles increase in size and number and eventually end up in

vesicles filled with pre-chylomicrons, which are transported to the Golgi apparatus.

Here, modification of pre-CM into mature CM occurs, followed by translocation to the

lateral surface of the enterocyte where CM are exocytosed into the interstitium,

ending up in mesenteric lymph. Nascent CM have diameters between 100-1000 nm.

Mesenteric lymph ducts drain into the thoracic duct, which enters the systemic

circulation at the level of the jugular vein.

In recent years, it has become appreciated that biliary PL secretion is necessary for

proper intestinal CM assembly and thus for secretion of dietary lipid into lymph.

Studies in rats with interrupted enterohepatic circulations, by cholestyramine feeding

or manipulation of bile composition by dietary means(96, 97), revealed a dietary lipid

accumulation in enterocytes. This phenomenon was also seen by Tso in bile-

diverted rats, where subsequent administration of bile acids only partially reinstated

lipid transport into lymph. Only after administrating biliary PL, lymphatic lipid trans-

port was fully restored(98).

Voshol et al. demonstrated a delayed plasma TG appearance after an oral lipid bolus

in Mdr2 -/- mice lacking biliary PL secretion. This aberrant postprandial plasma TG

response was accompanied by normal fecal fat excretion and accumulation of lipid

droplets in the intestinal wall, suggesting a relatively well-preserved intestinal lipid

uptake into enterocytes in the absence of biliary PL, but a delay in subsequent CM

secretion(99). Intestinal PL requirements for CM production might be comparable to

that of the liver for VLDL secretion. In choline deficiency, decreased hepatic PC syn-

thesis results in impaired VLDL production(100). Enterocytes might similarly require

biliary PL for appropriate intestinal CM production. A schematic overview of the

various steps involved in absorption of TG and PL is depicted in Figure 10.

24

Chapter 1


Absorption of enteral PL exceeds 90%, 20% of which is absorbed as intact PL, 40%

as lyso-PL, and 40%(101;102) is degraded to glycerophosphocholine and phosphoryl

choline, which is taken up via the portal vein. After absorption, PL acyl chains are

biologically highly available for the organism(103;104).

25


lymph TG

+

phospholipids bile acids

pancreas lipaseTG

2-MG FFA

ileum

BILE

micelle

+

chylomicron

lymph TG

+

phospholipids bile acids

pancreas lipaseTG

2-MG FFA

ileum

BILE

micelle

+

chylomicron

+phospholipase

1-lysoPC

FFA

+

chylomicron

OPO OO

choline

OPO OO

choline

OP

O

OO

ileum

PL

lymph PL

+phospholipase

1-lysoPC

FFA

+

chylomicron

OPO OO

OPO OO

choline

OPO OOOPO OO

choline

OP

O

OOO

PO

OO

ileum

PL

lymph PL

Figure 10a: Intestinal TG absorption

Figure 10b: Intestinal PL absorption


Alterations in lipid homeoastasis during cholestasisLipid malabsorption

Fat malabsorption as a consequence of disturbed bile secretion is associated with

weight loss due to energy deficiency (in children complicated by impaired growth

and development), with fat-soluble vitamin deficiencies and with EFA deficiency.

Several compensatory mechanisms for fat malabsorption during bile deficiency have

been described in animal models for cholestasis. Minich et al. reported on lipid mal-

absorption in rats with chronic bile diversion. In this model, bile is absent from the

intestinal lumen, but in contrast to the cholestatic condition, (toxic) biliary compo-

nents do not accumulate in the body. Bile-diverted rats appeared to compensate for

their markedly decreased dietary lipid absorption by strongly increasing their food

ingestion(25). Subsequent morphological examination of the small intestine revealed

that both villus and crypt height were significantly increased in bile-diverted rats

compared to controls.

Porter and Knoebel et al. described another compensatory mechanism designated

as ‘the absorptive reserve of the small intestine’(74;105). Under physiological conditions,

only the proximal part of the intestine is involved in fat absorption. In situations where

proximal fat absorption appears impaired, more distal parts can also contribute.

Minich et al. also demonstrated that the amount of dietary lipid strongly affects the

efficacy of lipid absorption in bile deficient states. In rats with chronic bile diversion,

absorption of dietary lipid remained highly efficient (84% of ingested lipid) when

regular low-fat chow was fed. However, when rats were fed a high-fat diet, lipid

absorption coefficients decreased to around 50%, indicating that compensatory

mechanisms for lipid absorption in bile deficiency have limited capacity(33). Detailed

knowledge of different compensation mechanisms and alternative routes of lipid

absorption during bile deficiency are important for developing dietary treatment

strategies for nutritional deficiencies in cholestatic patients.

Fat-soluble vitamin deficiency

Although great variability exists between studies in defining biochemical vitamin

deficiency, many reports have indicated the existence of significantly decreased fat-

soluble vitamin levels under cholestatic conditions(106-108). Fat-soluble vitamins (class I

polar lipids) are highly dependent on intraluminal solubilization by bile acids, and

lack of bile flow usually results in malabsorption and depletion of fat-soluble vitamin

stores. The extent of deficiency appears to be highly vitamin-species specific; Phillips

et al.(106) reported biochemical deficiencies of vitamin A, D, E and K in 34%, 13%, 2%

and 8% of PBC patients, respectively. Similar values were found by Kaplan and

Kowdley et al.(64;107;108). Vitamin A deficiency in chronic cholestasis can not only result

26

Chapter 1


from malabsorption, but hepatic secretion of retinol binding protein (RBP) may also

be diminished, leading to low plasma levels of retinol and impaired delivery to target

tissues such as retina and epithelial cells(60). Deficiency of vitamin K can lead to life-

threatening haemorrhages due to the vitamin K dependence of clotting factors II, VII,

IX, X and proteins S and C for haemostatic activity. Biliary atresia can present itself

with intracranial haemorrhage as first consequence of cholestasis-induced lipid

malabsorption(109). Although vitamin D deficiency during lipid malabsorption can par-

tially be circumvented by endogenous photosynthesis of vitamin D3 in the skin, many

chronically ill patients are not adequately exposed to sunlight, resulting in low

vitamin D and calcium levels and impaired bone mineralization(60;61). Prolonged

vitamin E deficiency in cholestatic children leads to degenerative neuromyopathy

eventually resulting in peripheral neuropathy, muscular weakness, ophthalmoplegia

and retinal dysfunction which appears partly irreversible. The irreversibility and

severity of many of the symptoms associated with fat-soluble vitamin deficiencies

mandate strict monitoring and correction of vitamin status in cholestatic patients.

EFA deficiency

EFA deficiency as a consequence of overall lipid malabsorption in cholestasis is well

recognized. Additionally, it has become apparent that EFA deficiency in itself can

impair lipid absorption. Levy et al. observed decreased bile salt secretion rates in

EFA-deficient rats(110), implying impaired bile formation as a possible cause for EFA

deficiency-induced fat malabsorption. EFA deficiency may also affect intracellular

events of dietary fat absorption occurring in the enterocyte. In both situations, EFA

deficiency during cholestasis may further compromise dietary fat absorption.

Lipid metabolism during cholestasis

During lipid absorption, both intestine and liver release large amounts of TG-rich

lipoproteins into the circulation. During fasting, the liver is the major source of TG-rich

lipoproteins by secreting VLDL. Both liver and intestine are capable of synthesizing

HDL, which are secreted as nascent particles containing predominantly PL and un-

esterified cholesterol. Another major lipoprotein, LDL, is predominantly formed in the

plasma compartment as a product of VLDL catabolism. Additionally, the liver synthe-

sizes apoproteins that are essential structural and enzymatic components of lipo-

proteins. Apoproteins act as cofactors for enzymes crucial for cholesterol esterifica-

tion or TG lipolysis. ApoA-I activates the cholesterol esterifying enzyme LCAT, which

is also synthesized in the liver. ApoC-II is required for lipoprotein lipase (LPL) activa-

tion, which hydrolyzes lipoprotein TG, thus converting CM into CM-remnants and

VLDL into IDL and ultimately LDL. ApoE and apoB are crucial for receptor-mediated

27



uptake of lipoproteins by peripheral cells, and for hepatic uptake of end products of

lipoprotein catabolism(111;112). Lipoprotein remnant uptake by the liver, mediated by

SR-B1 and hepatic lipase (HL), and dependent on lipoprotein type also by LDL-

receptor and LDL-receptor-related protein (LRP), provides a feedback inhibition

mechanism for cholesterol homeostasis by regulating activity of HMG-CoA

reductase, the key enzyme in hepatic cholesterol synthesis. Hepatocyte damage due

to toxic accumulation of bile acids in cholestasis may disrupt synthesis of apopro-

teins and other enzymes involved in lipoprotein formation and metabolism such as

LCAT, CETP and PLTP, with concomitant derangements in plasma and hepatic lipid

homeostasis which will be discussed below.

Lipoprotein X in cholestasis

Biliary excretion is the principal route for cholesterol disposal from the body (either

direct or after conversion into bile acids), and cholestasis thoroughly deranges body

sterol balance. The well-recognized increase in plasma free cholesterol observed in

some forms of cholestasis can be accompanied by an equimolar elevation of

plasma PL(113;114). Hypercholesterolemia in (extrahepatic) cholestasis is accompanied

by plasma appearance of the aberrant lipoprotein X (LpX)(115;116). LpX is a 40-100 nm

bilamellar vesicle with an aqueous lumen, predominantly composed of PL and free

cholesterol in equimolar amounts and containing only 3% of TG and 2% of choles-

teryl ester(117;118). Gradient ultracentrifugation revealed that LpX is isolated in the LDL-

fraction(119) and contains apoC and albumin. Manzato et al. hypothesized that LpX

particles represent biliary vesicles regurgitated from liver into plasma of cholestatic

subjects(114), since both LpX and bile vesicles are composed of PL and free choles-

terol. The presence of apoC and albumin, and the observation that the cholesterol-

PL ratio in LpX differs from that in bile, can be explained by plasma interactions of

LpX with other lipoproteins. LpX is not readily taken up by the liver, thus LpX-

cholesterol does not participate in feedback inhibition of hepatic cholesterol

synthesis. This could contribute to the paradox of increased hepatic cholesterol

neosynthesis in hypercholesterolemia during cholestasis. Felker observed LpX-like

vesicles in bile canaliculi of bile duct-ligated rats, indicating a biliary origin of the

particle(119-121). Oude Elferink et al. demonstrated the biliary origin of LpX in mice, since

in Mdr2 -/- mice, which secrete PL-free bile, bile duct ligation was not associated with

appearance of LpX(122).

HDL in cholestasis

Apart from increased lipid contents in the LDL fraction in the form of LpX, appearance

of TG-rich LDL and decreased plasma VLDL concentrations, chronic cholestasis is

28

Chapter 1


associated with strongly decreased plasma HDL concentrations (<10%)(123).

The underlying mechanism may involve either increased HDL clearance during

cholestasis, or decreased HDL synthesis. Recent work indicates that bile salts, accu-

mulating in hepatocytes during cholestasis, can suppress apoA-I gene transcription

via a negative farnesoid X receptor (FXR) response element mapped to the C-site of

the apoA-I promoter(124). As HDL is partly derived from CM surface remnants, low

plasma HDL levels could also result from decreased CM formation during intestinal

bile deficiency, or from defective HDL formation from CM surface remnants by

PLTP(125). Upon ultracentrifugation, HDL isolated in cholestasis is in the density range

of bilamellar discoidal particles, enriched in free cholesterol and PL with decreased

apoA-I and apoA-II contents and increased apoE, resembling so-called ‘nascent’

HDL particles. Such particles are normally not found in plasma in considerable

amounts because of rapid transformation by concerted actions of LCAT, CETP and

PLTP.

Lipoprotein-metabolizing enzymes in cholestasis

LCAT: Lecithin cholesterol acyl transferase (LCAT) and hepatic lipase (HL) are key

enzymes in lipoprotein metabolism. Both proteins are produced in the liver, but LCAT

is active in the circulation at the HDL surface whereas HL resides at the hepatic

endothelial cell lining. LCAT is a 60 kDa glycoprotein that converts cholesterol and

PL into cholesteryl esters and lyso-PL, and is activated by apoA-I. Its cholesterol

esterifying activity not only moves cholesterol from the HDL surface into the core,

thereby promoting the flux of cholesterol from cell membranes into HDL, but it also

leads to morphological changes of HDL particles. Nascent disc-shaped HDL

becomes spherical as cholesteryl esters accumulate in the HDL core. HL and LCAT

hydrolytic activities together account for over 80% of PL disappearance from

plasma(126). Impaired hepatic synthesis of these enzymes in cholestasis could thus

contribute to increased plasma PL concentrations. Both plasma cholesteryl ester and

LCAT concentrations are decreased in cholestatic subjects, and plasma appearance

of ‘nascent’ discoidal HDL particles is considered a direct result of defective LCAT

functioning. Furthermore, discoidal HDL has been associated with primary familial

LCAT deficiency(113).

CETP: Cholesteryl ester transfer protein (CETP) transfers excess cholesteryl esters

from HDL to VLDL and LDL in exchange for TG(127;128), thus participating in the so-

called reverse cholesterol transport. CETP activity results in homogenous fatty acid

species distribution between lipoprotein fractions. Activity of CETP is decreased 25%

in cholestasis, associated with a decreased LA content of VLDL-TG and CE

29



compared to HDL(127). Faust et al. demonstrated that fatty acid absorption regulates

CETP secretion in CaCo-2 cells(129;130), and several animal and human studies(131-134)

revealed that high-fat diets can increase CETP activity. Freeman et al. suggested that

serum TG levels >1.4 mmol/l are required for significant CETP-mediated lipid

exchange between LDL and VLDL(135). The mechanism for decreased CETP activity in

cholestatic subjects is not obvious because of the multiple origins of CETP

synthesis (liver, intestine, adipose tissue, macrophages)(136;137). However, since hepa-

tocytes are the predominant source of CETP, impaired hepatic CETP synthesis

remains a likely contributor to the decreased CETP activity in cholestasis(127).

PLTP: Plasma phospholipid transfer protein (PLTP) circulates bound to HDL and

mediates transfer of PL from apoB-containing lipoproteins into HDL, thus modulating

HDL size and lipid composition. PLTP activity generates pre-ß-HDL, the major accep-

tor of cholesterol in the reverse-cholesterol transport route. PLTP knockout mice have

markedly reduced HDL levels due to defective transfer of PL from TG-rich

lipoproteins into HDL(138). Liver, adipocytes and lung are presumably the major

sources of circulating PLTP. Impaired hepatic synthesis of PLTP in cholestatic

conditions can markedly reduce circulating HDL levels, and due to the stimulatory

effect of PLTP on CETP activity(139), it can further deteriorate the already impaired

CETP function. Recently, PLTP was identified as an FXR target gene(140), providing a

molecular basis for reduced PLTP gene expression under cholestatic conditions.

EFA metabolism in cholestasis

Socha et al. reported on decreased plasma arachidonic acid (AA) levels in pediatric

cholestatic patients, which was attributed to impaired hepatic microsomal desaturase

and/or elongase activity(141;142). However, Minich et al. demonstrated that conversion of

[13C]-LA to [13C]-AA was not significantly different in short-term bile duct-ligated rats

compared to controls. Accordingly, delta-6-desaturase activity determined in hepatic

microsomes was not altered(143). These results are in agreement with observations of

de Vriese et al., who found equal delta-9-, delta-6- and delta-3-desaturase activities

in liver microsomes of cholestatic and non-cholestatic rats(144). Decreased LA uptake

by cholestatic subjects appears the predominant cause of low plasma AA levels,

rather than impaired post-absorptive EFA metabolism. Cholestasis may also impair

hepatic ß-oxidative capacity, which on one hand may preserve EFA from oxidation

but on the other hand may inhibit the final step in forming C22:6n-3 and C22:5n-6. In

rats with long-term bile duct ligation impaired hepatic ß-oxidative capacity has been

reported(145).

30

Chapter 1


Nutritional therapy in cholestasis

Chronic cholestasis is often accompanied by nutritional deficiencies due to

inadequate dietary intake, maldigestion, malabsorption and/or defective metabolism

of nutrients. Additionally, requirements of energy or specific nutrients may increase

during cholestasis. The recommended caloric intake for chronic cholestatic patients

is ~130% of RDI, usually accomplished by dietary supplementation with glucose

polymers and/or MCT oil. For prevention or treatment of EFA deficiency, EFA-rich oils

are frequently recommended, but the effects hereof have remained un-

satisfactory(146;147). The enteral route is preferred for dietary supplementations but in

pediatric patients with severe chronic malabsorption, nasogastric and nocturnal

feedings are often required(127). For fat-soluble vitamin deficiency in cholestatic

children, adequate and rapid correction is required. To treat vitamin D deficiency, a

regimen of oral 25-OHD is recommended, at a dose of 2-4 µg/kg/d in children and

50-100 µg in adults, with regular measurements of cholecalciferol levels in plasma to

prevent toxicity. Vitamin K supplements of 2.5-5 mg 2-7 times a week are currently

recommended as prophylaxis for children with chronic cholestasis(60;61).

Most cholestatic children absorb the phylloquinone form of vitamin K to reach

functionally adequate levels, if supplied in high dosages. In adults, vitamin K supple-

ments are only recommended when blood tests suggest deficiency. For correction of

vitamin E deficiency, oral forms of vitamin E (alpha-tocopherol, alpha-tocopheryl

acetate, alpha-tocopheryl succinate) are recommended at doses from 10-25 IU/kg/d

up to 100-200 IU/kg/d. When normalization of plasma vitamin E levels is not reached,

a water-soluble form of vitamin E called d-alpha-tocopheryl polyethylene glycol-1000

succinate (TPGS) can improve vitamin E status in cholestatic patients(148). Argao et al.

demonstrated that absorption of other fat-soluble vitamins is greatly enhanced by

simultaneous administration with TPGS(149). Recommended dosage of vitamin A in

chronic cholestasis is 10000 IU if given with TPGS(148). Irrespective of the form of

vitamin supplementation that is chosen, plasma vitamin levels should be carefully

monitored to avoid excessive serum levels and toxicity.

ConclusionCholestatic liver disease can disturb many aspects of lipid absorption and metabo-

lism. Accumulation of potentially toxic bile components in hepatocytes due to dis-

ruption of the flux of bile from the liver to the intestine can damage hepatocytes,

resulting in impaired synthetic function and decreased production of enzymes

involved in lipoprotein metabolism. Also, lipoprotein secretion is disturbed during

cholestasis, reflected by decreased HDL levels and appearance of the aberrant

lipoprotein X in plasma. The absence of biliary components from the intestinal lumen

31



during cholestasis can strongly impair uptake of dietary fat and fat-soluble vitamins,

resulting in nutritional deficiencies such as EFA deficiency and vitamin A, D, E and K

deficiency. Current prolonged survival of patients with chronic cholestasis will require

critical evaluation of nutrient deficiencies and adequate treatment strategies in order

to prevent sequelae and to improve quality of life.

SCOPE OF THIS THESIS

Essential fatty acids and their long-chain polyunsaturated metabolites are crucial for

normal function and development of human and animal cells. Low levels of essential

fatty acids in the body are associated with dietary fat malabsorption, steatosis and

impaired neurological development in infants. Pediatric conditions with lipid malab-

sorption (cholestasis, cystic fibrosis, prematurity) can be complicated by EFA

deficiency. A deficiency of essential fatty acids is common in children with

cholestatic liver disease, despite high intakes of dietary EFA. Dietary EFA are

predominantly present in the form of triglycerides (TG), which are malabsorbed

during cholestasis, whereas phospholipids (PL) are relatively well absorbed during

bile deficiency, and have been postulated to have a high post-absorptive

bioavailability.

This thesis focuses on the role of oral phospholipids (PL) as compared to

triglycerides (TG) as vehicles for EFA supplementation under cholestatic conditions

in experimental animals (mice) and pediatric patients with end-stage liver disease.

We aim to clarify the role of EFA-rich phospholipids in intestinal and hepatic

lipoprotein formation, and in the development of fat malabsorption and steatosis

during EFA deficiency. Also, we aim to evaluate the role of fat malabsorption and EFA

metabolism on the frequently observed alterations of EFA status in cystic fibrosis, a

condition also often associated with EFA deficiency, by analyzing EFA concentrations

and metabolism in different body compartments of mouse models for CF.

32

Chapter 1


REFERENCES1. Burr GO, Burr MM. A new deficiency disease produced by the rigid exclusion of fat from the diet. JBC 1929; (2):345-367.2. Hansen AE, et al. Role of LA in infant nutrition. Pediatrics 1963; 31(1):171-92.3. Richardson TJ, Sgoutas D. EFA deficiency in 4 adult patients during total parenteral nutrition. AJCN 1975; 28:258-263.4. Paulsrud JR, Pensler L, Whitten CF, Stewart S, Holman RT. EFA deficiency in infants induced by fat-free intravenous feeding.Am J Clin Nutr 1972; 25:904.5. Caldwell MD, Johnsson H, Otherson H. EFA deficiency in infants receiving i.v. fat-free alimentation. J Pediatr 1972; 81:894-8.6. Sprecher H. Biochemistry of essential fatty acids. Prog Lipid Res 1981; 20:13-24.7. Brenner RR. The desaturation steps in animal biosynthesis of PUFA. Lipids 1971; 6:567-575.8. Holman RT. The ratio of trienoic:tetraenoic acids in tissue lipids as a measure of EFA requirement. J Nutr 1960; 70:405-410.9. Brenner RR. Nutritional and hormonal factors influencing desaturation of EFA in animals. Prog Lipid Res 1981; 20(41):47.10. Brenner RR. Hormonal modulation of d6 and d5 desaturases. Prostaglandins Leukot Essent Fatty Acids 2003; 68:151-162.11. Infante JP. Vitamin E and selenium participation in fatty acid desaturation. Mol Cell Biochem 1986; 69:93-108.12. Cunnane S. Problems with EFA: time for a new paradigm? Prog Lipid Res 2003; 42(6):544-568.13. Leyton J, et al. Differential oxidation of saturated and unsaturated fatty acids in vivo in the rat. Br J Nutr 1987; 57:383-393.14. Innis SM. Perinatal biochemistry and physiology of LCPUFA. J Pediat 2003;143:S1-8.15. Makrides M, Neumann MA, Byard RW. Fatty acid composition of brain, retina and RBC in breast- and formula-fed infants.Am J Clin Nutr 1994; 60:189-194.16. Uauy R, Mena P, Wegher B. Long-chain polyunsaturated fatty acid formation in neonates. Pediatr Res 2000; 47:127-135.17. Crawford MA. Placental delivery of AA and DHA. AJCN 2000; 71:S275-84.18. Carnielli VP, Wattimena DJ, Luijendijk IH, Boerlage A, Degenhart HJ, Sauer PJ. The VLBW premature infant is capable ofsynthesizing AA and DHA from linoleic and linolenic acids. Pediatr Res 1996; 40(1):169-174.19. Uauy R, Peirano P, Hoffman D, Mena P, Birch D, Birch E. Role of EFA in the function of the developing nervous system. Lipids1996; 31 Suppl:S167-76.:S167-S176.20. Elias SL, Innis SM. Infant plasma trans, n-6, and n-3 fatty acids and conjugated linoleic acids are related to maternal plas-ma fatty acids, length of gestation, and birth weight and length. Am J Clin Nutr 2001; 73(4):807-814.21. Lauritzen L, Hansen HS, Jorgensen MH, Michaelsen KF. The essentiality of long-chain n-3 fatty acids in relation to devel-opment and function of the brain and retina. Prog Lipid Res 2001; 40(1-2):1-94.22. Boehm G, et al. DHA and AA content of serum and RBC membrane PL of preterm infants fed breast milk, standard formu-la or formula supplemented with n-3 and n-6 LCPUFA. Eur J Pediatr 1996; 155:410-416.23. Horrobin DF, Huang YS, Cunnane SC, Manku MS. EFA in plasma, RBC and liver PL in common laboratory animals as com-pared to humans. Lipids 1984; 19(10):806-811.24. van Dorp D. Prostaglandin synthesis. Acta Biol Med Ger 1976; 35:1041-1049.25. Minich DM, Kalivianakis M, Havinga R, Van Goor H, Stellaard F, Vonk RJ et al. Bile diversion in rats leads to a decreasedplasma concentration of LA which is not due to decreased net intestinal absorption of dietary LA. BBA 1999; 1438(1):111-119.26. Enser M, et al. The effect of EFA deficiency on the fatty acid composition of the total lipid of the intestine. Biochem J 1962;85:607-614.27. Snipes RL. The effects of EFA deficiency on the ultrastructure and functional capacity of the jejunal epithelium. Lab Invest1968; 18(2):179-189.28. Melin T, Qi C, Nilsson A. Bile but not chyle lipoprotein is an important source of AA for the rat small intestine. Prostagl LeukotEssent Fatty Acids 1996; 55(5):337-343.29. Tso P. Intestinal lipid absorption. In: Physiology of the gastrointestinal tract. New York: Raven Press, 1994: 1867-1907.30. Carey MC, Hernell O. Digestion and absorption of fat. Seminars in gastrointestinal disease 1992; 3(4):189-208.31. Friedman HI, Nylund B. Intestinal fat digestion, absorption and transport. AJCN 1980;(33):1108-1139.32. Levy E, Garofalo C, Thibault L, Dionne S, Daoust L, Lepage G. Intraluminal and intracellular phases of fat absorption areimpaired in EFA deficiency. Am J Physiol 1992; 262(2 Pt 1):G319-G326.33. Verkade HJ, Tso P. Biophysics of intestinal luminal lipids. In: Tso P, Mansbach CM, Kuksis A, editors. Intestinal lipid metab-olism. New York: Kluwer, 2000: 1-18.34. Minich DM, Kuipers F, Vonk RJ, Verkade HJ. The role of bile in EFA absorption and metabolism. In: Riemersma RA,Armstrong R, Kelly RW, Wilson R, editors. EFA and eicosanoids-4th international congress. Champaign, Illinois, 1998: 43-47.35. Bennett Clark S, Ekkers TE, Singh A, Balint JA, Holt PR, Rodgers JB. Fat absorption in EFA deficiency: a model experi-mental approach to studies of the mechanism of fat malabsorption of unknown etiology. JLR 1973; 14:581-588.36. Lipkin M. Proliferation and differentiation of normal and diseased gastrointestinal cells. Johnson LR, Physiology of the gas-trointestinal tract. Raven Press,1987:255-84.37. Smith W.Prostanoid biosynthesis and mechanisms of action. AJP 1992; 263:F181-91.38. Strandvik B, Svensson E, Seyberth HW. Prostanoid biosynthesis in patients with cystic fibrosis. Prostaglandins LeukotEssent Fatty Acids 1996; 55:419-425.39. Van Biervliet S, Van Biervliet JP, Robberecht E, Christophe A. DHA trials in CF: A review of the rationale behind the clinicaltrials. J Cyst Fibros 2005; 4(1):27-34.

33



40. Zhao A, Yu J, Lew JL, Huang L, Wright SD, Cui J. PUFA are FXR ligands and differentially regulate expression of FXR tar-gets. DNA Cell Biol 2004; 23(8):519-526.41. Xu J, et al. SREBP-1 expression is suppressed by dietary PUFA. J Biol Chem 1999; 274(33):23577-23583.42. Dobrosotskaya IY, Seegmiller AC, Brown MS, Goldstein JL, Rawson RB. Regulation of SREBP processing and membranelipid production by PL in Drosophila. Science 2002; 296(5569):879-883.43. Staels B, Fruchart JC, Auwerx J. Fatty acids as transcriptional regulators of the LPL and apoC-III genes: consequences forplasma TG metabolism. In: Davignon J, Fruchart JC, Ordovas JM, editors. Omega-3, lipoproteins and atherosclerosis. Paris:John Libbey Eurotext, 1996: 41-48.44. Edwards PA, et al. BAREing it all: adoption of LXR and FXR and their roles in lipid homeostasis. JLR 2002; 43(1):2-12.45. Lambert G, Amar MJ, Guo G, Brewer HB, Jr., Gonzalez FJ, Sinal CJ. FXR is an essential regulator of cholesterol home-ostasis. JBC 2003; 278(4):2563-2570.46. Xu J, Teran-Garcia M, Park JH, Nakamura MT, Clarke SD. PUFA suppress hepatic SREBP-1 expression by accelerating tran-script decay. JBC 2001; 276(13):9800-7.47. Pawar A, et al.. The role of LXR-alpha in the fatty acid regulation of hepatic gene expression. J Biol Chem 2003; .48. Ou J, Tu H, Shan B, Luk A, DeBose-Boyd RA, Bashmakov Y et al. Unsaturated fatty acids inhibit transcription of the SREBP-1c gene by antagonizing ligand-dependent activation of the LXR. Proc Natl Acad Sci U S A 2001; 98(11):6027-6032.49. Smit MJ, Temmerman AM, Wolters H, Beynen AC, Vonk RJ, Kuipers F. Effects of dietary fish oil on hepatic cholesterol metab-olism. In: Windler E, Greten H, editors. Hepatic Endocytosis of Lipids and Proteins. Zuckschwerdt Verlag, 1992: 380-388.50. Simopoulos AP, Salem N, Jr. Egg yolk as a source of LCPUFA in infant feeding. Am J Clin Nutr 1992; 55(2):411-414.51. Innis SM. EFA in growth and development. Prog Lipid Res 1991; 30(1):39-103.52. Bjerve KS, Lovold-Mostad I, Thoresen L. ALA deficiency in patients on long-term gastric tube feeding: an estimation of LAand long-chain unsaturated n-3 fatty acid requirement in man. Am J Clin Nutr 1987; 45:66-77.53. Holman et al. A case of human LA deficiency involving neurological abnormalities. Am J Clin Nutr 1982; 35(3):617-623.54. Simopoulos AP, Leaf A, Salem N. Workshop statement on the essentiality of and recommended dietary intakes for omega-6 and omega-3 fatty acids. Prostaglandins Leukot Essent Fatty Acids 2000; 63:119-121.55. Cunnane S, Francescutti V, Brenna JT. DHA requirement and infant development. Nutrition 1999; 15:801-802.56. Carey MC, Small DM, Bliss CM. Lipid digestion and absorption. Ann Rev Phys 1983;45:651-77.57. Carey MC, Small DM. The characteristics of mixed micellar solutions with particular reference to bile. Am J Med 1970;49:590-608.:590-608.58. Lee MH, Lu K, Hazard S, Yu H, Shulenin S, Hidaka H et al. Identification of a gene, ABCG5, important in regulation of dietarycholesterol absorption. Nat Genet 2001; 27(1):79-83.59. Berge KE, Tian H, Graf GA, Yu L, Grishin NV, Schultz J et al. Accumulation of dietary cholesterol in sitosterolemia causedby mutations in adjacent ABC transporters. Science 2000; 290(5497):1771-5.60. Sokol RJ. Fat-soluble vitamins and their importance in patients with cholestatic liver diseases. Gastroenterol Clin North Am1994; 23(4):673-705.61. Sokol RJ. Vitamin deficiency and replacement in childhood cholestasis. In: Lentze M, Reichen J, editors. pediatric cholesta-sis. Kluwer Academic Publishers 1991: 289-303.62. Argao EA, Heubi JE. Fat-soluble vitamin deficiency in infants and children. Curr Opin Pediatr 1993; 5(5):562-566.63. Jeppesen PB, Hoy CE, Mortensen PB. Deficiencies of EFA, vitamin A and E and changes in plasma lipoproteins in patientswith reduced fat absorption or intestinal failure. European Journal of Clinical Nutrition 2000; 54:632-642.64. Meydani M, Martin KR. Intestinal absorption of fat-soluble vitamins. Mansbach, Tso, Kuksis: Intestinal lipid metabolism.Kluwer academic publishers, 2000: 367-78.65. Hollander D. Intestinal absorption of vitamins A, E, D, K. J Lab Clin Med 1981; 97(4):449-62.66. Borgstrom B. Digestion and absorption of lipids. Int Rev Physiol 1977; 12:305-23.:305-323.67. Liao TH, Hamosh P, Hamosh M. Fat digestion by lingual lipase. Pediatr Res 1984; 18(5):402-409.68. Mattson FH, Volpenhein RA. Carboxylic ester hydrolases of rat pancreatic juice. J Lipid Res 1966; 7(4):536-543.69. Lombardo D, et al.Studies on the substrate specificity of a carboxyl ester hydrolase from human pancreatic juice. BBA 1980;611(1):147-155.70. Hofmann AF, Borgstrom B. The intraluminal phase of fat digestion in man. J Clin Invest 1964; 43:247-257.71. Staggers JE, Hernell O, Stafford RJ, Carey MC. Physical-chemical behavior of dietary and biliary lipids during intestinaldigestion and absorption 1. Biochem 1990; 29(8):2028-40.72. Hernell O, Staggers JE, Carey MC. Physical-chemical behavior of dietary and biliary lipids during intestinal digestion andabsorption 2. Biochem 1990; 29(8):2041-2056.73. Narayanan VS, Storch J. Fatty acid transfer in taurodeoxycholate mixed micelles. Biochemistry 1996; 35(23):7466-7473.74. Porter HP, et al. Fat absorption in bile fistula man. Gastroenterology 1971; 60(6):1008-1019.75. Mansbach CM, Newton D, Stevens RD. Fat digestion in patients with bile acid malabsorption but minimal steatorrhea. DigDis Sci 1980; 25(5):353-362.76. Nishioka T, Havinga R, Tazuma S, Stellaard F, Kuipers F, Verkade HJ. Administration of PC-cholesterol liposomes partiallyreconstitutes fat absorption in chronically bile-diverted rats. Biochim Biophys Acta 2004; 1636(2-3):90-98.

34

Chapter 1


77. Voshol PJ, Havinga R, Wolters H, Ottenhoff R, Princen HM, Oude Elferink RP Reduced plasma cholesterol and increasedfecal sterol loss in mdr2 P-glycoprotein-deficient mice. Gastroenterology 1998; 114(5):1024-1034.78. Stremmel W, Lotz G, Strohmeyer G, Berk PD. Identification, isolation, and partial characterization of a fatty acid binding pro-tein from rat jejunal microvillous membranes. J Clin Invest 1985; 75(3):1068-1076.79. Schulthess G, Lipka G, Compassi S, Boffelli D, Weber FE, Paltauf F. Absorption of monoacylglycerols by small intestinalbrush border membrane. Biochemistry 1994; 33(15):4500-4508.80. Coe NR, Smith AJ, Frohnert BI, Watkins PA, Bernlohr DA. FATP1 is a very long-chain acyl-CoA synthetase. J Biol Chem1999; 274(51):36300-36304.81. Watkins PA, Pevsner J, Steinberg SJ. Human very long-chain acyl-CoA synthetase and two human homologs: initial char-acterization and relationship to fatty acid transport protein. Prostaglandins Leukot Essent Fatty Acids 1999; 60(5-6):323-328.82. Vassileva G, Huwyler L, Poirier K, Agellon LB, Toth MJ. FABP-i is not essential for dietary fat absorption in mice. FASEB J2000; 14(13):2040-2046.83. Stahl A, et al. Identification of the major intestinal fatty acid transport protein. Mol Cell 1999; 4(3):299-308.84. Harmon CM, Abumrad NA. Binding of sulfosuccinimidyl fatty acids to adipocyte membrane proteins: isolation and amino-terminal sequence of an 88-kD protein implicated in transport of long-chain fatty acids. J Membr Biol 1993; 133(1):43-49.85. Ibrahimi A, et al Role of CD36 in membrane transport of LCFA. Curr Opin Clin Nutr Metab Care 2002; 5(2):139-145.86. Abumrad NA, Sfeir Z, Connelly MA, Coburn C. Lipid transporters: membrane transport systems for cholesterol and fattyacids. Curr Opin Clin Nutr Metab Care 2000; 3(4):255-262.87. Salen G, Ahrens EH, Jr., Grundy SM. Metabolism of beta-sitosterol in man. J Clin Invest 1970; 49(5):952-967.88. Gregg RE, Connor WE, Lin DS, Brewer HB, Jr. Abnormal metabolism of shellfish sterols in a patient with sitosterolemia andxanthomatosis. JCI 1986;77(6):1864-1872.89. Bhattacharyya AK, Connor WE. Beta-sitosterolemia and xanthomatosis. A newly described lipid storage disease in two sis-ters. J Clin Invest 1974; 53(4):1033-1043.90. Buhman KF, Accad M, Farese RV. Mammalian acyl-CoA:cholesterol acyltransferases. BBA 2000; 1529(1-3):142-154.91. Chang CC, Sakashita N, Ornvold K, Lee O, Chang ET, Dong R. Immunological quantitation and localization of ACAT-1 andACAT-2 in human liver and small intestine. J Biol Chem 2000; 275(36):28083-28092.92. Martins IJ, Mortimer BC, Redgrave TG. Effect of the ACAT inhibitor CL 277,082 on apolipoprotein B48 transport in mesen-teric lymph and on plasma clearance of chylomicrons and remnants. Arterioscler Thromb Vasc Biol 1997; 17(1):211-216.93. Martins IJ, Mortimer BC, Miller J, Redgrave TG. Effects of particle size and number on plasma clearance of chylomicronsand remnants. JLR 1996; 37(12):2696-2705.94. Mortimer BC, Holthouse DJ, Martins IJ, Stick RV, Redgrave TG. Effects of triacylglycerol-saturated acyl chains on the clear-ance of chylomicron-like emulsions from the plasma of the rat. Biochim Biophys Acta 1994; 1211(2):171-180.95. Patton GM, Bennett Clark S, Fasulo JM, Robins SJ. Utilization of individual lecithins in intestinal lipoprotein formation in therat. J Clin Invest 1984; 73(1):231-240.96. Cassidy MM, Lightfoot FG, Grau L, Satchitanandum S, Vahouny GV. Lipid accumulation in jejunal and colonic mucosa fol-lowing chronic cholestyramine (Questran) feeding. Dig Dis Sci 1985; 30(5):468-476.97. Arjmandi BH, Ahn J, Nathani S, Reeves RD. Dietary soluble fiber and cholesterol affect serum cholesterol concentration,hepatic portal venous short-chain fatty acid concentrations and fecal sterol excretion in rats. J Nutr 1992; 122(2):246-253.98. Tso P, Kendrick H, Balint JA, Simmonds WJ. Role of biliary PC in the absorption and transport of dietary triolein in the rat.Gastroenterology 1981; 80(1):60-65.99. Voshol PJ, Minich DM, Havinga R, OudeElferink RPJ, Verkade HJ, Groen AK. Postprandial chylomicron formation and fatabsorption in mdr2 p-glycoprotein-deficient mice. Gastroenterology 2000; 118:173-182.100. Verkade HJ, Fast DG, Rusinol AE, Scraba DG, Vance DE. Impaired biosynthesis of PC causes a decrease in number ofVLDL particles in the Golgi but not in the endoplasmic reticulum of rat liver. J Biol Chem 1993; 268(33):24990-24996.101. Zierenberg O, Grundy SM. Intestinal absorption of polyenephosphatidylcholine in man. J Lipid Res 1982; 23(8):1136-1142.102. Bloom B, et al. Absorption of plasma PL. Manner of transport from intestinal lumen to lacteals. AJP 54 A.D.; 177:84-86.103. Lieber CS, Robins SJ, Li J, DeCarli LM, Mak KM, Fasulo JM. PC protects against fibrosis and cirrhosis in the baboon.Gastroenterology 1994; 106(1):152-159.104. Fox JM. Polyene phosphatidylcholine: pharmacokinetics after oral administration In: Avogaro P, Macini M, Ricci G, PaolettiR, ed. PL and atherosclerosis. Raven, 1983.105. Knoebel LK. Intestinal absorption in vivo of micellar and nonmicellar lipid. Am J Physiol 1972; 223(2):255-261.106. Philips JR et al. Fat-soluble vitamin levels in patients with PBC. Am J Gastroenterology 2001; 96:(9):2745-2750.107. Kowdley KV, Emond MJ, Sadowski JA, Kaplan MM. Plasma vitamin K1 level is decreased in primary biliary cirrhosis. AmJ Gastroenterol 1997; 92(11):2059-2061.108. Kaplan MM, Elta GH, Furie B, Sadowski JA, Russell RM. Fat-soluble vitamin nutriture in primary biliary cirrhosis.Gastroenterology 1988; 95(3):787-792.109. van den Anker JN, Sinaasappel M. Bleeding as presenting symptom of cholestasis. J Perinatol 1993; 13(4):322-324.110. Levy E, Garofalo C, Rouleau T, Gavino V, Bendayan M. Impact of essential fatty acid deficiency on hepatic sterol metabo-lism in rats. Hepatology 1996; 23(4):848-857.111. Sabesin SM, Bertram PD, Freeman MR. Lipoprotein metabolism in liver disease. Adv Intern Med 1980; 25:117-46.:117-146.

35



112. Brown MS, Goldstein JL. Lipoprotein receptors in the liver. J Clin Invest 1983; 72(3):743-747.113. Miller JP. Dyslipoproteinaemia of liver disease. Baillieres Clin Endocrinol Metab 1990; 4(4):807-832.114. Manzato E, et al Formation of lpX. Its relationship to bile compounds. J Clin Invest 1976; 57(5):1248-1260.115. Soros P, et al. LpX in patients with cirrhosis. Hepatology 1998; 28(5):1199-1205.116. Seidel D, Alaupovic P, Furman RH. A lipoprotein characterizing obstructive jaundice. J Clin Invest 1969; 48(7):1211-1223.117. Kostner GM, Laggner P, Prexl HJ, Holasek A. Investigation of the abnormal LDL occurring in patients with obstructive jaun-dice. Biochem J 1976; 157(2):401-407.118. Laggner P, et al. The lipid bilayer structure of the abnormal human plasma LpX. Eur J Biochem 1977; 77(1):165-171.119. Felker TE, et al. Secretion of LpX by perfused livers of rats with cholestasis. PNAS USA 1978; 75(7):3459-3463.120. Crawford AR, Smith AJ, Hatch VC, Oude Elferink RP, Borst P, Crawford JM. Hepatic secretion of PL vesicles in the mousecritically depends on mdr2 or MDR3 P-glycoprotein expression. J Clin Invest 1997; 100(10):2562-2567.121. Smith AJ, de Vree JM, Ottenhoff R, Oude Elferink RP, Schinkel AH, Borst P. Hepatocyte-specific expression of the humanMDR3 P-glycoprotein gene restores the biliary PC excretion absent in Mdr2 (-/-) mice. Hepatology 1998; 28(2):530-536.122. Oude Elferink RP, Ottenhoff R, van Marle J, Frijters CM, Smith AJ, Groen AK. Class III P-glycoproteins mediate formationof LpX in the mouse. JCI 1998;102(9):1749-57.123. Tallet F, et al. Characterization of lipoproteins during human cholestasis. Clin Chim Acta 1996; 244(1):1-15.124. Claudel T, Sturm E, Duez H, Torra IP, Sirvent A, Kosykh V et al. Bile acid-activated nuclear receptor FXR suppresses apoA-I transcription via a negative FXR response element. J Clin Invest 2002; 109(7):961-971.125. Huuskonen J, et al. The impact of PLTP on HDL metabolism. Atherosclerosis 2001; 155(2):269-281.126. Shamburek RD, Zech LA, Cooper PS, Vandenbroek JM, Schwartz CC. Disappearance of two major phosphatidylcholinesfrom plasma is predominantly via LCAT and hepatic lipase. Am J Physiol 1996; 271(6 Pt 1):E1073-E1082.127. Korsten MA, Lieber CS. Nutrition in pancreatic and liver disorders. In: Shils ME, Olson JA, Shike M, Ross CA, editors.Modern nutrition in health and disease. Lippincott Williams & Wilkins, 1999: 1066-1079.128. Swenson TL, et al.. Plasma CETP has binding sites for neutral lipids and phospholipids. JBC 1988; 263(11):5150-5157.129. Faust RA, Albers JJ. Synthesis and secretion of plasma CETP by human HepG2 cells. Arteriosclerosis 1987; 7(3):267-275.130. Faust RA, Albers JJ. Regulated vectorial secretion of CETP (LTP-I) by the CaCo-2 model of human enterocyte epithelium.J Biol Chem 1988; 263(18):8786-8789.131. Quinet EM, Agellon LB, Kroon PA, Marcel YL, Lee YC, Whitlock ME. Atherogenic diet increases CETP mRNA levels in rab-bit liver. J Clin Invest 1990; 85(2):357-363.132. Quig DW, Zilversmit DB. Plasma lipid transfer activity in rabbits. Atherosclerosis 1988; 70(3):263-271.133. Groener J, van Ramshorst E, Katan M, Mensink R, van Tol A. Diet modulates plasma neutral lipid transfer protein activityin normolipidemic human subjects. Klin Wochenschr 1990; 68 Suppl 22:106-7.:106-107.134. Son YS, Zilversmit DB. Increased lipid transfer activities in hyperlipidemic rabbits. Arteriosclerosis 1986; 6(3):345-351.135. Freeman DJ, Caslake MJ, Griffin BA, Hinnie J, Tan CE, Watson TD. The effect of smoking on post-heparin LPL and HL,CETP and LCAT activities in human plasma. Eur J Clin Invest 1998; 28(7):584-591.136. Nagashima M, McLean JW, Lawn RM. Cloning and mRNA tissue distribution of rabbit CETP. JLR 1988; 29(12):1643-1649.137. Jiang XC, Masucci-Magoulas L, Mar J, Lin M, Walsh A, Breslow JL et al. Down-regulation of mRNA for the LDL receptor intransgenic mice containing the gene for human CETP. J Biol Chem 1993; 268(36):27406-27412.138. Kawano K, et al. Role of HL and SRBI in clearing PL/cholesterol-rich lipoproteins in PLTP-/- mice. BBA 2002;1583(2):133139. Lagrost L et al., Opposite effects of CETP and PLTP on size distribution of plasma HDL. JBC 1996; 271(32):19058-65.140. Urizar NL, et al. FXR mediates bile acid activation of PLTP gene expression. JBC 2000; 275(50):39313-39317.141. Socha P, Koletzko B, Pawlowska J, Socha J. EFA status in children with cholestasis, in relation to serum bilirubin concen-tration. J Pediatr 1997; 131(5):700-706.142. Socha P, Koletzko B, Swiatkowska E, Pawlowska J, Stolarczyk A, Socha J. EFA metabolism in infants with cholestasis. ActaPaediatr 1998; 87(3):278-283.143. Minich DM, Havinga R, Stellaard F, Vonk RJ, Kuipers F, Verkade HJ. Intestinal absorption and postabsorptive metabolismof LA in rats with short-term bile duct ligation. AJP 2000 ;279 (6 ):G1242 -8 2000; 279(6):G1242-48.144. Vriese De SR, Savelli JL, Poisson JP, Narce M, Kerremans I, Lefebvre R. Fat absorption and metbolism in bile duct ligatedrats. Nutrition and Metabolism 1 A.D.; 45:209-216.145. Krahenbuhl S, Talos C, Reichen J. Mechanisms of impaired hepatic fatty acid metabolism in rats with long-term bile ductligation. Hepatology 1994; 19(5):1272-81.146. Moreno LA, Gottrand F, Hoden S, Turck D, Loeuille GA, Farriaux JP. Improvement of nutritional status in cholestatic chil-dren with supplemental nocturnal enteral nutrition. J Pediatr Gastroenterol Nutr 1991; 12(2):213-216.147. Bavdekar A, Bhave S, Pandit A. Nutrition management in chronic liver disease. Indian J Pediatr 2002; 69(5):427-431.148. Sokol RJ, et al Multicenter trial of d-alpha-tocopheryl polyethylene glycol 1000 succinate for treatment of vitamin E defi-ciency in children with cholestasis. Gastroenterology 1993; 104(6):1727-1735.149. Argao EA, Heubi JE, Hollis BW, Tsang RC. d-Alpha-tocopheryl polyethylene glycol-1000 succinate enhances absorptionof vitamin D in cholestatic liver disease of infancy and childhood. Pediatr Res 1992; 31(2):146-150.

36

Chapter 1


A. WernerD.M. Minich

H. Havinga V.W. Bloks

H. van GoorF. Kuipers

H.J. Verkade

2Am J Physiol Gastrointest Liver Physiol. 2002; 283(4): G900-G908

Fat malabsorptionin essential fatty acid-deficient mice

is not due to impaired bile formation


ABSTRACT

Background: Essential fatty acid (EFA) deficiency induces fat malabsorption, but the

pathophysiological mechanism hereof is unknown. Bile salts and EFA-rich biliary

phospholipids affect dietary fat solubilization and chylomicron formation, respec-

tively. We investigated whether altered biliary bile salt and/or phospholipid secretion

mediate EFA deficiency-induced fat malabsorption in mice.

Methods: FVB mice received EFA-sufficient (EFAS) or EFA-deficient (EFAD) chow for

eight weeks. Subsequently, fat absorption, bile flow and bile composition were deter-

mined. Identical dietary experiments were performed in Mdr2 -/- mice, secreting

phospholipid-free bile.

Results: After eight weeks, EFAD chow-fed wildtype and Mdr2 -/- mice were

markedly EFA-deficient (plasma triene (C20:3n-9) / tetraene (C20:4n-6) ratio >0.2).

Fat absorption decreased (70.1±4.2% vs. 99.1±0.3%, p<0.001) but bile flow and

biliary bile salt secretion increased in EFA-deficient mice compared to EFA-sufficient

controls (4.87±0.36 vs. 2.87±0.29 µl/min/100g bodyweight, p<0.001; 252±30 vs.

145±20 nmol/min/100g bodyweight, p<0.001). Bile salt composition was similar in

EFAS and EFAD chow-fed mice. Similar to EFA-deficient wildtype mice, EFA-deficient

Mdr2 -/- mice developed fat malabsorption associated with twofold increased bile flow

and bile salt secretion.

Conclusion: Fat malabsorption in EFA-deficient mice is not due to impaired biliary

bile salt or phospholipid secretion. We hypothesize that EFA deficiency affects

intracellular processing of dietary fat by enterocytes.

38

Chapter 2


39

Fat malabsorption in essential fatty acid-deficient mice is not due toimpaired bile formation

INTRODUCTION

Essential fatty acid (EFA) deficiency has been associated with several pathological

consequences in humans and in experimental animals, including fat malabsorp-

tion(1-5). The pathophysiological mechanisms underlying EFA deficiency-induced fat

malabsorption have not been elucidated. To address this issue, the consecutive intra-

luminal and intracellular steps in the absorption process have been investigated in

control and EFA-deficient rats, i.e., lipolysis of dietary triglycerides by lipases, solu-

bilization of lipolytic products by bile components, uptake by the enterocyte and

intracellular re-esterification to triglycerides and, finally, chylomicron formation and

secretion into lymph(3;4). Available data from rat models of EFA deficiency have indi-

cated that neither fat digestion nor fatty acid uptake by the enterocyte is responsible

for the fat malabsorption encountered in EFA deficiency(3;4). In rats, EFA deficiency

resulted in significantly decreased bile flow and biliary secretion of bile salts, phos-

pholipids (PL) and cholesterol(3;6). Acyl chain analysis of biliary PL from EFA-deficient

rats revealed that biliary PL contents of linoleic acid (C18:2n-6) and arachidonic acid

(C20:4n-6) were decreased to 10% and 26% of control values, respectively, and were

replaced by oleic acid (C18:1n-9)(4). In addition to alterations in bile, intracellular

events such as fatty acid re-esterification and chylomicron production were severely

impaired(3;4). It thus seems that EFA deficiency in rats affects bile formation and/or

intracellular processing of fat by the enterocyte, resulting in fat malabsorption. In a

previous study we demonstrated the role of biliary phospholipids in intestinal

chylomicron formation, using Mdr2 -/- mice (gene symbol: ABC-B4) which are unable

to secrete phospholipids into bile(7;8). We reasoned that this animal model could be

important to further delineate the pathophysiological mechanism of EFA deficiency-

induced fat malabsorption.

If altered biliary phospholipid secretion were involved in EFA deficiency-associated

fat malabsorption, one would expect fat absorption in Mdr2 -/- mice to be relatively

unaffected by changes in EFA status. If, however, the pathophysiological mechanism

would not involve alterations in biliary phospholipid secretion, EFA deficiency would

be expected to equally affect the process in Mdr2 -/- and in wildtype mice. In the

present study we investigated the role of bile formation in general and of biliary phos-

pholipid secretion in particular, in the pathophysiology of EFA deficiency-associated

fat malabsorption. We first developed and characterized a murine model for EFA

deficiency with respect to fat absorption and bile formation. Then, we compared fat

absorption and bile formation in EFA-deficient and EFA-sufficient Mdr2 -/- mice.


MATERIALS AND METHODS

AnimalsMice homozygous for disruption of the multidrug resistance gene-2 (Mdr2 -/-) and

wildtype (Mdr2 +/+) mice with a free virus breed (FVB) background were obtained

from the breeding colony at the Central Animal Facility, Academic Medical Center,

Amsterdam, the Netherlands (9). All mice were 2-4 months old and weighed 25-30 g.

Mice were housed in a light controlled (lights on 6 AM - 6 PM) and temperature

controlled (21°C) facility and allowed tap water and chow (Hope Farms B.V. 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.

Experimental proceduresFat absorption and bile secretion in EFA-deficient mice

FVB mice (n=14) were fed standard low-fat chow (6 weight% fat, 14 energy% fat) for

standardization for one month, after which they were anesthetized with halothane

and a baseline blood sample was obtained by tail bleeding for determination of EFA

status. Blood was collected in micro-hematocrit tubes containing heparin, and plas-

ma was separated by centrifugation at 9000 rpm for 10 min (Eppendorf centrifuge,

Eppendorf, Germany) and stored at -20°C until analysis. Subsequently, mice were

randomly assigned to either an EFA-sufficient (EFAS) or an EFA-deficient (EFAD)

diet. The EFAS and EFAD diets were isocaloric and contained 16 weight% fat. The

EFAS chow contained 20 energy%, 34 energy% and 46 energy% from protein, fat

and carbohydrate, respectively, and had the following fatty acid profile: 32 mol%

palmitic acid (C16:0), 6 mol% stearic acid (C18:0), 32 mol% oleic acid (C18:1n-9),

and 30 mol% linoleic acid (C18:2n-6) (custom synthesis, Hope Farms BV, Woerden,

the Netherlands). The EFAD diet had identical energy percentages derived from pro-

tein, fat and carbohydrate, and had the following fatty acid composition: 41 mol%

C16:0, 48 mol% C18:0, 8 mol% C18:1n-9, and 3 mol% C18:2n-6) (custom synthesis,

Hope Farms BV, Woerden, the Netherlands). At 2-weekly intervals, blood samples

were taken in the manner described above. After 4 and 8 weeks of experimental diet,

a 72h fecal fat balance was performed involving quantitative feces collection and

determination of chow intake. At 8 weeks, mice were anaesthetized by intraperitoneal

injection of Hypnorm (fentanyl/fluanisone) and diazepam, and gallbladders were

cannulated for collection of bile for 1h as described previously(10). At the end of bile

collection, a large blood sample (0.6-1.0 ml) was obtained by heart puncture.

40

Chapter 2


41

Fat malabsorption inessential fatty acid-deficient mice is not due toimpaired bile formation

Hepatic expression of Cyp7A and Cyp27

A separate group of male FVB mice was fed EFA-containing or EFA-deficient chow

for 8 weeks (n=6 per dietary group). After 8 weeks, animals were anaesthetized with

halothane and blood was collected by cardiac puncture for lipid analysis. Livers were

removed and liver samples were immediately frozen in liquid nitrogen, and stored at

-80°C for determination of mRNA levels of two key enzymes in bile salt biosynthesis,

cholesterol 7-alpha-hydroxylase (Cyp7A) and sterol 27-hydroxylase (Cyp27).

Plasma accumulation of [3H]-triolein and [14C]-oleic acid after Triton WR-1339

In a separate experiment, male FVB mice fed EFAS or EFAD chow for 8 weeks (n=5

per group) were injected i.v. with 12.5 mg Triton WR-1339 (12.5 mg/100 µL PBS) to

block lipolysis of lipoproteins in the circulation. Then an intragastric fat bolus was

administered containing 200 µL olive oil, in which 10 µCi [3H]-triolein (glycerol tri-

[9,10(n)-[3H]]-oleate) (Amersham, Buckinghamshire, UK) and 2 µCi [14C]-oleic acid

(NEN Laboratories, Boston, MA) were dispersed. Before (t=0) and at 1, 2, 3 and 4

hours after label administration, blood samples (75 µL) were taken by tail bleeding.

[3H] and [14C] in plasma (25 µL) were measured by scintillation counting

Plasma appearance of retinyl-palmitate after retinol administration

In addition to the absorption of fatty acids during EFA deficiency, the plasma

appearance of the less polar lipid molecule retinol (vitamin A) was investigated.

Retinol (5000 IU) in an olive oil bolus (100 µL per mouse) was administered intra-

gastrically to EFA-deficient mice and control mice (n=7 per group) and blood

samples were taken at 0, 2 and 4 hours after bolus administration.

Fat absorption and bile secretion in EFA-deficient Mdr2 -/- mice

Mdr2 -/- mice (n=14) were fed low-fat chow (6 weight% fat, 14 energy% fat) for

standardization for one month. Baseline blood samples were taken for determination

of plasma EFA status, after which mice were randomly assigned to either the EFA-

containing or the EFA-deficient diet. The diets were identical to those described

above. After eight weeks of feeding the respective diets, blood samples were

obtained by tail bleeding under halothane anesthesia. As in the wildtype mice, fat

absorption was measured by means of a 72h fecal fat balance. Retinol in a bolus of

olive oil was administered intragastrically to EFA-deficient Mdr2 -/- mice (n=7) and

their EFA-sufficient controls (n=7). Blood samples were taken at 0, 2 and 4 hours

after bolus administration. Gallbladders were cannulated and bile was collected for

1h as described above(10). At the end of bile collection, mice were sacrificed after

obtaining a large blood sample (0.6-1.0 ml) by heart puncture.


Analytical techniquesFatty acid status was analyzed by extracting, hydrolyzing and methylating total

plasma lipids, liver homogenates and biliary lipids according to the method

described by Lepage and Roy(11). To account for losses during lipid extraction,

heptadecanoic acid (C17:0) was added to all samples as an internal standard prior

to extraction and methylation procedures, and BHT was added as an antioxidant.

Fatty acid methyl esters were separated and quantified by gas liquid chromato-

graphy on a Hewlett Packard gas chromatograph model 6890, equipped with a

50mx0.2mm Ultra 1 capillary column (Hewlett Packard, Palo Alto, CA) and a FID

detector. The injector and detector were set at 260°C and 250°C, respectively. The

oven temperature was programmed from an initial temperature of 160°C to a final

temperature of 290°C in 3 temperature steps (160°C held 2 min; 160-240°C, ramp

2°C/min, held 1 min; 240-290°C, ramp 10°C/min, held 10 min). Helium was used as

a carrier gas with a constant flow rate of 0.5 ml per minute. Individual fatty acid methyl

esters were quantified by relating the areas of their chromatogram peaks to that of

the internal standard heptadecanoic acid (C17:0).

Plasma lipid levels (cholesterol, HDL-cholesterol, triglycerides, free fatty acids and

phospholipids) were measured using commercially available kits (Roche

Diagnostics, Mannheim resp. WAKO Chemicals Neuss, Germany) according to the

instructions provided. Cholesterol, cholesterol ester and triglycerides in liver tissue

were determined after Bligh and Dyer lipid extraction (12) and biliary bile salt composi-

tion was measured as described previously. Total protein concentrations of liver

homogenates were determined according to the method described by Lowry et al.(13).

Total RNA was isolated from liver tissue using Trizol Reagent (GIBCO BRL, Grand

Island, NY) according to the manufacturer's instructions. Single stranded cDNA was

synthesized from 4.5 g RNA and subsequently subjected to quantitative real-time

detection RT-PCR(14;15). The following primers and probes were used: for Cyp7A

(accession number: L23754): 5'-CAG GGA GAT GCT CTG TGT TCA-3' (forward

primer), 5'-AGG CAT ACA TCC CTT CCG TGA-3' (reverse primer), 5'-TGC AAA ACC

TCC AAT CTG TCA TGA GAC CTC C-3' (probe). For Cyp27 (M73231, M62401): 5'-

GCC TTG CAC AAG GAA GTG ACT-3' (forward primer), 5'-CGC AGG GTC TCC TTA

ATC ACA-3' (reverse primer), 5'-CCC TTC GGG AAG GTG CCC CAG-3' (probe), and

for ß-actin (M12481): 5'-AGC CAT GTA CGT AGC CAT CCA-3' (forward primer) 5'-TCT

CCG GAG TCC ATC ACA ATG-3' (reverse primer) 5'-TGT CCC TGT ATG CCT CTG

GTC GTA CCA C-3' (probe). For each real-time PCR, 4 µL cDNA was used in a final

volume of 20 µL, containing 500 nmol/L of forward and reverse primers and 200

42

Chapter 2


43


nmol/L of probe, 250 nmol/L MgCl2, 10 nmol/L deoxy-ribonucleoside triphosphate

mix, 5 µL Real-time PCR buffer (10x), and 1.25 U Hot GoldStar (Eurogentec). Real-

time detection PCR was performed on the ABI PRISM 7700 (PE Applied Biosystems)

initialized by 10 minutes at 95oC to denature the complementary DNA followed by 40

PCR cycles each at 95oC for 15 seconds and 60oC for 1 minute.

Feces and chow pellets were freeze-dried and then mechanically homogenized.

From aliquots of each, lipids were extracted, hydrolyzed and methylated(11). Resulting

fatty acid methyl esters were analyzed by gas chromatography for their fatty acid

content as described previously (16). Fatty acids were quantified using heptadecanoic

acid (C17:0) as internal standard.

Retinyl palmitate concentrations in plasma were determined after two extractions with

hexane(17). Retinyl acetate was added to plasma samples as an internal standard

before lipid extraction. Samples were resuspended in ethanol and analyzed by

reverse-phase HPLC using a 150 x 4,6 mm Symmetry RP18 column (Waters Corp.,

Milford, MA, USA)(18). Peak area of retinyl palmitate was normalized to that of retinyl

acetate. At each time point, concentrations were expressed as µmol retinyl palmitate

per L plasma.

CalculationsFatty acid status in plasma and in bile

Relative concentrations (mol%) of plasma, liver and biliary phospholipid fatty acids

were calculated using the summed areas of major fatty acid peaks (palmitic, stearic,

oleic, linoleic acids) and then expressing the area of each individual fatty acid as a

percentage of this amount. EFA deficiency was determined by calculating the triene

/tetraene ratio (C20:3n-9/C20:4n-6) in plasma of mice. A ratio of >0.2 was

considered deficient (19).

Fatty acid absorption using 72h balance techniques

Absorption of major dietary fatty acids (palmitic, stearic, oleic, linoleic acids) was

determined by subtracting the amount of each individual fatty acid excreted in feces

in 72h, from the amount of this dietary fatty acid ingested in 72h (net fat absorption).

This quantity was subsequently expressed as a percentage of the amount of fatty

acid ingested in 72h (coefficient of fat absorption; 100-fold the ratio between (amount

ingested - amount recovered in feces) and the amount ingested). Net amount of fat

absorption was calculated by subtracting the fecal loss of the four major fatty acids

from the amount ingested.


StatisticsAll results are presented as means ± S.D. for the number of animals indicated. Data

were statistically analyzed using Student's two-tailed t test. The level of significancefor all statistical analysis was set at p<0.05. Analyses were performed using SPSS

for Windows software (SPSS, Chicago, IL).

RESULTS

Body weight and chow ingestion in EFA-deficient and EFA-sufficient wildtype mice

Body weight was monitored at the start of experimental feeding, and then every two

weeks for eight weeks. No differences in basal or final weight were found for wildtypemice fed EFAS or EFAD chow (basal: 24.4±2.0 and 24.6±1.3 gram; final: 23.3±1.0

and 21.9±2.0 gram, respectively). Chow intake, measured after four and eight weeks

of experimental diet feeding, was similar in both dietary groups (data not shown).

Essential fatty acid statusTriene/tetraene ratio in plasma and in liverThe classical biochemical parameter describing EFA status, the triene/tetraene ratio

(C20:3n-9/C20:4n-6), was measured in plasma at baseline and every two weeks for

eight weeks, and in liver after eight weeks of feeding EFA-deficient chow. Baselineplasma triene/tetraene ratio was 0.02±0.00, which is well below the cutoff value for

EFA deficiency (0.20). Already after two weeks on EFA-deficient diet, plasma

triene/tetraene ratios reached the cutoff-value for EFA deficiency, i.e., 0.19±0.08 forEFAD chow-fed mice vs. 0.01±0.00 for controls. After eight weeks on EFAD chow,

mice had a pronounced EFA deficiency with triene/tetraene ratios of 0.66±0.05 for

plasma and 0.56±0.09 for liver, in contrast to control mice (0.01±0.00 for plasma,p<0.001 and 0.02±0.00 for liver, p<0.001) (data not shown)

Characterization of EFA deficiency in miceSince EFA deficiency has not been characterized before in a murine model, plasma

and liver lipids were measured. No differences were found in cholesterol, HDL-

cholesterol, free fatty acid or phospholipid concentrations in plasma between EFASand EFAD diet-fed mice (Table 1). Plasma triglyceride concentrations were

decreased in EFA-deficient mice compared with mice fed EFA-containing chow

(p<0.001). Liver fat analysis revealed a significant fat accumulation (triglyceride,unesterified cholesterol, cholesterol ester) in EFA-deficient mice compared with their

EFA-sufficient counterparts (Table 2).

44

Chapter 2


45


Fat absorptionFecal fatty acid balance

The fecal fat balance revealed a decreased absorption of total dietary fat in EFAD

mice compared with their EFAS controls (p<0.01) (Figure 1). The absorption co-

efficient for EFAD mice was 70.1±1.6%, compared with absorption coefficients

above 95% for mice fed EFA-containing chow. Individual fatty acid balances for

palmitic, stearic, oleic and linoleic acids were calculated (Figure 1). In EFAD mice,

absorption of saturated fatty acids (palmitic and stearic acids) was more affected

than absorption of unsaturated fatty acids (oleic and linoleic acids).

Bile secretion and composition

Table 3 shows that bile flow and biliary secretion of bile salt, cholesterol and phos-

pholipid during a 1h period immediately after interruption of the enterohepatic

circulation were higher in EFA-deficient mice compared with controls (for each

parameter, p<0.001). Theoretically, alterations in bile salt hydrophobicity could

contribute to fat malabsorption in EFA deficiency. However, bile salt composition

appeared to be similar between both dietary groups (Table 4).

lipid absorption

0

20

40

60

80

100

total FA C16:0 C18:0 C18:1 C18:2

abso

rptio

n pe

rcen

tage

(%) EFAS

EFADlipid absorption

0

20

40

60

80

100

total FA C16:0 C18:0 C18:1 C18:2

abso

rptio

n pe

rcen

tage

(%) EFAS

EFAD

EFAS

EFAD

Figure 1. Fat absorption coefficients of total dietary fat,and separately of major dietary fatty acids (C16:0, C18:0,C18:1n-9, C18:2n-6) in mice fed EFAS and EFAD chow for8 weeks. Feces were collected after a 72h period in whichchow intake was monitored by weighing chow containers.Absorption was calculated by subtracting fecal excretionof these fatty acids after 72h from their dietary intake in72h and then multiplying the result by 100, as detailed inthe Materials and Methods section. Data represent means± SD of 7 mice per group. For all lipid classes p<0.01

35.8 ± 9.4 *8.0 ± 2.0Cholesterol esters

75.6 ± 5.3 *34.9 ± 2.7Total cholesterol

219.9 ± 27.8 *151.9 ± 21.8Triglyceride

EFA DEFAS



219.9 ± 27.8 *151.9 ± 21.8Triglyceride

EFA DEFAS

3.11 ± 0.172.84 ± 0.44Phospholipids

0.60 ± 0.040.58 ± 0.10Free fatty acids

0.27 ± 0.05 *0.48 ± 0.07Triglyceride

1.41 ± 0.081.44 ± 0.16HDL-cholesterol

2.21 ± 0.122.04 ± 0.23Cholesterol

EFADEFAS

3.11 ± 0.172.84 ± 0.44Phospholipids

0.60 ± 0.040.58 ± 0.10Free fatty acids

0.27 ± 0.05 *0.48 ± 0.07Triglyceride

1.41 ± 0.081.44 ± 0.16HDL-cholesterol

2.21 ± 0.122.04 ± 0.23Cholesterol

EFADEFAS Table 1: Plasma lipid concentrations in mice fed EFA-con-taining (EFAS) or EFA-deficient (EFAD) chow for 8 weeks.Concentrations are in mM, * p<0.001; n=7 mice / group.

Table 2: Hepatic lipid concentrations in mice fed EFA-containing or EFA-deficient chow for 8 weeks.Concentrations are given in nmol/mg protein; * p<0.001;n=7 mice per group.

1


EFA deficiency-associated changes in acyl chain composition of biliary PL were

analyzed by gas chromatography. Relative concentrations (mol%) of C16:1n-7,

C18:1n-9 and C18:1n-7 were higher (p<0.001) and concentrations of C16:0,

C18:2n-6, C18:0 and C20:4n-6 were lower (p<0.05) in bile of EFA-deficient mice

compared to mice fed EFA-containing chow (data not shown).

Cyp7A and Cyp27 mRNA levels

To determine whether the increased bile salt secretion in EFA-deficient mice might be

due to increased hepatic bile salt synthesis, hepatic mRNA levels of Cyp7A and

Cyp27 were measured by quantitative real-time RT-PCR, using ß-actin mRNA as a

housekeeping signal. No significant differences in hepatic mRNA levels of Cyp7A

and Cyp27 were observed between mice fed EFAS and EFAD chow (1.00±0.64 vs.

0.63±0.29 for Cyp7A and 1.00±0.22 vs. 1.14±0.19 for Cyp27) (Figure 2). Values

represent the ratio of specific hepatic mRNA levels of Cyp7A and Cyp27 to the hepat-

ic mRNA level of ß-actin, normalized to the EFA-sufficient control group.

46

Chapter 2

0.0

0.3

0.5

0.8

1.0

1.3

1.5

cyp7a cyp27

rela

tive

mR

NA

leve

ls

EFAS

EFAD

0.0

0.3

0.5

0.8

1.0

1.3

1.5

cyp7a cyp27

rela

tive

mR

NA

leve

ls

EFAS

EFAD

EFAS

EFAD

Figure 2. Hepatic mRNA levels of Cyp7A and Cyp27 inmice fed EFAS or EFAD chow for 8 weeks determined byquantitative real-time RT-PCR.Values represent the ratio ofspecific hepatic mRNA levels of Cyp7A and Cyp27 to thatof ß-actin, normalized to the EFA-sufficient control group.Data represent means±SD of 6 mice per group. No sig-nificant differences were found between groups.

1.0 ± 0.41.1 ± 0.7Chenodeoxycholate

2.1 ± 0.91.6 ± 0.6Deoxycholate

2.5 ± 0.52.7 ± 0.5Alpha-muricholate

7.4 ± 1.56.5 ± 2.7Omega-muricholate

28.1 ± 4.6 *39.0 ± 1.3Beta-muricholate

58.9 ± 4.5 *49.0 ± 3.2Cholate

EFADEFAS





28.1 ± 4.6 *39.0 ± 1.3Beta-muricholate

58.9 ± 4.5 *49.0 ± 3.2Cholate

EFADEFAS

55.13 ± 9.23 *30.11 ± 7.72Phospholipids ²

4.96 ± 0.79 *2.23 ± 0.47Cholesterol ²

353 ± 74 *145 ± 48Bile salts ²

4.87 ± 0.87 *2.87 ± 0.71Bile flow ¹

EFADEFAS

55.13 ± 9.23 *30.11 ± 7.72Phospholipids ²

4.96 ± 0.79 *2.23 ± 0.47Cholesterol ²

353 ± 74 *145 ± 48Bile salts ²

4.87 ± 0.87 *2.87 ± 0.71Bile flow ¹

EFADEFAS Table 3: Bile flow and biliary secretion rates in mice fedEFA-sufficient or EFA-deficient chow for 8 weeks.Biliaryoutput rates are given in ¹ µl/min/100g body weight or² nmol/min/100 g body weight.* p<0.001; n=5-7 mice pergroup

Table 4: Biliary bile salt composition in mice fed EFA-containing or EFA-deficient chow for 8 weeks.>90% of allbile salts represented. Minor metabolites (<1% of totalarea) have been excluded. Values are expressed as a per-centage of the total amount. Data represent means ± SD, * p<0.01; n=5-7 mice per group.

2


47


Plasma accumulation of [3H]-triolein and [14C]-oleic acid after Triton WR-1339

administration

Figure 3a and 3b show the time course of plasma [3H] and [14C] radioactivity after

intragastric administration of [3H]-triolein and [14C]-oleic acid. EFA-deficient mice had

a slightly increased plasma concentration of both radioactive labels during the

studied time frame, reaching a significant difference at 3 hours after bolus adminis-

tration. If EFA deficiency would differentially affect lipolysis and fatty acid uptake, an

altered [3H]/[14C] ratio in plasma would be expected. However, when expressed as

the ratio between plasma [3H] and [14C] (figure 3c), no significant difference was

found between EFA-deficient and EFA-sufficient mice. Both for EFA-containing and

EFA-deficient chow-fed mice, thin layer chromatography revealed that [3H] as well as

[14C] radioactivity was predominantly (>95%) present in the triglyceride fraction.

Plasma appearance of retinyl-palmitate after retinol administration

The absorption of dietary oleic acid, a polar lipid (class III, soluble amphiphile)(20) was

only mildly impaired in EFA-deficient mice (Figure 1). In a separate experiment we

3b0

20000

40000

60000

0 1 2 3 43b0

20000

40000

60000

0 1 2 3 40

20000

40000

60000

0 1 2 3 4

3a0

20000

40000

60000

0 1 2 3 43a0

20000

40000

60000

0 1 2 3 4

0

2

4

6

8

10

time after administration (h)[3 H]-

trio

lein

/[14C

]-ol

eic

acid

rat

io

0 1 2 3 4

EFAS

3c

EFAD

0

2

4

6

8

10

time after administration (h)[3 H]-

trio

lein

/[14C

]-ol

eic

acid

rat

io

0 1 2 3 4

EFASEFAS

3c

EFADEFAD

Figure 3. Plasma appearance of [3H]-triolein ( 3a ) and[14C]-oleic acid ( 3b ) in mice after an intragastric load ofolive oil containing these fats and after intravenous injec-tion of Triton WR-1339 at time point zero. Mice were fedEFAS (solid line) or EFAD (dotted line) chow for 8 weeks.Blood samples were taken hourly for 4 hours and fatswere extracted from plasma following extraction. Thinlayer chromatography was performed to isolate thetriglyceride fraction for determination of radioactivity byscintillation counting. Scanning of the thin layer platesindicated that essentially all radioactivity (95%) was in thetriglyceride fraction. Data represent means ± SD of 5mice per group (dps per milliliter plasma). Statistical sig-nificance was reached for [3H]-triolein and [14C]-oleic acidat 3h after administration (p<0.05). ( 3c ) The ratio of[3H]/[14C] in plasma of in mice fed EFAS (solid line) orEFAD (dotted line) chow for 8 weeks after an intragastricload of olive oil containing [3H]-triolein and [14C]-oleic acidand after intravenous injection of Triton WR-1339 at timepoint zero. The [3H]/[14C] in the infusate was 4.43±0.05.Data represent means ± SD of 5 mice per group. No sig-nificant differences were found between groups at anytime point.


investigated plasma appearance of retinol (vitamin A), a less polar lipid (class I,

insoluble non-swelling amphiphile), after its intragastric administration. Plasmaconcentrations of retinyl palmitate were significantly lower in EFA-deficient mice

compared to controls at 2 and 4 hours after retinol administration (Figure 4).

Body weight and chow ingestion in EFA-deficient and EFA-sufficient Mdr2 -/- miceBasal body weights of Mdr2 -/- mice entering the EFAS and EFAD group were similar

(26.1±1.6 vs. 25.4±1.5, respectively). However, body weights of Mdr2 -/- mice fed

EFAD chow gradually decreased compared to Mdr2 -/- mice fed EFAS chow, and byeight weeks this resulted in a significantly lower body weight for EFAD- compared to

EFAS chow-fed Mdr2 -/- mice (20.2±1.0 g vs. 25.3±1.1g, respectively; p<0.01).

No significant difference in chow intake was observed between the two dietarygroups after four or eight weeks on either diet (data not shown).

Essential fatty acid statusTriene/tetraene ratio in plasma and in liver

The baseline triene/tetraene ratios (C20:3n-9/C20:4n-6) in plasma and liver were sig-

nificantly higher in Mdr2 -/- compared to Mdr2 +/+ mice (0.035±0.003 vs. 0.018±0.001,p<0.001), but were still well below the cutoff value for EFA deficiency (0.2). Mdr2 -/-

mice fed EFAD chow for 8 weeks had developed EFA deficiency according to

triene/tetraene ratios in plasma and in liver (0.46±0.03 for plasma and 0.38±0.04 forliver), in contrast to the EFA-sufficient Mdr2 -/- controls (0.01±0.01 for plasma,

p<0.001 and 0.02±0.00 for liver, p<0.01).

Characterization of EFA deficiency in Mdr2 -/- mice

Plasma concentrations of cholesterol and phospholipid were increased, whereas the

plasma triglyceride level was decreased in EFA-deficient Mdr2 -/- mice compared totheir EFA-sufficient controls (p<0.05) (Table 5). Similar to the situation in wildtype

mice, a pronounced hepatic fat accumulation characterized by increased levels of

triglyceride, unesterified and esterified cholesterol, was observed in EFA-deficientMdr2 -/- mice (Table 6).

48

Chapter 2

4

0

5

10

15

20

25

0 2 4hours after bolus administrationpl

asm

a re

tinyl

pal

mita

te (µ

mol

/L)

EFASEFAD

4

0

5

10

15

20

25

0 2 4hours after bolus administrationpl

asm

a re

tinyl

pal

mita

te (µ

mol

/L)

EFASEFADEFAD

Figure 4:Time course of retinyl palmitate concentration inplasma of mice fed EFAS (solid line) or EFAD chow (dot-ted line) for 8 weeks after intragastric administration of5000 IU retinol at time zero. Data represent means ± SDof 7 mice per group. p<0.05 at 2 and 4 hours after bolusadministration.


49


The fecal fat balance revealed a decreased dietary fat absorption in EFA-deficient

Mdr2 -/- mice compared to EFA-sufficient controls (p<0.01, Figure 5). Absorption

coefficients for saturated fatty acids (palmitic and stearic acids) were lower than

absorption coefficients of unsaturated fatty acids (oleic and linoleic acids) in EFA-

deficient Mdr2 -/- mice. Plasma concentrations of retinyl palmitate after intragastric

administration of retinol were significantly lower in EFA-deficient deficient Mdr2 -/-

mice compared to EFA-sufficient controls (p<0.01, Figure 6)

Bile secretion and composition

As in EFAD Mdr2 +/+ mice (Table 3), bile flow was increased in EFAD Mdr2 -/- mice

compared with their EFAS counterparts (p<0.05). Bile flow and bile salt secretion

were higher in Mdr2 -/- mice than in Mdr2 +/+ mice (p<0.01) (Table 7). Bile salt

composition was virtually identical between EFAS and EFAD Mdr2 -/- mice (Table 8).

lipid absorption in Mdr2 -/- mice

0

20

40

60

80

100

total FA C16:0 C18:0 C18:1 C18:2

abso

rptio

n pe

rcen

tage

(%)

EFASEFAD

lipid absorption in Mdr2 -/- mice

0

20

40

60

80

100

total FA C16:0 C18:0 C18:1 C18:2

abso

rptio

n pe

rcen

tage

(%)

EFASEFAD

6

0

5

10

15

0 2 4time after bolus administrationpl

asm

a re

tinyl

pal

mita

te (µ

mol

/L)

EFASEFAD

6

0

5

10

15

0 2 4time after bolus administrationpl

asm

a re

tinyl

pal

mita

te (µ

mol

/L)

EFASEFADEFASEFAD

Figure 5. Fat absorption coefficients of total dietary fat,and separately of major dietary fatty acids (C16:0, C18:0,C18:1n-9, C18:2n-6) in Mdr2(-/-) mice fed EFAS or EFADchow for 8 weeks. Feces were collected after a 72h peri-od in which chow intake was monitored by weighingchow containers. Absorption was calculated by subtract-ing fecal excretion of these fatty acids after 72h from theirdietary intake in 72h and multiplying the result by 100.Data represent means ± SD of 7 mice per group. For alllipid classes p<0.01.

Figure 6. Time course of retinyl palmitate concentration inplasma of Mdr2(-/-) mice fed EFAS (circles) or EFAD chow(squares) for 8 weeks after intragastric administration of5000 IU retinol at time zero. Data represent means ± SDof 7 mice per group. p<0.05 at 2 and 4 hours after bolusadministration.



225.8 ± 95.9 *101.0 ± 22.2Triglyceride

EFADEFAS



225.8 ± 95.9 *101.0 ± 22.2Triglyceride

EFADEFAS

2.82 ± 0.24 *2.18 ± 0.18Phospholipids

0.66 ± 0.110.61 ± 0.08Free fatty acids

0.34 ± 0.12 *0.46 ± 0.09Triglyceride

1.30 ± 0.19 *1.04 ± 0.12HDL-cholesterol

1.82 ± 0.21 *1.39 ± 0.14Cholesterol

EFADEFAS

2.82 ± 0.24 *2.18 ± 0.18Phospholipids

0.66 ± 0.110.61 ± 0.08Free fatty acids

0.34 ± 0.12 *0.46 ± 0.09Triglyceride

1.30 ± 0.19 *1.04 ± 0.12HDL-cholesterol

1.82 ± 0.21 *1.39 ± 0.14Cholesterol

EFADEFAS Table 5: Plasma lipid concentrations in Mdr2 -/- mice fedEFA-containing or EFA-deficient chow for 8 weeks.Concentrations are in mM, * p<0.05; n=7 mice / group.

Table 6: Hepatic lipid concentrations in Mdr2 -/- mice fedEFA-containing or EFA-deficient chow for 8 weeks.Concentrations are given in nmol/mg protein; * p<0.05;n=7 mice per group

5


DISCUSSION

We investigated the mechanism of fat malabsorption in EFA deficiency in mice, with

particular emphasis on the possible role of altered bile formation as proposed by

Levy et al. based on studies in rats(3;6). In order to specify the contribution of biliary

phospholipid secretion, studies were performed in EFA-deficient and EFA-sufficient

wildtype and Mdr2 -/- mice; the latter are unable to secrete phospholipids into their

bile. Yet, we first had to develop and characterize a murine model for EFA deficiency.

Our data indicate that dietary fat absorption is reduced in EFA-deficient mice

compared to EFA-sufficient controls. The mechanism underlying this EFA deficiency-

associated fat malabsorption does not likely involve alterations in bile formation

(including bile flow, bile salt secretion rate, bile salt composition or phospholipid

secretion rate) nor changes in fat digestion (lipolysis). Rather, our data strongly

indicate that EFA deficiency in mice affects intracellular events of fat absorption that

occur in the enterocyte.

Biochemical EFA deficiency, conventionally defined by a molar ratio of eicosatrienoic

acid (C20:3n-9) and arachidonic acid (C20:4n-6) above 0.20 in plasma(19), was

already reached after only 2 weeks of EFA-deficient diet feeding to mice. This

rapidity of onset makes the mouse an attractive and versatile model for studying EFA

deficiency.

Similar to EFA-deficient rats(21;22), EFA-deficient mice experienced changes in plasma

and liver fat contents. Specifically, EFA-deficient mice have decreased plasma

triglycerides and increased hepatic triglyceride and cholesterol levels. In EFA-

deficient rats, Wanon et al.(23) observed alterations in HDL-composition, with defective

translocation of HDL-cholesterol into bile and concomitantly increased hepatic VLDL-

50

Chapter 2





35.9 ± 4.128.0 ± 8.0Beta-muricholate

56.2 ± 3.463.7 ± 7.3Cholate

EFADEFAS





35.9 ± 4.128.0 ± 8.0Beta-muricholate

56.2 ± 3.463.7 ± 7.3Cholate

EFADEFAS

0.07 ± 0.160.10 ± 0.26Phospholipids ²

1.31 ± 0.350.97 ± 0.28Cholesterol ²

540 ± 182 *260 ± 53Bile salts ²

9.27 ± 2.17 *6.57 ± 1.57Bile flow ¹

EFADEFAS

0.07 ± 0.160.10 ± 0.26Phospholipids ²

1.31 ± 0.350.97 ± 0.28Cholesterol ²

540 ± 182 *260 ± 53Bile salts ²

9.27 ± 2.17 *6.57 ± 1.57Bile flow ¹

EFADEFAS Table 7: Bile flow and biliary secretion rates in Mdr2 -/-

mice fed EFA-containing or EFA-deficient chow for 8weeks. Biliary output rates are given in ¹ µl/min/100g bodyweight or ² nmol/min/100 g body weight.* p<0.05; n=5-7mice per group.

Table 8: Biliary bile salt composition in Mdr2 -/- mice fedEFA-containing or EFA-deficient chow for 8 weeks.>90%of all bile salts represented. Minor metabolites (<1% oftotal area) have been excluded. Values expressed as per-centage of total amount. Data represent means ± SD;n=5-7 mice per group.


51


cholesterol secretion. Lipoprotein abnormalities were also found by Levy et al.(24) in

EFA-deficient cystic fibrosis (CF) patients. Compared to their EFA-sufficient counter-

parts, EFA-deficient CF-patients had increased plasma triglyceride levels

(specifically in the VLDL, LDL, HDL2 and HDL3 fractions), and decreased plasma

HDL- and LDL-cholesterol. Also, lipoprotein size was altered in these EFA-deficient

patients, with larger VLDL, LDL, and HDL2 particles and smaller HDL3 particles.

It could be speculated that EFA, as constituents of triglycerides, phospholipids and

cholesterol esters, may be essential for regulation of lipoprotein metabolism.

Although the effects of EFA-deficiency are species specific, it appears that an

adequate EFA-status is required for efficient intestinal and hepatic processing of

lipoproteins.

In addition to the changes in plasma and liver lipids, EFA deficiency in mice was

associated with fat malabsorption. The coefficients of fat absorption in EFA-deficient

mice (60-70%) were somewhat lower than corresponding values in EFA-deficient rats

(80-90%)(2;4;5). In all of these studies, EFA deficiency was induced by feeding the ani-

mals high-fat EFA-deficient diets almost entirely composed of saturated fatty acids.

Apart from species specificity, the difference in coefficients of fat absorption could be

related to the amount and type of fat in the diet. In the present study, mice were fed

high fat chow (16 weight%) whereas Hjelte et al.(2) used chow diets containing only

7 weight% fat. Levy et al.(3) reported that lipolytic activity in EFA-deficient rats was

unchanged compared with control rats. Our present results in EFA-deficient mice are

compatible with this observation. The appearance of [3H]-triolein in plasma after its

intragastric administration was similar in EFA-deficient mice and control mice.

If anything, the [3H]-label was recovered from plasma of EFAD mice even at higher

concentrations compared to EFAS controls. The explanation for this phenomenon

may involve the choice of the lipid, oleic acid. The absorption of dietary oleic acid

was only mildly impaired during EFA deficiency (Figure 1). In addition, a tracer effect

could not be excluded. When we investigated the absorption of a less polar lipid,

retinol, after its intragastric administration, both EFA-deficient wildtype and EFA-

deficient Mdr2 -/- mice showed decreased plasma concentrations compared to their

EFA-sufficient controls, which underlines the occurrence of lipid malabsorption

during EFA deficiency.

Rather than differences in fat digestion (lipolysis), it could have been expected that

alterations in bile formation contributed to fat malabsorption in EFA-deficient mice, in

analogy to the situation in EFA-deficient rats(6). The pathophysiology of EFA deficien-

cy-associated fat malabsorption in mice could be due to decreased biliary bile salt


52

Chapter 2

secretion rates, analogous to previous data in rats. However, bile flow and biliary bile

salt and phospholipid secretion rates were increased during EFA deficiency. Robins

and Fasulo(25) reported that EFA-deficient hamsters have increased hepatic bile flow

and biliary bile salt and cholesterol secretion compared to controls. It is not known,

however, whether EFA-deficient hamsters have a decreased coefficient of fat absorp-

tion. Our observation in EFA-deficient mice, together with the available data on EFA-

deficient rats and hamsters, indicate that the effects of EFA deficiency on bile forma-

tion are species specific. Present data exclude that EFA-deficient fat malabsorption

is due to decreased rates of biliary bile salt secretion in mice, in contrast to the situ-

ation in rats(3;6). Theoretically, an increase in the contribution of hydrophilic bile salts

(to total bile salts) could contribute to impaired solubilization of dietary fats.

Yet, biliary bile salt composition was virtually unchanged in EFA-deficient mice

compared to controls. The increased biliary secretion rate of bile salts, immediately

after interruption of the enterohepatic circulation, strongly suggests an expansion of

the bile salt pool size in EFA deficiency.

Bile salts negatively affect their own biosynthesis by repressing the expression of

Cyp7A via the FXR-SHP1-LRH1-Cyp7A pathway, with Cyp7A encoding the enzyme

cholesterol 7A-hydroxylase that catalyzes the first step of the neutral pathway in bile

salt synthesis(26-30). In our experiments however, Cyp7A and Cyp27 mRNA levels were

similar in livers from EFAS and EFAD mice. We speculate that EFA deficiency in mice

impairs the capacity of bile salts to exert negative feedback inhibition on their own

hepatic biosynthesis, but the mechanism hereof remains unclear.

Not only biliary bile salts, but also biliary phospholipids play a role in dietary fat

absorption, for example in supplying surface components for the assembly of

chylomicron particles in enterocytes(31;32). Under physiological conditions, overall fat

absorption in Mdr2 -/- mice is only slightly decreased compared to control mice

(95 vs. 98%), based on 72h fecal fat balance measurements. On the other hand,

kinetics of chylomicron formation are clearly delayed in Mdr2 -/- mice(33). Therefore,

a quantitative alteration in biliary phospholipid secretion was not likely to contribute

to EFA deficiency-associated fat malabsorption. Yet, fat malabsorption during EFA

deficiency could still be due to qualitative changes in biliary phospholipid

composition. Replacement of polyunsaturated acyl chains (linoleoyl-, arachidonoyl-)

by monounsaturated or saturated species could theoretically be responsible for

impaired chylomicron assembly and secretion. In accordance with findings by

Bennett Clark et al. in EFA-deficient rats(4), the acyl chain composition of biliary PL in

EFA-deficient mice showed less essential (i.e., C18:2n-6 and C20:4n-6) and more

non-essential fatty acids (i.e., C18:1n-9, C18:1n-7, C16:1n-7). If acyl chain


53


composition of bile phospholipids were important for fat malabsorption during EFA

deficiency, one would expect that EFA deficiency would not, or to a much lesser

extent, affect fat absorption in Mdr2 -/- mice. Present data, however, clearly indicate

that fat absorption in EFA-deficient Mdr2 -/- mice was affected similarly as in EFA-

deficient wildtype mice. In EFA-deficient Mdr2 -/- mice, as in EFA-deficient wildtype

controls, bile flow and biliary bile salt secretion were increased. Based on the similar

fat absorption coefficients we found in EFA-deficient mice with and without biliary PL

secretion, we conclude that the effect of biliary PL acyl chain composition is not of

pathophysiological relevance for EFA deficiency-associated fat malabsorption.

Rather than to alterations in bile secretion, fat malabsorption during EFA deficiency

may be due to other steps involved in fat absorption. Intestinal mucosal phospho-

lipids normally contain large amounts of C18:2n-6 and C20:4n-6, and during EFA

deficiency the levels of these fatty acids are markedly decreased(34-36). The resultant

structural changes in membranes, and the increased cellular turnover rate in the

intestinal mucosa reported in EFA-deficient rats(5) could be responsible for decreased

dietary fat absorption. Based on our study and previous studies in EFA-deficient

rats(3;4), the intraluminal events involved in fat absorption (i.e., lipolysis of dietary

triglyceride by pancreatic lipase, solubilization of lipolytic products and uptake by the

enterocyte) seem to be relatively undisturbed in EFA deficiency. By inference, it is

therefore more likely that defects in one of the several intracellular events

(i.e., re-esterification, chylomicron assembly and/or secretion) are involved in EFA

deficiency-associated fat malabsorption.

AcknowledgementsThe authors thank R. Boverhof and C. Hulzebos for their technical assistance.

REFERENCES1. Tso P. Intestinal lipid absorption. In: Tso P, ed. Physiology of the gastrointestinal tract. New York: Raven Press 1994:1867-907.2. Carey MC, Hernell O. Digestion and absorption of fat. Seminars in gastrointestinal disease 1992;3:189-208.3. Verkade HJ, Tso P. Biophysics of intestinal luminal lipids. In: Tso, Mansbach, Kuksis: Intestinal lipid metabolism. New York:Kluwer 2000:1-18.4. Minich DM, Kuipers F, Vonk RJ, Verkade HJ. The role of bile in EFA absorption and metabolism. In: Riemersma RA, ArmstrongR, Kelly RW, Wilson R, eds. EFA and eicosanoids - 4th international congress. Champaign, Illinois: OACS Press 1998:43-7.5. Friedman HI, Nylund B. Intestinal fat digestion, absorption and transport. American journal of clinical nutrition 1980;1108-39.6. Kalivianakis M, Elsrodt J, Havinga R et al. Validation in an animal model of the carbon 13-labeled mixed triglyceride breathtest for the detection of intestinal fat malabsorption. Journal of pediatrics 1999;135:444-50.7. Sealy MJ, Muskiet FAJ, Martini IA et al. EFA deficientie bij pediatrische patienten. Tijdschrift Kindergeneesk 1997;65:144-50.8. Smit, E. N., Fokkema, M. R., Brouwer, D. A. J., Lanting, C. I., Woltil, H. A., Boersma, E. R., and Muskiet, F. A. J. Erythrocytelinoleic acid cut-off value for establishment of subclinical essential fatty acid deficiency. 2000. Unpublished Work9. Kalivianakis M, Verkade HJ. The mechanism of fat malabsorption in cystic fibrosis patients. Nutrition 1999;15:167-9.


10. Otte JB, De Ville De Goyet J, Reding R et al. Sequential treatment of biliary atresia with Kasai portoenterostomy and livertransplantation: a review. Hepatology 1994;20:41S-8S.11. Holman RT, Pusch F, Svingen B, Dutton HJ. Unusual isomeric polyunsaturated fatty acids in liver phospholipids of rats fedhydrogenated oil. Proc.Natl.Acad.Sci.U.S.A. 1991;88:4830-4.12. Balistreri WF. Liver and biliary system - development and function. In: Behrman RE, Kliegman RM, Arvin AM, Nelson WE,eds. Nelson textbook of paediatrics. Philadelphia: W.B.Saunders company 1996:1125-58.13. Burr GO, Burr MM. A new deficiency disease produced by the rigid exclusion of fat from the diet. Journal of biological chem-istry 1929;LXXXII:345-67.14. Schumm DE. Lipid biosynthesis, Lipid degradation. In: Schnittman ER, Marnhout R, eds. Essentials of biochemistry. Little,Brown and company inc. 1995:203-25.15. Carey MC, Small DM, Bliss CM. Lipid digestion and absorption. Ann.Rev.Physiol. 1983;45:651-77.16. Minich DM, Voshol PJ, Havinga R et al. Biliary phospholipid secretion is not required for intestinal absorption and plasmastatus of linoleic acid in mice. Biochimica et biophysica acta 1999;1441:14-22.17. Yamanaka WK, Clemans GW, Hutchinson ML. Essential fatty acid deficiency in humans. Prog.Lipid Res. 1981;19:187-215.18. Bottina NR. Metabolism of unsaturated fatty acids in the intestine. Lipids 1966;2:155-60.19. Verkade HJ, Bijleveld CMA, Werner A. Intestinale vetmalabsorptie: pathofysiologie en diagnostiek. TijdschriftKindergeneeskunde 2000;68:175-82.20. Voshol PJ, Minich DM, Havinga R et al. Postprandial chylomicron formation and fat absorption in multidrug resistance gene2 p-glycoprotein-deficient mice. Gastroenterology 2000;118:173-82.21. Setchell KDR, O'Connell NC. Inborn errors of bile acid biosynthesis. Bile acids in gastroenterology. 2000:129-36.22. Suchy FJ. Neonatal cholestasis. In: Manns, Boyer, Jansen, Reichen: Cholestatic liver diseases. Kluwer 1998:124-34.23. Dellert SF, Balistreri WF. Neonatal cholestasis. In: Walker, Durie, Hamilton, Walker-Smith, Watkins, eds. Paediatric gastroin-testinal disease. 1996:999-1012.24. Pleskow RG, Grand RJ. Wilson's disease. In: Walker, Durie, Hamilton, Walker-Smith, Watkins, eds. Paediatric gastrointesti-nal disease. 1996:1233-42.25. Storch J. The role of fatty acid binding proteins in enterocyte fatty acid transport. In: Tso P, Mansbach CM, Kuksis A, eds.Intestinal lipid metabolism. Kluwer 2001:153-70.26. Jansen PLM, Strautnieks SS, Jacquemin E et al. Liver pancreas and biliary tract - Hepatocanalicular bile salt export pumpdeficiency in patients with progressive familial intrahepatic cholestasis. Gastroenterology 1999;117:1370-9.27. Bezerra JA, Balistreri WF. Intrahepatic cholestasis: order out of chaos. Gastroenterology 1999;117:1496-8.28. Bull LN, van Eijk MJT, Pawlikowska L. Progressive familial intrahepatic cholestasis types 1, 2, and 3. Gut 1998;42:766-7.29. Whitington PF, Freese DK, Alonso EM, Schwarzenberg SJ, Sharp HL. Clinical and biochemical findings in progressive famil-ial intrahepatic cholestasis. Journal of paediatric gastroenterology and nutrition 1994;18:134-41.30. Knisely AS. Progressive familial intrahepatic cholestasis: a personal perspective. Pediatr.Dev.Pathol. 2000;3:113-25.31. Kleinman RE, D'Agata ID, Vacanti JP. Liver transplantation. In: Walker, Durie, Hamilton, Walker-Smith, Watkins, eds.Paediatric gastrointstinal disease. 1996:1340-58.32. Minich DM, Vonk RJ, Verkade HJ. Intestinal absorption of essential fatty acids under physiological and essential fatty acid-deficient conditions. Journal of lipid research 1997;38:1709-21.33. Boehm G, Borte M, Boehles HJ, Mueller H, Kohn G, Moro G. DHA and AA content of serum and RBC membrane PL ofpreterm infants fed breast milk, standard formula or formula supplemented with n-3 and n-6 LCPUFA. Eur J Pediatr1996;155:410-6.34. Horrobin DF, Huang YS, Cunnane SC, Manku MS. EFA in plasma, red blood cells and liver phospholipids in common lab-oratory animals as compared to humans. Lipids 1984;19:806-1135. Hoffman DF, Uauy R. Essentiality of dietary omega-3 fatty acids for premature infants: plasma and red blood cell fatty acidcomposition. Lipids 1992;27:886-95.36. Bernard A, Caselli C, Carlier H. Linoleic acid chyloportal partition and metabolism during its intestinal absorption. Ann nutrmetab 1991;35:98-110.

54

Chapter 2


Essential fatty acid deficiencyin mice is associated with

hepatic steatosis and secretion of large VLDL particles3

Am J Physiol Gastrointest Liver Physiol. 2005; 288(6): G1150-1158

A. WernerH. Havinga

T. Bos V.W. Bloks F. Kuipers

H.J. Verkade


56

Chapter 3

ABSTRACT

Background: Essential fatty acid (EFA) deficiency in mice decreases plasma

triglyceride (TG) concentrations and increases hepatic TG content. We evaluated

in vivo and in vitro whether decreased hepatic secretion of TG-rich VLDL contributes

to this consequence of EFA deficiency.

Methods: EFA deficiency was induced in mice by feeding an EFA-deficient (EFAD)

diet for eight weeks. Hepatic VLDL secretion was quantified in fasted EFAD and EFA-

sufficient (EFAS) mice using the Triton WR-1339 method. In cultured hepatocytes

from EFAD and EFAS mice, VLDL secretion into medium was measured by

quantifying [3H]-glycerol incorporation into TG and phospholipids (PL). Hepatic

expression of genes involved in VLDL synthesis and clearance was measured, as

were plasma activities of lipolytic enzymes.

Results: TG secretion rates were quantitatively similar in EFAD and EFAS mice in vivo

and in primary hepatocytes from EFAD and EFAS mice in vitro. However, EFA

deficiency increased the size of secreted VLDL particles, as determined by

calculation of particle diameter, particle sizing by light scattering and evaluation of

TG-to-apoB ratio. EFA deficiency did not inhibit hepatic lipase and lipoprotein lipase

activities in plasma, but increased hepatic mRNA levels of apoAV and apoC-II,

both involved in control of lipolytic degradation of TG-rich lipoproteins.

Conclusions: EFA deficiency does not affect hepatic TG secretion rate in mice,

but increases the size of secreted VLDL particles. Present data suggest that

hypotriglyceridemia during EFA deficiency is related to enhanced clearance of

altered VLDL particles.


57

EFA deficiency in mice is associated with hepatic steatosis and secretion of large VLDL particles

INTRODUCTION

Development of hepatic steatosis is a well-established manifestation of essential fatty

acid (EFA) deficiency in animal models. It was first described in 1958 in rats by Alfin-

Slater et al.(1) and in 1970 by Fukazawa and Sinclair et al.(2; 3). The excess lipid deposit-

ed in the liver during EFA deficiency can theoretically result from increased uptake of

circulating lipids, enhanced de novo lipogenesis, decreased fatty acid oxidation,

decreased hepatic lipoprotein secretion, or from a combination of these. Increased

hepatic lipogenesis and decreased fatty acid oxidation could indeed contribute,

since polyunsaturated fatty acids are physiological suppressors of fatty acid synthe-

sis through down-regulation of SREBP1c(4-8), and inducers of hepatic fatty acid

oxidation through activation of PPAR-alpha (PPARa), respectively. The quantitative

contribution of increased lipogenesis and decreased oxidation to EFA deficiency-

induced hepatic steatosis has not been established. While the effects of EFA

deficiency on induction of hepatic steatosis are fairly consistent in the literature, the

consequences for hepatic lipoprotein secretion and plasma lipid profiles are less

clear. Fukazawa et al.(3) reported decreased triglyceride (TG) and phospholipid (PL)

secretion from perfused livers of EFAD rats. However, EFA deficiency has also been

associated with enhanced hepatic TG secretion rates in rats(9-12). Similarly, data on

lipoprotein clearance during EFA deficiency are equivocal. Activities of plasma

lipoprotein lipase (LPL) and hepatic lipase (HL) were reported to be increased in

EFAD rats by Nilsson et al.(2; 13-14), whereas Levy and colleagues described decreased

plasma LPL activity in EFAD rats(15; 16).

Recently, we characterized a mouse model for EFA deficiency in which hepatic TG

levels are increased and plasma TG concentrations are decreased(17). To preclude the

confusion from isolated studies on EFA deficiency in different species and models,

we have chosen to characterize this mouse model in detail. Previously, we reported

characteristics of EFAD mice with respect to growth, intestinal fat absorption, bile

formation and fatty acid composition in specific organs(17-19). In the present study, we

investigated whether EFA deficiency affects hepatic VLDL secretion in mice in vivo

and in isolated mouse hepatocytes in vitro. Our data indicate that EFA deficiency in

mice does not quantitatively affect hepatic VLDL-TG secretion but increases VLDL

particle size. We hypothesize that clearance rate of these large lipoproteins is

increased to yield low plasma TG levels.


58

Chapter 3


MaterialsTriton WR-1339, Triton X-100, fatty acid-free bovine serum albumin (BSA), oleic acid

and heptadecanoic acid were obtained from Sigma Chemical Co. (St. Louis, MO,

USA). [3H]-glycerol was purchased from New England Nuclear (Boston, MA, USA),

glycerol tri-[9,10(n)-[3H]-oleate from Amersham Biosciences (Buckinghamshire, UK)

and glycerol trioleate from Fluka Chemie, Sigma-Aldrich (Zwijndrecht, the

Netherlands). 4-15% SDS ready gels were from Biorad (Hercules, CA, USA), heparin

was obtained from Leo Pharma BV (Weesp, the Netherlands) and all cell culture

materials were from Costar (Cambridge, MA, USA).

AnimalsMale wildtype mice with a free virus breed (FVB) background were obtained from

Harlan (Horst, the Netherlands). When starting the experimental diets, mice were

approximately eight weeks old and were housed in a light-controlled (lights on 6 AM-

6 PM) and temperature-controlled (21°C) facility with free access to tap water and

standard laboratory chow (RMH-B, Arie Blok BV, Woerden, the Netherlands).

The experimental protocols were approved by the Ethics Committee for Animal

Experiments, Faculty of Medical Sciences, University of Groningen, the Netherlands.

Experimental dietsThe EFAD diet contained 20 energy% protein, 46 energy% carbohydrate and 34

energy% fat, respectively, and had the following fatty acid composition: 41.4 mol%

palmitic acid (C16:0), 47.9 mol% stearic acid (C18:0), 7.7 mol% oleic acid

(C18:1n-9) and 3 mol% linoleic acid (C18:2n-6). An isocaloric EFA-sufficient (EFAS)

diet was used as control diet, containing 20 energy% protein, 43 energy% carbo-

hydrate and 37 energy% fat with 32.1 mol% C16:0, 5.5% C18:0, 32.2 mol%

C18:1n-9 and 30.2% C18:2n-6 (custom synthesis, diet numbers 4141.08 (EFAD) and

4141.07 (EFAS), respectively; Arie Blok BV, Woerden, the Netherlands).

Experimental proceduresInduction of EFA deficiency in mice

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

switched to EFAD or EFAS diet at 8 weeks of age. After 8 weeks on EFAD or EFAS

diet, 6 mice of each dietary group were anesthetized by halothane/NO2 and a large

blood sample was obtained by cardiac puncture for determination of plasma lipid

levels, lipoprotein profile and plasma and erythrocyte fatty acid composition.


59


Blood was collected in heparinized tubes and plasma and erythrocytes were

separated by centrifugation at 2400 rpm for 10 min (Eppendorf Centrifuge,

Eppendorf, Germany). Fresh erythrocyte samples were hydrolyzed and methylated(20)

for gas-chromatographic analysis of fatty acid profiles. After liver excision, tissue

aliquots (30 mg) were immediately stored in liquid nitrogen for mRNA isolation.

The remaining liver tissue was stored at -80°C until further analysis.

Fast Protein Liquid Chromatography (FPLC)

For plasma lipoprotein size fractionation, 200 µl of pooled plasma from EFAD- and

EFAS-diet fed mice (n=6 per group) was separated by FPLC on a Superose 6

HR10/30 column (Amersham Pharmacia Biotech, Uppsala, Sweden). Triglyceride,

phospholipid and cholesterol concentrations in the obtained fractions (0.5 ml) were

measured as described below.

In vivo VLDL secretion in EFAD and EFAS mice

In mice fed EFAD or EFAS diet for 8 weeks (n=6 per group), plasma lipolysis was

blocked by retro-orbital injection of Triton WR-1339 (12.5 mg/100 µl phosphate-

buffered saline) after an overnight fast. Blood samples (75 µl) were obtained from the

retro-orbital plexus under halothane anesthesia, before and at 60 min intervals after

Triton injection for 4 hours. Blood was collected in micro-hematocrit tubes containing

heparin, and was centrifuged at 2400 rpm for 10 min (Eppendorf Centrifuge,

Eppendorf, Germany) for isolation of plasma and blood cells. At the end of the

experiment, a large blood sample was obtained by cardiac puncture, after which the

liver was removed and stored at -80°C until further analysis. From the last blood

sample, the plasma VLDL fraction (d 0.93-1.006 g/ml) was isolated by ultra-

centrifugation. For this purpose, 800 µl of NaCl solution with a density of 1.006 g/ml,

containing 0.02% NaN3, was added to 200 µl plasma, followed by centrifugation for

100 min at 120 000 rpm at 4°C in an Optima TM LX table top centrifuge (Beckman

Instruments, Inc., Palo Alto, CA, USA). The top layer containing the VLDL fraction was

isolated by tube slicing, and the volume was recorded by weight. A 30 µl portion was

used for particle size determination using dynamic light scattering (details see below)

and the remaining VLDL fraction was stored at -80°C until further analysis.

Post-heparin HL and LPL activity in plasma of EFAD and EFAS mice

Separate groups of EFAD and EFAS mice (n=6 per group) were fasted for 4 hours,

after which a baseline blood sample (150 µl) was obtained by orbital bleeding under

halothane anesthesia, for determination of baseline plasma lipase activity.

Subsequently, an intravenous bolus of 0.1 U of heparin per gram bodyweight was


60

Chapter 3

injected, and 10 min later a post-heparin blood sample (150 µl) was obtained by

orbital bleeding. Blood was collected in heparinized micro-hematocrit tubes, imme-

diately centrifuged at 2400 rpm for 10 min (Eppendorf Centrifuge, Eppendorf,

Germany) and isolated plasma was frozen in 10% glycerol in liquid nitrogen and

stored at -80°C until in vitro analysis of LPL and HL activities(21).

For the LPL and HL assay, 10 µl of plasma was incubated with 200 µl of ultrasonified

substrate containing 1 ml Triton X-100 (1%), 1 ml Tris-HCl (1M), 2 ml of heat-

inactivated human serum, 2 ml of fat-free BSA (10%), 42 mg triolein and 5 µl glycerol-

tri-(9,10(n)-[3H])-oleate (5mCi/ml), with or without addition of 50 µl NaCl (5M) to block

LPL activity. After 30 min incubation at 37°C, lipolysis was stopped by adding 3.25 ml

of heptane / methanol / chloroform (100/128/137, v/v/v) and 1 ml of 0.1M K2CO3.

After centrifugation for 15 min at 3600 rpm at room temperature, extracted

hydrolyzed fatty acids were quantified by scintillation counting. Lipase activities were

calculated according to the formula: [dps sample - dps blank] / dps 200 µl LPL-

substrate x factor, in which the factor = (2.45 (volume aqueous phase) x 4.74 (total

added FFA in mol) / [0.76 (extraction efficiency) x 0.5 (reaction time in h) x 0.01

(plasma volume in ml)]). Post-heparin LPL activity was calculated by subtracting

post-heparin hepatic lipase activity (i.e., lipase activity inhibited by 1M NaCl) from the

total post-heparin lipase activity.

In vitro VLDL secretion from cultured EFAD and EFAS hepatocytes

Isolation of hepatocytes from mice fed EFAD or EFAS diet for 8 weeks was performed

as described previously(22; 23). Hepatocytes were plated in 35 mm 6-well plastic dishes

pre-coated with collagen (Serva Feinbiochemica, Heidelberg, Germany) at a density

of 1.0x106 cells per well, suspended in 2 ml of William's E medium (Gibco BRL, Grand

Island, NY, USA) supplemented with 10% fetal calf serum (FCS), 0.20 U/ml insulin,

100 U/ml penicillin, 100 µg/ml streptomycin, µ50 g/ml gentamycin and 50 nM dexa-

methasone. Cells were maintained in a humidified incubator at 37°C and 5% CO2.

After a 5 hour attachment period the medium was refreshed. Cells were cultured

overnight, medium was removed and hepatocytes were washed and incubated for 4

hours with hormone-free and FCS-free (HF-SF) William's E medium supplemented

with 1.7% fat-free albumin, insulin, penicillin / streptomycin and gentamycin. Medium

was replaced by 1 ml HF-SF William's E medium per well containing 22 µM [3H]-

glycerol (4.4 µCi per well), 3 µM glycerol, 0.75 mM oleic acid (C18:1) complexed with

fatty acid-free BSA. After 24 hours incubation, medium was collected and centrifuged

for 2 min at 13000 rpm to remove debris, and stored at -80°C until further analysis.

Hepatocytes were washed with ice-cold Hank's balanced salt solution (HBSS) and

scraped into 2 ml of HBSS for lipid extraction.


61


Analytical techniquesPlasma lipids were measured using commercially available assay kits from Roche

(Mannheim, Germany) for triglyceride and total cholesterol, and from WAKO chemi-

cals GmbH (Neuss, Germany) for phospholipids. ApoB protein levels were deter-

mined by Western blotting. Proteins from plasma VLDL fractions (10 µl VLDL/lane)

were separated on 4-15% ready gels and blotted onto nitrocellulose membranes

(Hybond ECL, Amersham Pharmacia Biotech, Buckinghamshire, UK) by tank-

blotting. Membranes were blocked overnight in a 4% skimmed milk power solution

in Tris-buffered saline containing 0.1% Tween-20 (TTBS) and subsequently incu-

bated with the primary antibody (human polyclonal anti apoB, cross-reactive with

mouse, Roche, Mannheim Germany, 726494) diluted 1:100000 in TTBS for 2 hours

at room temperature. After washing, anti-sheep IgG linked to horseradish peroxidase

(Calbiochem, San Diego, CA, 402100), diluted 1:10000 in TTBS was added for 1

hour. Detection was carried out using ECL, according to manufacturer's instructions

(Amersham, Roosendaal, the Netherlands) and bands of apoB were quantified using

Image Masters VDS system (Amersham Pharmacia Biotech, Uppsala, Sweden).

VLDL size and volume distribution profiles were analyzed by dynamic light

scattering, using a Nicomp model 370 submicron particle analyzer (Nicomp Particle

Sizing Systems, Santa Barbara, CA, USA). Particle diameters were calculated from

the volume distribution patterns provided by the analyzer. TG-rich lipoprotein diame-

ters were also estimated using the equation according to Fraser and Harris et al.(24; 25:

diameter (nm) = 60 x ([0.211xTG/PL] + 0.27).

Essential fatty acid status was analyzed by hydrolyzing, methylating and extracting

plasma and erythrocyte lipids as described previously(20). For fatty acid analysis of

cultured hepatocytes, lipids were extracted from aliquots of mechanically homoge-

nized cell suspensions(26), followed by methylation procedures as described above.

Butylated hydroxytoluene was added as antioxidant. Heptadecanoic acid (C17:0)

was added to all samples as internal standard prior to extraction. Fatty acid methyl

esters were separated and quantified by gas liquid chromatography on a Hewlett

Packard gas chromatograph model 6890, equipped with a 50mx0.2mm Ultra 1

capillary column (Hewlett Packard, Palo Alto, CA) and a FID detector, using program

conditions as described previously(17). Individual fatty acid methyl esters were

quantified by relating areas of their chromatogram peaks to that of the internal

standard C17:0. Relative concentrations (mol%) of erythrocyte and hepatocyte fatty

acids were calculated by summation of fatty acid peak areas and subsequent

expression of the area of each individual fatty acid as a percentage of this amount.


62

Chapter 3

Lipids secreted into medium by EFAD and EFAS mouse hepatocytes and cellular

lipids were subjected to the lipid extraction procedure mentioned above. [3H]-TG and

[3H]-PL fractions were isolated from lipid extracts using thin-layer chromatography

(TLC) (20x20 cm, Silica gel 60 F254, Merck), with hexane/diethyl-ether/acetic acid

(80:20:1, v/v/v) as solvent. After iodine staining, the [3H]-containing TG- and PL-spots

were delineated and scraped into vials and assayed for radioactivity by scintillation

counting. A portion of the extracted lipids was dissolved in chloroform containing

2% Triton X-100. After chloroform evaporation and resuspension in H2O, total cellular

TG concentration was determined using the TG assay kit mentioned previously.

Protein concentrations in isolated mouse hepatocytes were determined according to

Lowry et al.(27), using Pierce bovine serum albumin as standard. Secreted apoB in

medium of EFAD and EFAS mouse hepatocytes was concentrated with fumed silica

and delipidated as described by Vance et al.(28) ApoB protein was separated by SDS-

PAGE using 4-15% gradient gels at 100V for 30 min followed by 150V for 90 min.

Subsequently, gels were subjected to the silver staining procedure as described by

Curtin et al.(29). The relative intensities of apoB100 and apoB48 bands were

determined using a CCD camera of Image Masters VDS system (Amersham

Pharmacia Biotech, Uppsala, Sweden).

For measurement of mRNA expression levels by real-time PCR, total RNA from EFAD

and EFAS liver tissue aliquots was isolated using TRI Reagent (Sigma T9424) accor-

ding to the manufacturer's instructions. Isolated total RNA was converted to single-

stranded cDNA with M-MLV reverse transcriptase by the reverse transcription proce-

dure from the manufacturer's protocol (Sigma). Real-time quantitative PCR was

performed by using the ABI prism 7700 Sequence Detector (Applied Biosystems,

Foster City, CA, USA). Primers were obtained from Invitrogen and a template-specif-

ic 3'-TAMRA, 5'-6-FAM labeled Double Dye Oligonucleotide probe was obtained from

Eurogentec, Seraing, Belgium. Primers and probes used in these studies for Acc1,

ApoB, Fas, Mttp, Srebp1a, Srebp1c, 18S, ß-actin, hmgCoAS-m have been described

previously(30-32). Acc2 forward primer: CAT ACA CAG AGC TGG TGT TGG ACT, reverse

primer: CAC CAT GCC CAC CTC GTT AC, probe: CAG GAA GCC GGT TCA TCT

CCA CCA G, GenBank accession NM_133904, ApoAV forward primer: GAC TAC TTC

AGC CAA AAC AGT TGG A, reverse primer: AAG CTG CCT TTC AGG TTC TCC T,

probe: CTT CTG TGG CTG GCC CAT CAC GC, GenBank accession NM_080434,

ApoC-I forward primer: GGG CAG CCA TTG AAC ATA TCA, reverse primer: TTG CCA

AAT GCC TCT GAG AAC, probe: CCC GGG TCT TGG TCA AAA TTT CCT TC,

GenBank accession NM_007469, ApoC-II forward primer: TTA CTG GAC CTC TGC


63


CAA GGA, reverse primer: CCC TGA GTT TCT CAT CCA TGC, probe: CCA AAG ACC

TGT ACC AGA AGA CAT ACC CGA, GenBank accession NM_009695, ApoC-III

forward primer: CCA AGA CGG TCC AGG ATG C, reverse primer: ACT TGC TCC

AGT AGC CTT TCA GG, probe: CCA TCC AGC CCC TGG CCA CC, GenBank

accession NM_023114, Cpt1a forward primer: CTC AGT GGG AGC GAC TCT TCA,

reverse primer: GGC CTC TGT GGT ACA CGA CAA, probe: CCT GGG GAG GAG

ACA GAC ACC ATC CAA C, GenBank accession NM_013495, Cpt1b forward primer:

CCC ATG TGC TCC TAC CAG ATG, reverse primer: CAC GTG CCT GCT CTC TGA

GA, probe: CCC AGG CAA AGA GAC AGA CTT GCT ACA GC, GenBank accession

NM_009948. All expression data were subsequently standardized for ß-actin, which

was analyzed in separate runs.

Calculations and statisticsAll results are presented as means ± S.D. for the number of animals indicated. Data

were statistically analyzed using Student's t-test or, in absence of normal distribution,

Mann-Whitney-U test. Level of significance was set at p<0.05. Analyses were

performed using SPSS for Windows software (SPSS, Chicago, IL).

RESULTS

In previous studies(17; 18), we developed and characterized a murine model for diet-

induced EFA deficiency by feeding mice an EFAD diet for eight weeks, which

resulted in pronounced biochemical hallmarks of EFA deficiency such as increased

triene/tetraene ratios(33; 34). After eight weeks of EFAD diet feeding, body weights of

mice were significantly lower compared to EFA-fed counterparts (29.5±1.7 g vs.

32.6±3.3 g, p<0.001), which is likely related to impaired dietary fat absorption

during EFA deficiency, as described previously(17; 35; 36). No other clinical characteristics

of EFA deficiency, such as alopecia or tail necrosis, were observed in EFAD mice.

Figure 1 shows that the EFAD diet decreased concentrations of EFA and LCPUFA

and increased levels of non-essential fatty acids in plasma VLDL and in erythrocytes.

plasma VLDL

EFASEFAD

*

0.0

1.0

2.0

**m

ol%

of t

otal

fatty

aci

ds

22:6n-3 TT-ratio20:5n-318:2n-6 20:4n-6 18:1n-9

* *

*

0

50

25

RBC

EFASEFAD

mol

% o

f tot

al fa

tty a

cids

*0.0

1.0

2.0

*

22:6n-3 TT-ratio20:5n-318:2n-6 20:4n-6 18:1n-9

* *0

50

25

plasma VLDL

EFASEFAD

*

0.0

1.0

2.0

**m

ol%

of t

otal

fatty

aci

ds

22:6n-3 TT-ratio20:5n-318:2n-6 20:4n-6 18:1n-9

* *

*

0

50

25

plasma VLDL

EFASEFAD

plasma VLDL

EFASEFADEFASEFAD

*

0.0

1.0

2.0

0.0

1.0

2.0

**m

ol%

of t

otal

fatty

aci

ds

22:6n-3 TT-ratio20:5n-318:2n-6 20:4n-6 18:1n-9 22:6n-3 TT-ratio20:5n-318:2n-6 20:4n-6 18:1n-9

* *

*

0

50

25

0

50

25

RBC

EFASEFAD

mol

% o

f tot

al fa

tty a

cids

*0.0

1.0

2.0

*

22:6n-3 TT-ratio20:5n-318:2n-6 20:4n-6 18:1n-9

* *0

50

25

RBC

EFASEFAD

RBC

EFASEFADEFASEFAD

mol

% o

f tot

al fa

tty a

cids

*0.0

1.0

2.0

0.0

1.0

2.0

*

22:6n-3 TT-ratio20:5n-318:2n-6 20:4n-6 18:1n-9 22:6n-3 TT-ratio20:5n-318:2n-6 20:4n-6 18:1n-9

* *0

50

25

0

50

25

Figure 1: Linoleic acid (C18:2n-6), arachidonic acid (C20:4n-6), oleic acid (C18:1n-9), eicosapentaenoic acid (C20:5n-3),docosahexaenoic acid (C22:6n-3) and the TT-ratio in plasma VLDL and in RBC of EFAD and EFAS mice. Fatty acid concentra-tions are in mol% of total fatty acids.Data represent means±SD of 6 mice per group. *p<0.05 for EFAD vs. EFAS.

1a 1b


Effects were more pronounced in VLDL than in erythrocytes. Similar to previous

studies, plasma triglyceride concentration was decreased in EFAD mice (EFAD

0.4±0.1 mM, EFAS 0.8±0.6 mM, p<0.05), which was predominantly due to a

decrease in the VLDL-sized lipoprotein fraction (Figure 2). Cholesterol concentrations

were higher in fractions 15-20 of the FPLC profile of EFAD mice, which may indicate

the presence of large HDL particles, or increased amounts of IDL/LDL-sized particles.

Figure 3 shows that, upon lipolysis blockage by Triton WR-1339, plasma accu-

mulation of TG was similar in EFAD and EFAS mice, indicating equal VLDL-TG

production rates (193±78 vs. 173±46 µmol TG/kg/h for EFAD and EFAS mice,

respectively; NS). Figure 4 shows that TG and PL levels in the VLDL fraction isolated

4 hours after Triton administration were similar in EFAD and EFAS mice.

64

Chapter 3

cho

l(µM

)

0

100

200

300

400

500

600

0 5 10 15 20 25 30 35 40

PL

(µM

)

0

100

200

300

400

0 5 10 15 20 25 30 35 40

FPLC plasma

TG

(µM

)

0

25

50

75

100

0 5 10 15 20 25 30 35 40

EFASEFAD

VLDL HDLIDL/LDL

cho

l(µM

)

0

100

200

300

400

500

600

100

200

300

400

500

600

0 5 10 15 20 25 30 35 400 5 10 15 20 25 30 35 40

PL

(µM

)

0

100

200

300

400

0 5 10 15 20 25 30 35 40

PL

(µM

)

0

100

200

300

400

0 5 10 15 20 25 30 35 400

100

200

300

400

0

100

200

300

400

0 5 10 15 20 25 30 35 40

FPLC plasma

TG

(µM

)

0

25

50

75

100

0 5 10 15 20 25 30 35 40

EFASEFAD

VLDL HDLIDL/LDL

FPLC plasma

TG

(µM

)

0

25

50

75

100

0

25

50

75

100

0 5 10 15 20 25 30 35 400 5 10 15 20 25 30 35 40

EFASEFADEFASEFAD

VLDL HDLIDL/LDLVLDLVLDL HDLHDLIDL/LDLIDL/LDL

Figure 2: Plasma triglyceride (TG), cholesterol (chol) andphospholipid (PL) concentrations in FPLC fractions ofEFA-deficient (EFAD) and EFA-sufficient (EFAS) mice.Data represent lipid concentrations in pooled plasmasamples of 6 mice per group.

plasma TG

incr

ease

in T

G c

once

ntra

tion

(mM

)

EFAS

EFAD

h after Triton 0

10

20

30

0 1 2 3 4

µmol

TG

/kg/

h

0

100

300

EFAS EFAD

200

plasma TG

incr

ease

in T

G c

once

ntra

tion

(mM

)

EFAS

EFAD

EFAS

EFAD

h after Triton 0

10

20

30

0 1 2 3 4

µmol

TG

/kg/

h

0

100

300

EFAS EFAD

200

µmol

TG

/kg/

h

0

100

300

EFAS EFAD

200

Figure 3: Increase in plasma triglyceride (TG) concentra-tion in mice fed EFA-deficient (EFAD) or EFA-sufficient(EFAS) diet for 8 weeks, before and at 60 min intervalsafter Triton WR-1339 injection. Plasma accumulation ofTG was similar in EFAD and EFAS mice, indicating equalVLDL-TG production rates (193±78 vs. 173±46 µmolTG/kg/h for EFAD and EFAS mice, respectively; NS). Data represent means ± SD of 6 mice per group.

2a

2b

2c

3


ApoB100 concentrations in this fraction were also similar in EFAD and EFAS mice,

yet, apoB48 was significantly lower in VLDL of EFAD mice (Figure 4c, p<0.05). The

decreased apoB48 concentration but the unaffected TG secretion rate suggests an

increased size of secreted VLDL particles during EFA deficiency.

Lipoprotein particle size in plasma of EFAD mice was estimated by three methods.

The core-to-surface ratio in the isolated TG-rich lipoprotein fraction (4 hours after

Triton) was significantly higher in EFAD than in EFAS mice (6.6±0.9 vs. 4.9±1.2,

p<0.05, Figure 5a). Upon calculation of lipoprotein diameters according to Fraser

and Harris et al.(24; 25), TG-rich lipoproteins from EFAD mice similarly appeared larger

than from EFAS mice (Figure 5b, p<0.05). Finally, determination of TG-rich lipo-

protein size using dynamic light scattering also indicated that plasma lipoproteins

from EFAD mice were larger than from EFAS mice; however, this difference did not

reach statistical significance (p=0.37, Figure 5c).

65


EFADEFAS

VLDL lipids 4 h after Triton

0

10

20

30

chol TG PL

mM

EFADEFASEFADEFADEFASEFAS


0

10

20

30

chol TG PL

mM


0

10

20

30

chol TG PL0

10

20

30

0

10

20

30

chol TG PL

mM

Triglyceride (TG), phospholipid (PL) and cholesterol in plasma VLDL(4a), relative TG, PL and cholesterol in plasma VLDL,expressed as % of total lipid.(4b), ApoB100 and apoB48 in plasma VLDL of EFAD and EFAS mice (4c). Plasma VLDL was iso-lated 4h after Triton administration. Data represent means±SD of 6 mice / group, *p<0.05 for apoB48 of EFAD vs. EFAS mice.

VLDL lipids 4h after Triton

0

20

40

60

80

100

%chol %TG %PL

%of

tota

l lip

id EFADEFAS


0

20

40

60

80

100

%chol %TG %PL

%of

tota

l lip

id


0

20

40

60

80

100

%chol %TG %PL0

20

40

60

80

100

0

20

40

60

80

100

%chol %TG %PL

%of

tota

l lip

id EFADEFASEFADEFADEFASEFAS

VLDL apoB 4h after Triton

band

inte

nsity

/ m

l VLD

L

0

20

40

60

apoB100 apoB48

*

EFADEFAS


band

inte

nsity

/ m

l VLD

L

0

20

40

60

apoB100 apoB48

*


band

inte

nsity

/ m

l VLD

L

0

20

40

60

apoB100 apoB48

*

0

20

40

60

0

20

40

60

apoB100 apoB48apoB100 apoB48

*

EFADEFASEFADEFADEFASEFAS

EFASEFAD

[TG

] / [P

L]

core / surface ratio

EFAS EFAD

*

0

10

5

EFASEFAD

[TG

] / [P

L]

core / surface ratio

EFAS EFADEFAD

*

0

10

5

0

10

5

Core-to-surface ratio in the isolated VLDL fraction (4h after Triton) of EFAD and EFAS mice, estimated by the ratio of TG (mM)and PL (mM). Data represent means ± SD of 5-6 mice per group, *p<0.05 for EFAD vs. EFAS mice. (5a); Lipoproteindiameter (nm) calculated as: diameter (nm) = 60 x ([0.211 x TG/PL] +0.27). TG-rich lipoproteins from EFAD mice were signif-icantly larger than those from EFAS mice (5b). Data represent means ± SD of 6 mice / group, *p<0.05 for EFAD vs. EFAS mice.TG-rich lipoprotein size (nm) measured by dynamic light scattering (5c). Data represent means ± SD of 6 mice per group,

p=0.37 for EFA-deficient vs. EFA-sufficient mice.

nm

calculated diameter

0

50

100

150

EFAS EFAD

*EFASEFAD

nm

calculated diameter

0

50

100

150

0

50

100

150

EFAS EFAD

*EFASEFAD

particle size determined by light scattering

0

50

100

150

EFAS EFAD

nm

EFASEFAD

particle size determined by light scattering

0

50

100

150

0

50

100

150

EFAS EFAD

nm

EFASEFAD

4a 4b 4c

5a 5b 5c


In addition to quantifying VLDL-TG production in mice in vivo, VLDL-TG production

was determined in cultured hepatocytes from EFAD and EFAS mice in vitro, to

exclude a possible influence of confounding metabolic effects of the systemic

circulation on VLDL production.

Figure 6 shows that primary hepatocytes from EFAD mice cultured for 48 hours

displayed the classical biochemical markers of EFA deficiency, including decreased

levels of LA, ALA and their long-chain metabolites AA and DHA and increased

concentrations of non-essential fatty acids of the n-7 and n-9 family. Biochemical

indications for EFA deficiency were more pronounced in TG than in PL of

hepatocytes, as described previously(19).

After 24 hours of incubation with [3H]-glycerol and oleic acid, EFAD hepatocytes had

incorporated significantly more label into TG and PL than EFAS hepatocytes, com-

patible with higher rates of TG and PL synthesis (Figure 7a). Total intracellular TG

mass was approximately two-fold higher in hepatocytes from EFAD mice, compared

with those from EFAS mice (figure 7b). Specific cellular TG activity was similar in

EFAD and EFAS cells (Figure 7c). EFAD hepatocytes secreted similar amounts of

[3H]-labeled TG into the medium, but significantly less phospholipids compared to

EFAS mice (Figure 7d). The apoB content was lower in medium from EFAD than from

EFAS cells (Figure 7e), indicating secretion of a decreased number of particles.

Since hepatic TG production was not affected by EFA deficiency in mice, hypo-

triglyceridemia is likely attributable to accelerated clearance of TG-rich lipoproteins.

Determination of mRNA levels of genes involved in hepatic lipogenesis and VLDL

assembly, using real-time quantitative PCR, showed a significantly increased

expression in EFAD mice of Acc1, which produces malonyl-CoA for fatty acid

synthesis (Figure 8).

66

Chapter 3

0

5

10

15

20

mo

l% o

f to

tal f

atty

aci

ds EFAS

EFAD

18:2n-6 18:3n-3 20:4n-6 22:6n-3 16:1n-7 18:1n-7

**

**

* *

fatty acid composition hepatocyte TG

fatty acid composition hepatocyte PL

0

5

10

15

20

mo

l% o

f to

tal f

atty

aci

ds EFAS

EFAD

18:2n-6 18:3n-3 20:4n-6 22:6n-3 16:1n-7 18:1n-7

**

*

*

*

0

5

10

15

20

mo

l% o

f to

tal f

atty

aci

ds EFAS

EFADEFASEFASEFADEFAD

18:2n-6 18:3n-3 20:4n-6 22:6n-3 16:1n-7 18:1n-7

**

**

* *





0

5

10

15

20

mo

l% o

f to

tal f

atty

aci

ds EFAS

EFADEFASEFASEFADEFAD

18:2n-6 18:3n-3 20:4n-6 22:6n-3 16:1n-7 18:1n-7

**

*

*

*

Figure 6: Fatty acid composition of phospholipids (PL)and triglycerides (TG) of cultured hepatocytes from EFA-deficient (EFAD) and EFA-sufficient (EFAS) mice.Concentrations of linoleic acid (LA, C18:2n-6), alpha-linolenic acid (ALA, C18:3n-3), arachidonic acid (AA,C20:4n-6), docosahexaenoic acid (DHA, C22:6n-3),palmitoleic acid (PA, C16:1n-7) and oleic acid (OA,C18:1n-9), are expressed as mol% of total fatty acids.Data represent means ± SD of 6 mice per group.*p<0.05for differences between EFAD and EFAS hepatocytes.

6a

6b


Expression of Fas, a critical gene for lipogenesis, tended to be higher in EFAD livers,

but the difference did not reach statistical significance. Hepatic expression of Acc2,

Cpt1a and Cpt1b was significantly increased in EFAD mice, compatible with

increased hepatic fatty acid oxidation. mRNA levels of ApoB and Mttp, key regulators

67


3H incorporation hepatocytes

0

50

100

150

200

250

EFASTG

EFASPL

dpm

/ mg

prot

ein

EFADTG

*

EFAD PL

*

3H incorporation hepatocytes

0

50

100

150

200

250

0

50

100

150

200

250

EFASTG

EFASTG

EFASPL

EFASPL

dpm

/ mg

prot

ein

EFADTG

*

EFADTG

*

EFAD PL

*

EFAD PL

*

Figure 7a: [3H]-incorporation into TG and PL by EFAD and EFAS hepatocytes after 24h of incubation with [3H]-glycerol andoleic acid. Data represent mean values ± SD measured in hepatocytes from 2 individual mice per group, n=6 separate meas-urements per mouse. *p<0.001 for differences between EFAD and EFAS hepatocytes.Figure 7b: Total intracellular TG mass(µmol/mg protein) in hepatocytes from EFAD and EFAS mice. Data represent mean values±SD measured in hepatocytes from2 mice per group, n=6 separate measurements per mouse. *p<0.001 for differences between EFAD and EFAS hepatocytes.Figure 7c: Specific cellular TG activity (dpm/nmol TG) in cultured hepatocytes from EFAD and EFAS mice (n=2 mice pergroup). Data represent means ± SD from 6 separate measurements per mouse, p=0.08 for EFAD vs. EFAS hepatocytes.Figure 7d: [3H]-TG and [3H]-PL secretion into medium (dpm x 103 / mg protein) by hepatocytes from EFAD and EFAS mice(n=2 mice per group). Data represent means ± SD from 6 measurements per mouse. *p<0.001 for differences between[3H]-PL secretion by EFAD and EFAS hepatocytes. [3H]-TG levels were not significantly different between groups.

Intracellular TG mass

TG (

µmol

/mg

prot

ein)

EFAS EFAD0.0

0.5

1.0

1.5

2.0

2.5 *Intracellular TG mass

TG (

µmol

/mg

prot

ein)

EFAS EFAD0.0

0.5

1.0

1.5

2.0

2.5

EFAS EFAD0.0

0.5

1.0

1.5

2.0

2.5

0.0

0.5

1.0

1.5

2.0

2.5 *

specific cellular TG activity

EFAS EFAD

dpm

/ nm

ol T

G

0

500

1000

1500

specific cellular TG activity

EFAS EFAD

dpm

/ nm

ol T

G

0

500

1000

1500

0

500

1000

1500

*

3H-TG secretion in medium

0

2

4

6

8

10

EFASTG

EFADTG

EFASPL

EFADPL

dpm

x 10

3/ m

g pr

otei

n

*

3H-TG secretion in medium

0

2

4

6

8

10

0

2

4

6

8

10

EFASTG

EFASTGTG

EFADTG

EFADTGTG

EFASPL

EFASPL

EFADPL

EFADPL

dpm

x 10

3/ m

g pr

otei

n

0.0

B48 B100

rela

tive

inte

nsity apoB in medium

1.0

2.0

**EFAD EFADEFAS EFAS

0.0

B48 B100B48 B100

rela

tive

inte

nsity apoB in medium

1.0

2.0

**EFAD EFADEFAS EFAS

7a 7b 7c 7d

Figure 7e: ApoB48 and apoB100 in medium from EFADand EFAS hepatocytes (n=2 mice per group). Datarepresent means ± SD from 6 separate measurementsper mouse. *p<0.01 for differences in apoB content inmedium of EFAD and EFAS hepatocytes.

Hepatic mRNA expression levels of genes involved in VLDL formation and clearance

0.0

0.5

1.0

1.5

2.0

2.5

Acc1

Srebp

1a

Srebp

1cCpt1

aCpt1

bFa

s

m-HmgC

oA-S

ApoB

Mttp

rela

tive

gene

exp

ress

ion

(A.U

.)

EFAS

EFAD

**

ApoA

V18

SApo

CII

ApoCIII

**

*

Acc2

*

ApoC

I

Hepatic mRNA expression levels of genes involved in VLDL formation and clearance

0.0

0.5

1.0

1.5

2.0

2.5

0.0

0.5

1.0

1.5

2.0

2.5

Acc1

Srebp

1a

Srebp

1cCpt1

aCpt1

bFa

s

m-HmgC

oA-S

ApoB

MttpApoB

Mttp

rela

tive

gene

exp

ress

ion

(A.U

.)

EFAS

EFAD

EFASEFAS

EFADEFAD

****

ApoA

V18

SApo

CII

ApoCIII

**

ApoA

V18

SApo

CII

ApoCIII

**

****

**

Acc2

**

ApoC

I

Figure 8: Hepatic mRNA expression levels of genes involved in VLDL formation and clearance, normalized to ß-actin. Datarepresent mean values ± SD of 6 mice per group, p<0.05 for differences in mRNA levels in EFAD vs. EFAS mice.

7e

8


of VLDL formation, were similar in EFAD and EFAS livers. Expression of ApoC-III, a

lipoprotein lipase (LPL) inhibitor, was similar in EFAD and EFAS mice but mRNA

levels of ApoC-II and apoA-V, involved in modulation of lipoprotein lipase activity,

were significantly higher in livers of EFAD mice compared to controls. Figure 9 shows

the in vitro activities of post-heparin plasma hepatic lipase (HL) and lipoprotein lipase

(LPL) in EFAD and EFAS mice. No significant differences in either HL or LPL

activities were detected between the two groups.

DISCUSSION

We previously demonstrated that EFA deficiency in mice is associated with increased

hepatic and decreased plasma TG concentrations(17). In the present study, we

investigated whether decreased hepatic VLDL secretion contributes to these meta-

bolic consequences of EFA deficiency. Our in vivo and in vitro data indicate that EFA

deficiency does not affect quantitative hepatic TG secretion, but alters VLDL size and

composition. We speculate that large VLDL particles, perhaps in combination with

increased apoCII expression, may be subject to increased clearance rates, resulting

in hypotriglyceridemia.

To test this hypothesis experimentally, clearance rates and plasma lipid levels could

be measured in EFAD and EFAS mice after infusing a defined lipoprotein emulsion

of labeled particles, fractionated into homogenous size populations as described by

Rensen et al.(37). Alternatively, although technically more challenging, VLDL particles

could be isolated from EFAD and EFAS mice, labeled ex vivo, and then clearance

rates of EFAD-derived VLDL and of EFAS-derived VLDL could be determined, each

in EFAD and EFAS mice.

Fatty acid profile analyses of erythrocytes, plasma VLDL and isolated hepatocytes

confirmed the presence of EFA deficiency in our mouse model. As previously

demonstrated(17), plasma TG levels were decreased in EFAD mice, particularly in the

68

Chapter 3

Plasma lipase activity

0

50

100

150

HL LPL

TG

hyd

rola

se a

ctiv

ity

(µm

ol F

FA

/h/m

l)

EFAS

EFAD

Plasma lipase activity

0

50

100

150

0

50

100

150

HL LPL

TG

hyd

rola

se a

ctiv

ity

(µm

ol F

FA

/h/m

l)

EFAS

EFAD

EFAS

EFAD

Figure 9: Post-heparin hepatic lipase (HL) and lipoproteinlipase (LPL) activities expressed as triglyceride hydrolaseactivity (µmol FFA/h/ml) measured in plasma of EFAD andEFAS mice (n=5-6 mice per group). No significant differ-ences were detected between the two groups.

9


VLDL fraction as determined by FPLC. Hepatic VLDL-TG secretion, however, was not

decreased in EFAD mice in vivo (determined by the Triton method), or in EFAD

hepatocytes (determined by [3H]-glycerol incorporation) in vitro. Decreased hepatic

TG and PL secretion has been reported in studies with perfused livers of EFAD rats(3).

However, the isolated perfused liver model has its limitations regarding physiological

lipoprotein secretion, due to lack of hormonal and metabolic feedback from the

circulation, with the perfusate usually only containing erythrocytes and fatty acids.

The discrepancy regarding in vivo studies on lipoprotein clearance in EFAD rats(13; 15)

and our EFAD mouse model may be explained by species specificity. Previous

studies have demonstrated that EFA deficiency has different effects on bile formation

in rats and in mice(17; 36; 38).

Although no quantitative differences were detected in hepatic TG secretion rates,

secreted VLDL particles were significantly larger under EFAD conditions. The

production of larger VLDL particles could be deduced from several independent

observations. During EFA deficiency, the concentration of PL in the VLDL fraction was

more profoundly decreased than that of TG (-80% vs. -50%, respectively; Figure 2),

indicating production of particles with increased core-surface ratio. The plasma VLDL

fraction isolated by ultracentrifugation (after Triton) contained less apoB48 in EFAD

than in EFAS mice. Since a single apoB48 molecule is present per VLDL particle, this

indicates secretion of a reduced number of VLDL particles in vivo. In line with these

observations, estimations of particle size by various means also indicated that VLDL

particles in EFAD mice were larger than in controls. In vitro, EFAD hepatocytes

similarly secreted equal amounts of labeled TG but lower amounts of PL and apoB

into the medium. An increase in VLDL particle size has previously been reported in

EFAD rats(15) and EFAD guinea pigs(39; 40). We hypothesize that during EFA deficiency,

the increased concentration of saturated acyl chains of hepatic PL or the decreased

concentration of unsaturated acyl chains (that is, insufficient hepatic EFA-rich PL

availability) affect the surface coating of nascent lipoproteins, resulting in relative TG-

oversaturation of secreted VLDL. Thus, hepatic VLDL-TG secretion rates apparently

are not quantitatively affected during EFA deficiency in mice. By inference, the

decreased plasma TG concentration must be due to increased VLDL clearance.

Differences in VLDL clearance during EFA deficiency in mice could result from

increased activities of lipolytic enzymes such as hepatic lipase and lipoprotein lipase.

However, we found no indications that EFA deficiency affects the in vitro activities of

hepatic or lipoprotein lipase in EFAD mice. Alternatively, VLDL clearance could be

enhanced secondary to alterations of VLDL particles, as was also suggested by

69



Sinclair et al. (2) for enhanced TG clearance from plasma of EFAD rats. In EFAD mice,

intravascular lipoprotein metabolism could be influenced by altered interactions with

apoC-II, with the recently identified apoAV, or with LPL or phospholipid transfer pro-

tein (PLTP)(41; 42). The decreased EFA content of VLDL surface- and core-lipid acyl

chains, as well as the decreased PL/TG ratio, affects the physical structure of VLDL

particles during EFAD, which could increase the affinity or binding sites for apoC-II or

apoA-V. In addition, it could be hypothesized that a more saturated surface layer in

EFAD VLDL can accommodate slightly better the appearance of TG from the core of

the lipoprotein at the interface, where TG serves as substrate for lipases. Hamilton

and Small(43-45) demonstrated that lipoprotein TG is not completely segregated into the

core oil phase, but is also present in small proportions (±3%) intercalated in the PL

surface layer. Although the exact mechanism by which LPL gains access to VLDL-TG

is not known, the surface TG, with carbonyl groups arranged at the aqueous inter-

face, provides the main pool for interaction with lipolytic enzymes. The decreased

PUFA content of lipoprotein-TG and -PL in EFAD mice may enhance incorporation of

TG in the PL monolayer at the aqueous interface, thus increasing accessibility to

lipases. During lipoprotein TG hydrolysis, the transfer of excess surface PL to HDL is

mediated PLTP. Rao et al.(46) reported that small VLDL have less affinity for PLTP than

large. The EFAD VLDL size and particle surface packing may thus affect the binding

affinity for PLTP and thereby its efficiency as a PL carrier, and VLDL metabolism.

Interestingly, both apoA-V and apoC-II mRNA levels were significantly increased

during EFA deficiency in mice, which may be compatible with enhanced VLDL

catabolism. The increased VLDL particle size could also account for increased

clearance from the plasma. In chylomicron studies, Quarfordt and Goodman(47) and

Chajek-Shaul et al.(48) demonstrated that large particles are cleared more rapidly from

plasma than small particles. Production of VLDL with a larger size implies that fewer

particles are being secreted to account for the similar TG production rates. Martins

et al.(49) postulated that particle number strongly affects lipoprotein clearance rate,

with small numbers of particles being cleared more rapidly than large numbers,

possibly due to a receptor-saturable process involving the availability of apoE.

The relation between murine EFA deficiency and VLDL particle size may be related

to the availability of PL for lipoprotein assembly. Under conditions of reduced PL

availability for VLDL assembly, e.g., during choline deficiency in rats, VLDL particles

with an increased core-to-surface ratio are produced(50; 51). We speculate that a similar

situation may apply in murine EFA deficiency. Previously, we demonstrated that EFA

deficiency in mice profoundly increases the amount of PL secreted into bile(17). Biliary

PL are predominantly composed of phosphatidylcholine (PC), similar to PL used for

70

Chapter 3


VLDL assembly. The increased biliary secretion of PC into the intestine may limit the

availability of PC for hepatic lipoprotein assembly, thus leading to production of VLDL

particles of increased size.

In addition to decreasing plasma TG levels, EFA deficiency in mice increased hepat-

ic TG content, both in vivo and in vitro. Interestingly, genes involved in fatty acid

oxidation (Cpt1a, Cpt1b, Acc2) were upregulated in livers of EFAD mice, compatible

with activation of transcription factor PPARa(52). It is well-known that EFA and LCPUFA

are natural ligands for PPARa and it was unexpected that PPARa-regulated genes

were upregulated during EFA deficiency. Possibly, increased levels of non-essential

LCPUFA (n-9 and n-7 family) in EFAD livers can also activate PPARa. Increased de

novo synthesis of n-9, n-7 and saturated fatty acids from acetyl-CoA may engender

increased rates of hepatic TG synthesis during EFA deficiency. Present data suggest

that unimpaired VLDL-TG secretion rates, in combination with increased hepatic TG

synthesis, causes hepatic TG accumulation in EFAD mice.

For speculations on the potential clinical implications of our current observations, we

attempted to relate our findings on the effects of EFA deficiency on lipoprotein

metabolism in mice to reports on cystic fibrosis (CF) patients, in whom EFA

deficiency is frequently observed. Unfortunately, no human CF data are available in

which information is simultaneously provided on presence of steatosis, plasma TG

concentrations, and VLDL particle size and clearance. Yet, indirect indications offer

some support for extrapolation of our present findings to the human condition,

although caution is warranted. Levy and Lepage(53) reported on the combination of

hypertriglyceridemia and diminished plasma PL concentrations in EFAD CF patients

compared to non-CF siblings. Interestingly, plasma VLDL of EFAD CF patients was

relatively TG-enriched compared to non-CF sibs, suggestive of increased particle

size of these lipoproteins in CF. However, a similar finding was reported for non-EFAD

CF patients, and no data on steatosis were provided. In 1999, Lindblad et al.(54)

reported that 35% of CF patients had steatosis, and that the level of the EFA linoleic

acid in plasma PL negatively correlated with the degree of steatosis.

We conclude that the steatosis and hypotriglyceridemia during EFA deficiency in

mice is a combined result of unimpaired hepatic TG secretion, increased hepatic

synthesis of non-essential fatty acids and secretion of large VLDL particles which

may be subject to rapid clearance rates.

AcknowledgementsThe authors would like to thank Baukje Elzinga, Stijntje Bor and Patrick Rensen for

their technical expertise and assistance in the experiments described in this article.

71



REFERENCES1. Alfin Slater RB and Bernick S. Changes in tissue lipids and tissue histology resulting from essential fatty acid deficiency inrats. Am J Clin Nutr 6: 613-624, 1958.2. Sinclair AJ and Collins FD. The effect of dietary essential fatty acids on the concentration of serum and liver lipids in the rat.Br J Nutr 24: 971-982, 1970.3. Fukazawa T, Privett OS and Takahashi Y. Effect of EFA deficiency on lipid transport from liver. Lipids 6: 388-393, 1971.4. Allmann D and Gibson D. Fatty acids synthesis during early LA deficiency in the mouse. J Lipid Res 79:51-62.: 51-62, 1965.5. Mater MK, Thelen AP and Jump DB. AA and PGE2 regulation of hepatic lipogenic gene expression. JLR 40: 1045-1052, 1999.6. Yahagi N, Shimano H, Hasty AH, Amemiya-Kudo M, Okazaki H, Tamura Y, Iizuka Y, Shionoiri F, Ohashi K, Osuga J, HaradaK, Gotoda T, Nagai R, Ishibashi S and Yamada N. A crucial role of sterol regulatory element-binding protein-1 in the regulationof lipogenic gene expression by polyunsaturated fatty acids. J Biol Chem 274: 35840-35844, 1999.7. Hannah VC, Ou J, Luong A, Goldstein JL and Brown MS. Unsaturated fatty acids down-regulate srebp isoforms 1a and 1cby two mechanisms in HEK-293 cells. J Biol Chem 276: 4365-4372, 2001.8. Shimano H, Yahagi N, Amemiya-Kudo M, Hasty AH, Osuga J, Tamura Y, Shionoiri F, Iizuka Y, Ohashi K, Harada K, Gotoda T,Ishibashi S and Yamada N. Sterol regulatory element-binding protein-1 as a key transcription factor for nutritional induction oflipogenic enzyme genes. J Biol Chem 274: 35832-35839, 1999.9. Williams MA, Tinoco J, Hincenbergs I and Thomas B. Increased plasma triglyceride secretion in EFA-deficient rats fed dietswith or without saturated fat. Lipids 24: 448-453, 1989.10. Huang MT and Williams MA. EFA deficiency and plasma triglyceride turnover in rats. Am J Physiol 238: E499-E505, 1980.11. Bird MI and Williams MA. Triacylglycerol secretion in rats: effects of essential fatty acids and influence of dietary sucrose,glucose or fructose. J Nutr 112: 2267-2278, 1982.12. Williams MA, Tinoco J, Yang YT, Bird MI and Hincenbergs I. Feeding pure docosahexaenoate or arachidonate decreasesplasma triacylglycerol secretion in rats. Lipids 24: 753-758, 1989.13. De Pury GG and Collins FD. VLDL and LPL in serum of rats deficient in essential fatty acids. J Lipid Res 13: 268-275, 1972.14. Nilsson A, Hjelte L, Nilsson-Ehle P and Strandvik B. Adaptive regulation of lipoprotein lipase and salt-resistant lipase activ-ities in essential fatty acid deficiency: an experimental study in the rat. Metabolism 39: 1305-1308, 1990.15. Levy E, Thibault L, Garofalo C, Messier M, Lepage G, Ronco N and Roy CC. Combined (n-3 and n-6) essential fatty defi-ciency is a potent modulator of plasma lipids, lipoprotein composition, and lipolytic enzymes. J Lipid Res 31: 2009-2017, 1990.16. Levy E, Delvin E, Peretti N, Bouchard G and Seidman E. Combined effects of EFA deficiency and tumor necrosis factor-alpha on circulating lipoproteins in rats. Lipids 38: 595-601, 2003.17. Werner A, Minich DM, Havinga R, Bloks V, Van Goor H, Kuipers F and Verkade HJ. Fat malabsorption in essential fatty acid-deficient mice is not due to impaired bile formation. Am J Physiol Gastrointest Liver Physiol 283: G900-G908, 2002.18. Minich DM, Voshol PJ, Havinga R, Stellaard F, Kuipers F, Vonk RJ and Verkade HJ. Biliary phospholipid secretion is notrequired for intestinal absorption and plasma status of linoleic acid in mice. Biochimica et biophysica acta 1441: 14-22, 1999.19. Werner A, Havinga R, Kuipers F and Verkade HJ. Treatment of essential fatty acid deficiency with dietary triglycerides orphospholipids in a murine model of extrahepatic cholestasis. Am J Physiol Gastrointest Liver Physiol 286: G822-G832, 2004.20. Muskiet FA, van Doormaal JJ, Martini IA, Wolthers BG and van der SW. Capillary gas chromatographic profiling of total long-chain fatty acids and cholesterol in biological materials. J Chromatogr 278: 231-244, 1983.21. Zechner R. Rapid and simple isolation procedure for lipoprotein lipase from human milk. BBA 1044: 20-25, 1990.22. Kuipers F, Jong MC, Lin Y, Eck M, Havinga R, Bloks V, Verkade HJ, Hofker MH, Moshage H, Berkel TJ, Vonk RJ and HavekesLM. Impaired secretion of VLDL-triglycerides by apoE-deficient mouse hepatocytes. J Clin Invest 100: 2915-2922, 1997.23. Klaunig JE, Goldblatt PJ, Hinton DE, Lipsky MM, Chacko J and Trump BF. Mouse liver cell culture. I. Hepatocyte isolation.In Vitro 17: 913-925, 1981.24. Harris WS, Hustvedt BE, Hagen E, Green MH, Lu G and Drevon CA. N-3 fatty acids and chylomicron metabolism in the rat.J Lipid Res 38: 503-515, 1997.25. Fraser R. Size and lipid composition of chylomicrons of different Svedberg units of flotation. J Lipid Res 11: 60-65, 1970.26. Bligh EG and Dyer WJ. A rapid method for total lipid extraction and purification. Can J Biochem Physiol 37: 911-917, 1959.27. Lowry OH, Rosebrough NJ, Farr AL and Randal RJ. Protein measurement with the Folin phenol reagent. J Biol Chem 193:265-275, 1951.28. Vance DE, Weinstein DB and Steinberg D. Isolation and analysis of lipoproteins secreted by rat liver hepatocytes. BiochimBiophys Acta 792: 39-47, 1984.29. Curtin A, Deegan P, Owens D, Collins P, Johnson A and Tomkin GH. Elevated triglyceride-rich lipoproteins in diabetes. Astudy of apolipoprotein B-48. Acta Diabetol 33: 205-210, 1996.30. Kok T, Wolters H, Bloks VW, Havinga R, Jansen PL, Staels B and Kuipers F. Induction of hepatic ABC transporter expres-sion is part of the PPARalpha-mediated fasting response in the mouse. Gastroenterology 124: 160-171, 2003.31. Plosch T, Kok T, Bloks VW, Smit MJ, Havinga R, Chimini G, Groen AK and Kuipers F. Increased hepatobiliary and fecal cho-lesterol excretion upon activation of the liver X receptor is independent of ABCA1. J Biol Chem 277: 33870-33877, 2002.

72

Chapter 3


32. Grefhorst A, Elzinga BM, Voshol PJ, Plosch T, Kok T, Bloks VW, van der Sluijs FH, Havekes LM, Romijn JA, Verkade HJ andKuipers F. Stimulation of lipogenesis by pharmacological activation of the liver X receptor leads to production of large, triglyc-eride-rich very low density lipoprotein particles. J Biol Chem 277: 34182-34190, 2002.33. Holman RT. The ratio of trienoic:tetraenoic acids in tissue lipids as a measure of EFA requirement. J Nutr 70: 405-410, 1960.34. Yamanaka WK, Clemans GW and Hutchinson ML. EFA deficiency in humans. Prog Lipid Res 19: 187-215, 1981.35. Bennett Clark S, Ekkers TE, Singh A, Balint JA, Holt PR and Rodgers JB. Fat absorption in EFA deficiency: a model exper-imental approach to studies of the mechanism of fat malabsorption of unknown etiology. JLR 14: 581-588, 1973.36. Levy E, Garofalo C, Thibault L, Dionne S, Daoust L, Lepage G and Roy CC. Intraluminal and intracellular phases of fatabsorption are impaired in essential fatty acid deficiency. Am J Physiol 262: G319-G326, 1992.37. Rensen PC, Herijgers N, Netscher MH, Meskers SC, Van Eck M and Van Berkel TJ. Particle size determines the specificityof apolipoprotein E-containing triglyceride-rich emulsions for the LDL receptor versus hepatic remnant receptor in vivo. J LipidRes 38: 1070-1084, 1997.38. Levy E, Garofalo C, Rouleau T, Gavino V and Bendayan M. Impact of essential fatty acid deficiency on hepatic sterol metab-olism in rats. Hepatology 23: 848-857, 1996.39. Abdel-Fattah G, Fernandez ML and McNamara DJ. Regulation of very low density lipoprotein apo B metabolism by dietaryfat saturation and chain length in the guinea pig. Lipids 33: 23-31, 1998.40. Abdel-Fattah G, Fernandez ML and McNamara DJ. Regulation of guinea pig very low density lipoprotein secretion rates bydietary fat saturation. J Lipid Res 36: 1188-1198, 1995.41. Mortimer BC, Holthouse DJ, Martins IJ, Stick RV and Redgrave TG. Effects of triacylglycerol-saturated acyl chains on theclearance of chylomicron-like emulsions from the plasma of the rat. Biochim Biophys Acta 1211: 171-180, 1994.42. Redgrave TG, Rakic V, Mortimer BC and Mamo JC. Effects of sphingomyelin and phosphatidylcholine acyl chains on theclearance of TG-rich lipoproteins from plasma. Studies with lipid emulsions in rats. Biochim Biophys Acta 1126: 65-72, 1992.43. Miller KW and Small DM. Surface-to-core and interparticle equilibrium distributions of triglyceride-rich lipoprotein lipids. JBiol Chem 258: 13772-13784, 1983.44. Hamilton JA, Miller KW and Small DM. Solubilization of triolein and cholesteryl oleate in egg phosphatidylcholine vesicles.J Biol Chem 258: 12821-12826, 1983.45. Hamilton JA, Vural JM, Carpentier YA and Deckelbaum RJ. Incorporation of medium chain triacylglycerols into phospho-lipid bilayers: effect of long-chain triacylglycerols, cholesterol, and cholesteryl esters. J Lipid Res 37: 773-782, 1996.46. Rao R, Albers JJ, Wolfbauer G and Pownall HJ. Molecular and macromolecular specificity of human plasma phospholipidtransfer protein. Biochemistry 36: 3645-3653, 1997.47. Quarfordt SH and Goodman DS. Heterogeneity in the rate of plasma clearance of chylomicrons of different size. BiochimBiophys Acta 116: 382-385, 1966.48. Chajek-Shaul T, Eisenberg S, Oschry Y and Olivecrona T. Metabolic heterogeneity of post-lipolysis rat mesenteric lymphsmall chylomicrons produced in vitro. J Lipid Res 24: 831-840, 1983.49. Martins IJ, Mortimer BC, Miller J and Redgrave TG. Effects of particle size and number on the plasma clearance of chy-lomicrons and remnants. J Lipid Res 37: 2696-2705, 1996.50. Yao ZM and Vance DE. The active synthesis of phosphatidylcholine is required for very low density lipoprotein secretionfrom rat hepatocytes. J Biol Chem 263: 2998-3004, 1988.51. Verkade HJ, Fast DG, Rusinol AE, Scraba DG and Vance DE. Impaired biosynthesis of phosphatidylcholine causes adecrease in the number of very low density lipoprotein particles in the Golgi but not in the endoplasmic reticulum of rat liver. JBiol Chem 268: 24990-24996, 1993.52. Vu-Dac N, Gervois P, Jakel H, Nowak M, Bauge E, Dehondt H, Staels B, Pennacchio LA, Rubin EM, Fruchart-Najib J andFruchart JC. Apolipoprotein A5, a crucial determinant of plasma triglyceride levels, is highly responsive to peroxisome prolifer-ator-activated receptor alpha activators. J Biol Chem 278: 17982-17985, 2003.53. Levy E, Lepage G, Bendayan M, Ronco N, Thibault L, Galeano N, Smith L and Roy CC. Relationship of decreased hepaticlipase activity and lipoprotein abnormalities to EFA deficiency in cystic fibrosis patients. J Lipid Res 30: 1197-1209, 1989.54. Lindblad A, Glaumann H and Strandvik B. Natural history of liver disease in cystic fibrosis. Hepatology 30: 1151-1158, 1999.

73



74


Lymphatic chylomicron sizeis inversely related to biliary

phospholipid secretion in mice 4Conditionally accepted for publication in Am J Physiol 2005

A. WernerH. Havinga

F. PertonF. Kuipers

H.J. Verkade


ABSTRACT

Background: Biliary phospholipids (PL) stimulate dietary fat absorption by facilitating

intraluminal lipid solubilization and by providing surface components for chylomicron

assembly. Impaired hepatic PL availability induces secretion of large VLDL, but it is

unclear whether chylomicron size depends on biliary PL availability. Biliary PL

secretion is absent in Mdr2 -/- mice, whereas it is strongly increased in essential fatty

acid (EFA) deficient mice. We investigated lymphatic chylomicron size and compo-

sition in mice with absent (Mdr2 -/-) or enhanced (EFA deficiency) biliary PL secretion

and in their respective controls, under basal conditions and during intraduodenal

lipid administration.

Methods: EFA deficiency was induced by feeding mice a high-fat EFA-deficient diet

for eight weeks. Lymph was collected by mesenteric lymph duct cannulation, with or

without intraduodenal lipid administration. Lymph was collected in 30-minute

fractions for 4 hours, and lymphatic lipoprotein size was determined by dynamic light

scattering techniques. Lymph lipoprotein subfractions were isolated by ultra-

centrifugation and lipid composition was measured.

Results: Lymphatic lipoproteins were significantly larger in Mdr2 -/- mice than in

Mdr2 +/+ controls, both without (+50%) and with (+25%) intraduodenal lipid

administration. In contrast, EFA-deficient mice secreted significantly smaller lipo-

proteins into lymph than EFA-sufficient controls (164±28 nm vs. 234±49 nm;

p<0.001). Chylomicron size increased during fat absorption in both EFA-deficient

and EFA-sufficient mice, but the difference between the groups persisted.

Conclusions: Present results strongly suggest that the availability of biliary PL is a

major determinant of the size of intestinally produced lipoproteins, both under basal

conditions and during lipid absorption. Altered chylomicron size may have

physiological consequences for postprandial chylomicron processing.

76

Chapter 4


INTRODUCTION

Biliary phospholipids (PL) have a well-documented function in transport of dietary

lipids from the intestinal lumen into lymph. Apart from their role in intraluminal lipid

solubilization, biliary phospholipids have been implicated as crucial components for

adequate intestinal chylomicron (CM) assembly and secretion into lymph(23). Biliary

phosphatidylcholines (PC), which compose ~95% of biliary phospholipids, provide

the main source for chylomicron surface coating(13; 17; 20), and the supply of bile PC to

the intestine increases the synthesis of apoB48, the apolipoprotein needed for

chylomicron formation(6; 14). Additionally, the high essential fatty acid (EFA) content of

biliary PC may be required for maintenance of normal intestinal mucosal membrane

composition and function(7; 22). Reduced availability of PC for hepatic VLDL assembly

in rats has been associated with decreased VLDL secretion and with assembly of

relatively large VLDL particles(24). It is well-established that fat absorption and

intestinal lipoprotein secretion are strongly impaired in situations of disturbed bile

formation, such as cholestasis. Studies in rats with interrupted enterohepatic

circulation by means of permanent bile diversion(23) demonstrated a substantial lipid

accumulation in intestinal mucosal cells. Administration of bile salts to these bile-

diverted rats partially restored lymphatic lipid transfer, but only when both bile salts

and biliary phospholipids were supplemented, lymphatic lipid transport was fully

reinstated. Yet, a direct relationship between biliary phospholipid availability and

intestinal lipoprotein size has not been established.

We recently characterized fat absorption in two mouse models with altered biliary

phospholipid secretion(30). EFA deficiency increases biliary phospholipid secretion by

~80%, in conjunction with a decrease by 30-40% in dietary lipid absorption.

In Mdr2 -/- mice, in which biliary phospholipid secretion is absent(21), plasma

appearance of enterally administered lipid is delayed and lipid accumulates in

enterocytes. Quantitatively, however, lipid absorption is unaffected in these mice(26).

In the present study, we investigated whether and to what extent quantitative or

qualitative alterations in biliary phospholipid secretion affect chylomicron secretion

into lymph. Chylomicron size and composition were measured after mesenteric

lymph duct cannulation, using the aforementioned mouse models with altered intra-

luminal biliary phospholipid availability, i.e., Mdr2 -/- mice (no biliary PL secretion),

EFA-deficient mice (increased biliary PL secretion) and corresponding control mice.

Our results show that the absence of biliary phospholipid secretion in mice is

accompanied by production of intestinal lipoproteins of increased size and

decreased phospholipid content, whereas EFA deficiency, associated with increased

biliary phospholipid secretion, has the opposite effect

77

Lymphatic chylomicron size is inversely related to biliary phospholipid secretion in mice



AnimalsMice homozygous for disruption of the multidrug resistance gene-2 (Mdr2 -/-) and

wildtype (Mdr2 +/+)mice with a free virus breed (FVB) background were obtained from

the breeding colony at the Central Animal Facility, Academic Medical Center,

Amsterdam, the Netherlands(8). For dietary induction of EFA deficiency in mice, male

wildtype FVB mice were obtained from Harlan (Horst, the Netherlands). All mice were

8 weeks old, weighed 25 to 35 gram and were housed in a light-controlled (lights on

6 AM - 6 PM) and temperature-controlled (21°C) facility. Mice were allowed tap water

and chow ad libitum. The experimental protocols were approved by the Ethics

Committee for Animal Experiments, University of Groningen, the Netherlands.

Experimental dietsThe standard laboratory low-fat chow (RMH-B, Arie Blok BV, Woerden, the

Netherlands) contained 14 energy% fat. The high-fat EFA-deficient (EFAD) diet

contained 34 energy% fat, and had the following fatty acid composition: 41.4 mol%

palmitic acid (C16:0), 47.9 mol% stearic acid (C18:0), 7.7 mol% oleic acid

(C18:1n-9) and 3 mol% linoleic acid (C18:2n-6). An isocaloric EFA-sufficient (EFAS)

diet was used as control diet, containing 37 energy% fat with 32.1 mol% C16:0, 5.5

mol% C18:0, 32.2 mol% C18:1n-9 and 30.2 mol% C18:2n-6 (diet numbers 4141.08

(EFAD) and 4141.07 (EFAS), Arie Blok BV, Woerden, the Netherlands).

Experimental proceduresInduction of EFA deficiency in mice

All mice were fed standard laboratory chow from weaning. For induction of EFA

deficiency, wildtype FVB mice were fed the EFA-deficient (EFAD) diet for eight weeks.

A control group of FVB mice was fed the isocaloric EFA-sufficient (EFAS) diet for

eight weeks. This method for induction of EFA deficiency was previously applied in

mice and characterized by our group(15; 29; 30).

Mesenteric lymph duct cannulation in mice

Cannulation of the mesenteric lymph duct was performed according to procedures

described by Wang et al.(27; 28) Non-fasted mice were anesthetized with halothane/NO2

and dorsally arched over a cotton cylinder for optimal visualization of the mesenteric

lymph duct. After extra-abdominal displacement of the intestine, the common mesen-

teric lymph duct was exposed by removal of surrounding tissues and membranes

using a blunt mini-kocher. A 0.305 x 0.635 mm (id x od) silicone catheter was

78

Chapter 4


introduced through the abdominal wall and positioned parallel to the lymph duct.

Subsequently, the catheter was carefully inserted into a small incision in the

lymphatic duct and fixated by drops of tissue glue at the junction of lymph duct and

catheter. A subgroup of mice subsequently received an intraduodenal 200 µl bolus

of a parenteral lipid emulsion (Intralipid® 20%, Fresenius Kabi, 's Hertogenbosch, the

Netherlands), mixed with glucose 3.3% and NaCl 0.3% (50:50), after collection of a

30-minute baseline lymph sample. After repositioning the intestine, the abdominal

incision was closed with 8-10 sutures and mice were placed in restrainer cages in a

37°C incubator. Analgesia was maintained with intraperitoneal injection of

buprenorfine (Temgesic®) 0.1 mg/kg. Lymph was collected by gravity into EDTA-

containing microtubes, in 30-minute fractions for 4 hours after cannulation of the

mesenteric lymph duct.

Determination of lymphatic lipoprotein size and composition

Lymphatic lipoprotein size and volume distribution profiles were analyzed within 6

hours after lymph collection by dynamic light scattering techniques, using a Nicomp

model 370 submicron particle analyzer (Nicomp Particle sizing Systems, Santa

Barbara, CA, USA). Particle diameters were calculated from the volume distribution

patterns provided by the analyzer. The lymphatic chylomicron (CM) fraction

(d<1.006) was isolated after complementing collected lymph with a 1.006 g/ml NaCl

solution containing 0.02% NaN3 to a final volume of 1 ml, and subsequent centrifu-

gation at 40000 rpm at 4°C in an Optima TM LX table top centrifuge (Beckman

Instruments, Inc., Palo Alto, CA, USA) for 15 minutes. The top layer containing the

chylomicron fraction was isolated by tube slicing and the volume was recorded by

weight. The remaining lymph solution was again complemented with 1.006 g/ml

NaCl to a final volume of 1 ml and centrifuged at 120000 rpm for 1 hour and 40

minutes for isolation of the very low density lipoprotein (VLDL) fraction.

Analytical techniquesPlasma lipids were measured using commercially available assay kits from Roche

(Mannheim, Germany) for triglycerides and total cholesterol, and from WAKO

chemicals GmbH (Neuss, Germany) for phospholipids.

Calculations and statisticsAll results are presented as means ± S.D. for the number of animals indicated. Data

were statistically analyzed using Student's t-test or ANOVA test with post-hoc

Bonferroni correction. Level of significance was set at p<0.05. Analyses were

performed using SPSS for Windows software (SPSS, Chicago, IL).

79



RESULTS

Lymph flow was highly variable between and within individual mice, ranging between

1.0 and 15.8 µl/min per 100 g body weight, but no significant differences in lymph

flow were observed between experimental groups and their respective controls (data

not shown).

In Mdr2-Pgp-deficient (Mdr2 -/-) mice, biliary phospholipid secretion into the intestine

is virtually absent (<0.5 nmol/min per 100 g)(21). Voshol et al. described that post-

prandial plasma appearance of chylomicrons is impaired in Mdr2 -/- mice(26). Figure 1

shows that non-fed Mdr2 -/- mice secreted chylomicrons of significantly greater size

(+51%) into lymph than Mdr2 +/+ controls during the first 2 hours of lymph collection

(131±23 vs. 87±26 for Mdr2 -/- and Mdr2 +/+ mice, respectively, p<0.001).

Concentrations of triglycerides (TG), phospholipids (PL) and cholesterol were

significantly lower in lymph of Mdr2 -/- mice than of controls (Figure 2a). The decrease

in PL and cholesterol content (-57% and -93%, respectively) was considerably more

pronounced than that of triglyceride (-26%). In the isolated lymphatic chylomicron

(Figure 2b) and very low density lipoprotein (VLDL) fractions (Figure 2c), similar

differences in lipid content were detected: in the first hour of lymph collection,

concentrations of phospholipid, triglyceride and cholesterol were decreased in

lipoproteins of Mdr2 -/- mice compared to controls, but the relative triglyceride

concentration slightly increased (CM fraction: 86±3% vs. 78±7%; VLDL fraction:

85±1% vs. 80±1% for Mdr2 -/- and Mdr2 +/+ mice, respectively, p<0.005). The core-to-

surface ratio of lymphatic lipoproteins (i.e., [TG]/[PL], Figure 2d) was increased in

total lymph as well as in the isolated CM and VLDL fractions of Mdr2-/- mice during

the first hour of lymph collection, indicating secretion of larger lipoproteins in Mdr2 -

/- mice than in controls. In the later fractions, a similar trend was observed, but the

increased core-surface ratio only reached significance in the VLDL fraction.

80

Chapter 4

lymph lipoprotein size no lipid bolus

time (min)

part

icle

siz

e (n

m)

0

50

100

150

200

250

0

* * * *

30 60 90 120 150 180

Mdr2 +/+

Mdr2 -/-

lymph lipoprotein size no lipid bolus

time (min)

part

icle

siz

e (n

m)

0

50

100

150

200

250

00

50

100

150

200

250

0

* * * *

30 60 90 120 150 180

Mdr2 +/+

Mdr2 -/-Mdr2 +/+

Mdr2 -/-

Figure 1: Lymphatic lipoprotein size (nm), measured bydynamic light scattering, in lymph of non-fasted Mdr2-Pgp-deficient (Mdr2(-/-)) mice (grey circles) and Mdr2(+/+)

mice (black circles). Lipoprotein size was measured in 30-minute fractions of collected mesenteric lymph, andlymph was collected for 3 hours. Data represent means ±SD of 6-10 mice per group. *p<0.05 for differencesbetween Mdr2(-/-) and Mdr2(+/+) mice.

1


We also assessed the effects of absence of biliary phospholipids on chylomicron

formation during the active phase of lipid absorption. After collection of a 30-minute

baseline sample, a 200 µl lipid bolus was administered intraduodenally to Mdr2 -/- and

Mdr2 +/+ mice. Similar to the non-fed state, the average diameter of lymph lipo-

proteins during lipid absorption was significantly larger in Mdr2 -/- mice than in

Mdr2 +/+ controls (183±41 nm vs. 152±44 nm; p<0.001). The increase in lipoprotein

size after lipid administration occurred more rapidly in Mdr2 -/- mice (Figure 3).

81


Mdr2+/+

Mdr2-/-

lipid composition lymph 0-60 min

0

2

4

6

8

10

12

PL TG chol

relative lipid composition lymph 0-60 min

0

20

40

60

80

100

PL TG chol

###

*

*##

##

lipid composition lymph >60 min

0

2

4

6

8

10

12

PL TG chol

relative lipid composition lymph >60 min

0

20

40

60

80

100

PL TG chol

Mdr2+/+

Mdr2-/-

## #

#

mM

mM

% o

f tot

al li

pid

% o

f tot

al li

pidMdr2+/+

Mdr2-/-


0

2

4

6

8

10

12

PL TG chol


0

20

40

60

80

100

PL TG chol

###

*

*##

##Mdr2+/+

Mdr2-/-


0

2

4

6

8

10

12

0

2

4

6

8

10

12

PL TG chol


0

20

40

60

80

100

0

20

40

60

80

100

PL TG chol

###

*

*##

##


0

2

4

6

8

10

12

PL TG chol


0

20

40

60

80

100

PL TG chol

Mdr2+/+

Mdr2-/-

## #

#


0

2

4

6

8

10

12

0

2

4

6

8

10

12

PL TG chol


0

20

40

60

80

100

PL TG chol


0

20

40

60

80

100

PL TG chol0

20

40

60

80

100

0

20

40

60

80

100

PL TG chol

Mdr2+/+Mdr2+/+

Mdr2-/-Mdr2-/-

## #

#

mM

mM

% o

f tot

al li

pid

% o

f tot

al li

pid

mM

mM

% o

f tot

al li

pid

% o

f tot

al li

pid

% o

f tot

al li

pid

% o

f tot

al li

pid

Figure 2a: Absolute (mM) and relative (%) concentrations of phospholipid (PL), triglyceride (TG) And cholesterol (chol) in lymphof Mdr2(-/-) mice (grey bars) and Mdr2(+/+) mice (black bars) in the first hour of lymph collection and after the first hour. Data rep-resent means±SD of 6-10 mice per group. *p<0.05, #p<0.005, ##p<0.001 for differences between Mdr2(-/-) and Mdr2(+/+) mice.

lipid composition

lymph CM 0-60 min relative lipid composition

lymph CM 0-60min

0

20

40

60

80

100

PL TG chol

Mdr2+/+

Mdr2-/-

PL TG chol

##

##

#

#

0

2

4

6

8

10

12

PL TG chol

lipid composition lymph CM >60 min

relative lipid composition lymph CM >60min

0

20

40

60

80

100

PL TG chol

Mdr2+/+

Mdr2-/-

**0

2

4

6

8

10

12

** **

mM

mM

% o

f tot

al li

pid

% o

f tot

al li

pid

lipid composition lymph CM 0-60 min

relative lipid composition lymph CM 0-60min

0

20

40

60

80

100

PL TG chol

Mdr2+/+

Mdr2-/-

PL TG chol

##

##

#

#

0

2

4

6

8

10

12

lipid composition lymph CM 0-60 min

relative lipid composition lymph CM 0-60min

0

20

40

60

80

100

0

20

40

60

80

100

PL TG chol

Mdr2+/+

Mdr2-/-

Mdr2+/+Mdr2+/+

Mdr2-/-Mdr2-/-

PL TG chol

##

##

#

#

0

2

4

6

8

10

12

PL TG chol



0

20

40

60

80

100

PL TG chol

Mdr2+/+

Mdr2-/-

**0

2

4

6

8

10

12

** **PL TG chol



0

20

40

60

80

100

PL TG chol

Mdr2+/+

Mdr2-/-

**0

2

4

6

8

10

12

** **

mM

mM

% o

f tot

al li

pid

% o

f tot

al li

pid

% o

f tot

al li

pid

% o

f tot

al li

pid

Figure 2b: Absolute (mM) and relative (%) concentrations of phospholipid (PL), triglyceride (TG) and cholesterol (chol) in theisolated lymphatic chylomicron (CM) fraction of Mdr2(-/-) mice (grey bars) and Mdr2(+/+) controls (black bars) in the first hour oflymph collection and after the first hour. Data represent means ± SD of 6-10 mice per group. *p<0.01, #p<0.005, ##p<0.001for differences between Mdr2(-/-) and Mdr2(+/+) mice.

lipid composition

lymph VLDL 0-60 min

0

20

40

60

80

100

PL TG chol

relative lipid composition lymph VLDL 0-60min

Mdr2+/+

Mdr2-/-

##

##

PL TG chol##

0

2

4

6

8

10

12

## ##

lipid composition lymph VLDL >60 min

relative lipidcompositionlymph VLDL >60min

0

20

40

60

80

100

PL TG chol

Mdr2+/+

Mdr2-/-

##

##

PL TG chol##

0

2

4

6

8

10

12

mM

mM

% o

f tot

al li

pid

% o

f tot

al li

pid

lipid composition lymph VLDL 0-60 min

0

20

40

60

80

100

PL TG chol


Mdr2+/+

Mdr2-/-

##

##

PL TG chol##

0

2

4

6

8

10

12

## ##

lipid composition lymph VLDL 0-60 min

0

20

40

60

80

100

0

20

40

60

80

100

PL TG chol


Mdr2+/+

Mdr2-/-

##

##

PL TG chol##

0

2

4

6

8

10

12

## ##

lipid composition lymph VLDL >60 min

relative lipidcompositionlymph VLDL >60min

0

20

40

60

80

100

PL TG chol

Mdr2+/+

Mdr2-/-

##

##

0

20

40

60

80

100

0

20

40

60

80

100

PL TG chol

Mdr2+/+

Mdr2-/-

##

##

PL TG chol##

0

2

4

6

8

10

12

PL TG chol##

0

2

4

6

8

10

12

mM

mM

% o

f tot

al li

pid

% o

f tot

al li

pid

% o

f tot

al li

pid

% o

f tot

al li

pid

Figure 2c: Absolute (mM) and relative (%) concentrations of phospholipid (PL), triglyceride (TG) and cholesterol (chol) in theisolated lymphatic very low density lipoprotein (VLDL) fraction of Mdr2(-/-) mice (grey bars) and Mdr2(+/+) controls (black bars) inthe first hour of lymph collection and after the first hour. Data represent means ± SD of 6-10 mice per group. **p<0.01,#p<0.005, ##p<0.001 for differences between Mdr2(-/-) and Mdr2(+/+) mice.

2a

2b

2c


Figure 4 shows the absolute and relative lipid concentrations in lymph, at baseline

and 1 and 2 hours after intraduodenal lipid administration. Absolute phospholipid,

triglyceride and cholesterol concentrations (Figure 4a) were significantly lower in

Mdr2 -/- mice than in controls, and did not increase over time. Relatively, however,

lymph of Mdr2 -/- mice contained more triglyceride and less cholesterol and phos-

pholipid than that of Mdr2 +/+ controls (Figure 4b), suggesting the presence of larger

particles. Indeed, core-to-surface ratio's were increased in Mdr2 -/- mice compared to

controls (4.1±1.4 vs. 2.7±0.9, p<0.001; Figure 4c). Thus, a quantitative decrease in

biliary phospholipid secretion in Mdr2 -/- mice was associated with secretion of larger

lymphatic lipoproteins.

To assess the effects of increased biliary phospholipid secretion, we measured lym-

phatic lipoprotein size and composition in mice after dietary induction of EFA defi-

ciency. Previously, we characterized the effects of EFA deficiency on fat absorption

and biliary phospholipid secretion(30). EFA-deficient (EFAD) mice have a dietary lipid

malabsorption ranging between 60-70% of the amount ingested, combined with a

70% increased bile flow and an 83% increased biliary phospholipid excretion. Biliary

82

Chapter 4

core-surface ratio (0-60 min)

[TG

] / [P

L]

0

2

4

6

8

10

total lymph CM VLDL

**

#

core-surface ratio (>60 min)

0

2

4

6

8

10

total lymph CM VLDL

[TG

] / [P

L]

#

Mdr2+/+

Mdr2-/-Mdr2+/+

Mdr2-/-


[TG

] / [P

L]

0

2

4

6

8

10

total lymph CM VLDL

**

#


[TG

] / [P

L]

0

2

4

6

8

10

total lymph CM VLDL

**

#

[TG

] / [P

L]

0

2

4

6

8

10

0

2

4

6

8

10

total lymph CM VLDL

**

#


0

2

4

6

8

10

total lymph CM VLDL

[TG

] / [P

L]

#


0

2

4

6

8

10

0

2

4

6

8

10

total lymph CM VLDL

[TG

] / [P

L]

#

total lymph CM VLDL

[TG

] / [P

L]

#

Mdr2+/+

Mdr2-/-Mdr2+/+

Mdr2-/-Mdr2+/+

Mdr2-/-Mdr2+/+

Mdr2-/-

Figure 2d: Core-to-surface ratio, estimated by the ratio of triglyceride (TG) and phospholipid (PL) concentration (mM) in totallymph and in the isolated lymphatic chylomicron (CM) and very low density lipoprotein (VLDL) fractions of Mdr2(-/-) mice (greybars) and Mdr2(+/+) mice (black bars) in the first hour of lymph collection and after the first hour. Data represent means ± SD of6-10 mice per group. *p<0.05, #p<0.005 for differences between Mdr2(-/-) and Mdr2(+/+) mice.

lymph lipoprotein size after lipid bolus

time (min)

0 30 60 90 120 150 180

part

icle

siz

e (n

m)

* **

* * *

*

*

-30

lipid bolus 50

100

150

200

250

0

*Mdr2 +/+

Mdr2 -/-

lymph lipoprotein size after lipid bolus

time (min)

0 30 60 90 120 150 180

part

icle

siz

e (n

m)

* **

* * *

**

*

-30

lipid bolus 50

100

150

200

250

0

50

100

150

200

250

0

*Mdr2 +/+

Mdr2 -/-Mdr2 +/+

Mdr2 -/-

Figure 3: Lymphatic lipoprotein size (nm), determined bydynamic light scattering, in lymph of Mdr2(-/-) mice (greycircles) and Mdr2(+/+) controls (black circles) during activelipid absorption. Mesenteric lymph was collected andlipoprotein size was measured in 30-minute fractions,prior to, and for 3 hours after intraduodenal lipid bolusadministration. Data represent means ± SD of 5-9 miceper group. *p<0.05 for differences between Mdr2(-/-) andMdr2(+/+) mice

2d

3


83


lipid composition lymph

baseline

0

5

10

15

PL TG chol

mM

Mdr2+/+

Mdr2-/-

##

lipid composition lymph 1 hour after lipid bolus

0

5

10

15

PL TG chol

mM

Mdr2+/+

Mdr2-/-

*

#

#

lipid composition lymph 2 hours after lipid bolus

0

5

10

15

PL TG chol

mM

Mdr2+/+

Mdr2-/-

*

##

lipid composition lymph baseline

0

5

10

15

PL TG chol

mM

Mdr2+/+

Mdr2-/-

##


0

5

10

15

PL TG chol

mM

Mdr2+/+

Mdr2-/-

*

#

#


0

5

10

15

PL TG chol

mM

Mdr2+/+

Mdr2-/-

*

##


0

5

10

15

0

5

10

15

PL TG chol

mM

Mdr2+/+

Mdr2-/-Mdr2+/+Mdr2+/+

Mdr2-/-Mdr2-/-

##


0

5

10

15

0

5

10

15

PL TG chol

mM

Mdr2+/+


Mdr2-/-Mdr2-/-

*

#

#


0

5

10

15

0

5

10

15

PL TG chol

mM

Mdr2+/+


Mdr2-/-Mdr2-/-

*

##

Figure 4a: Concentrations of PL, TG and cholesterol (chol) in mM, in lymph of Mdr2(-/-) mice (grey bars) and Mdr2(+/+) controls(black bars) at baseline, and at 1 and 2 hours after intraduodenal lipid administration. Data represent means ± SD of 5-9 miceper group. *p<0.05, #p<0.005 for differences between Mdr2(-/-) and Mdr2(+/+) mice.

Mdr2+/+ Mdr2-/-

core-surface ratio after lipid bolus

0

2

4

6

8

10

[TG

] / [P

L]

*

Mdr2+/+ Mdr2-/-


0

2

4

6

8

10

0

2

4

6

8

10

[TG

] / [P

L]

* Figure 4c: Core-to-surface ratio, estimated by the ratio of triglyceride (TG) and phos-pholipid (PL) concentrations (mM) in lymph of Mdr2(-/-) mice (grey bars) and Mdr2(+/+)

controls (black bars). Data represent means ± SD of 5-9 mice per group. *p<0.001for differences between Mdr2(-/-) and Mdr2(+/+) mice.

Figure 4b: Relative concentrations of PL, TG and cholesterol,expressed as % of total lipid, in lymph of Mdr2(-/-) mice (grey bars)and Mdr2(+/+) controls (black bars) at baseline, and at one and two hours after intraduodenal lipid administration.Data represent means ± SD of 5-9 mice per group. *p<0.05, #p<0.005 for differences between Mdr2(-/-) and Mdr2(+/+) mice.

0

20

40

60

80

100

relative lipid composition lymph 1 hour after lipid bolus

PL TG chol

% o

f tot

al li

pid

Mdr2+/+

Mdr2-/-

#

#

0

20

40

60

80

100

relative lipid composition lymph 2 hours after lipid bolus

PL TG chol

% o

f tot

al li

pid

Mdr2+/+

Mdr2-/-

#

#

0

20

40

60

80

100

relative lipid composition lymph baseline

PL TG chol

% o

f tot

al li

pid

Mdr2+/+

Mdr2-/-

#

#

*

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid

Mdr2+/+

Mdr2-/-

#

#

0

20

40

60

80

100

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid

Mdr2+/+


Mdr2-/-Mdr2-/-

#

#

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid

Mdr2+/+

Mdr2-/-

#

#

0

20

40

60

80

100

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid

Mdr2+/+


Mdr2-/-Mdr2-/-

#

#

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid

Mdr2+/+

Mdr2-/-

#

#

*

0

20

40

60

80

100

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid

Mdr2+/+


Mdr2-/-Mdr2-/-

#

#

*

4a

4b

4c


phospholipid acyl chains of EFA-deficient mice contained significantly less essential

fatty acids and their long-chain metabolites (i.e., C18:2n-6, C18:3n-6 and C20:4n-6),

and more non-essential fatty acids (C16:1n-7, C18:1n-7, C18:1n-9) than EFA-

sufficient controls (Figure 5a). Figure 5b shows that at baseline, before enteral lipid

administration, lymphatic chylomicrons of EFA-deficient mice were significantly

smaller (-32%) than those of EFA-sufficient controls. This decreased particle size in

EFA-deficient mice persisted during the active phase of fat absorption, i.e., for two

hours after lipid bolus administration (164±28 nm vs. 234±49 nm; p<0.001). Lymph

of EFA-deficient mice contained, both absolutely and relatively, more phospholipid

and less triglyceride and cholesterol than lymph of EFA-sufficient controls (Figure 6a

and 6b). This was associated with a lower core-to-surface ratio in EFA-deficient mice

(3.4±0.9 vs. 5.7±1.6 for EFA-deficient and EFA-sufficient mice, respectively;

p<0.001, Figure 6c), indicating secretion of smaller lymphatic lipoproteins.

84

Chapter 4

0

10

20

30

40

50

16:0 16:1n-7 18:0 18:3n-6 18:2n-6 18:1n-9 18:1n-7 20:4n-6

mo

l% o

f to

tal f

atty

aci

ds

Biliary fatty acids in EFAD and EFAS mice

**

*

*

*

**

*

EFAD EFAS

0

10

20

30

40

50

0

10

20

30

40

50

16:0 16:1n-7 18:0 18:3n-6 18:2n-6 18:1n-9 18:1n-7 20:4n-6

mo

l% o

f to

tal f

atty

aci

ds

Biliary fatty acids in EFAD and EFAS mice

**

*

*

*

**

*

EFAD EFAS EFAD EFAD EFAS EFAS

Figure 5a: Relative fatty acid composition of biliaryphospholipids from EFA-deficient (EFAD, open bars)and EFA-sufficient (EFAS, black bars) mice.Individual concentrations of palmitic acid (C16:0),palmitoleic acid (C16:1n-7), stearic acid (C18:0),dihomo-gamma-linolenic acid (C18:3n-6), linoleicacid (C18:2n-6), oleic acid (C18:1n-9) and arachi-donic acid (C20:4n-6) are expressed as molar per-centages of total fatty acids. Data represent means± SD of 7 mice per group. *p<0.001 for differencesbetween EFAD and EFAS mice.

* * * * *lymph lipoprotein size after lipid bolus

time (min)

part

icle

siz

e (n

m)

0

50

100

150

200

250

300

0 30 60 90 120 150 180 210

EFAS

EFAD

*

lipid bolus

-30

* * * * *lymph lipoprotein size after lipid bolus

time (min)

part

icle

siz

e (n

m)

0

50

100

150

200

250

300

0

50

100

150

200

250

300

0 30 60 90 120 150 180 210

EFAS

EFAD

EFAS EFAS

EFAD EFAD

*

lipid bolus

-30

Figure 5b: Lymphatic lipoprotein size (nm), deter-mined by dynamic light scattering, in lymph of EFA-deficient (EFAD) mice (open circles) and EFA-suffi-cient (EFAS) controls (black circles), during activelipid absorption. Mesenteric lymph was collectedand lipoprotein size was measured in 30-minutefractions, prior to, and for 3.5 hours after intraduo-denal lipid bolus administration. Data representmeans ± SD of 6-9 mice per group. *p<0.05 for dif-ferences between EFAD and EFAS mice.

5a

5b


85


0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid

EFASEFAD

#

#

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid

EFASEFAD

0

20

40

60

80

100

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid

EFASEFADEFASEFASEFADEFAD

#

#

*

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid

EFASEFAD

*

*

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid

EFASEFAD

0

20

40

60

80

100

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid


*

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid

EFASEFAD

#

#

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid

EFASEFAD

0

20

40

60

80

100

0

20

40

60

80

100


PL TG chol

% o

f tot

al li

pid


#

#

EFAS EFAD


0

2

4

6

8

10

[TG

] / [P

L]

*

EFAS EFAD


0

2

4

6

8

10

0

2

4

6

8

10

[TG

] / [P

L]

*Figure 6c: Core-to-surface ratio, estimated by the ratio of triglyceride (TG) concen-tration (mM) and phospholipid (PL) concentration (mM) in lymph of EFA-deficient(EFAD, white bars) mice and EFA-sufficient (EFAS, black bars) controls. Data repre-sent means ± SD of 6-9 mice per group. *p<0.001 for differences between EFADand EFAS mice.

mM

0

5

10

15

20

25 EFASEFAD


PL TG chol

# mM

PL TG chol0

5

10

15

20

25 EFASEFAD


mM

PL TG chol0

5

10

15

20

25 EFASEFAD


#

*

mM

0

5

10

15

20

25

0

5

10

15

20

25 EFASEFADEFASEFASEFADEFAD


PL TG cholPL TG chol

# mM

PL TG chol0

5

10

15

20

25

0

5

10

15

20



mM

PL TG cholTG chol0

5

10

15

20

25

0

5

10

15

20



#

*

Figure 6a: Concentrations of PL, TG and cholesterol (chol) in mM, in lymph of EFA-deficient (EFAD, white bars) mice andEFA-sufficient (EFAS, black bars) controls at baseline, and at one and two hours after intraduodenal lipid administration.Data represent means ± SD of 6-9 mice per group. *p<0.05, #p<0.005 for differences between EFAD and EFAS mice.

Figure 6b: Relative concentrations of PL, TG and cholesterol), expressed as % of total lipid, in lymph of EFAD (white bars) miceand EFAS (black bars) controls at baseline, and at 1 and 2 hours after intraduodenal lipid administration. Data represent means± SD of 6-9 mice per group. *p<0.05, #p<0.005 for differences between EFAD and EFAS mice.

6a

6b

6c


DISCUSSION

Biliary phospholipids (PL) facilitate efficient transport of dietary lipids from the intes-

tinal lumen into lymph, primarily by providing the surface coat for chylomicrons, but

also by stimulating apoB48 synthesis and maintaining adequate enterocyte

membrane composition.

Several conditions can alter biliary phospholipid secretion; essential fatty acid (EFA)

deficiency in mice is associated with decreased EFA contents of biliary phospholipid

and profoundly increased bile flow, whereas Mdr2-Pgp-deficiency is associated with

virtual absence of biliary phospholipid secretion. We previously demonstrated that in

both of these murine models, postprandial plasma appearance of enterally adminis-

tered lipids is decreased. In EFA-deficient mice, this is combined with decreased net

intestinal fat absorption compared to EFA-sufficient controls, as determined by fecal

fat balance. In Mdr2 -/- mice, however, net intestinal fat absorption is only marginally

affected, indicating that despite postprandial hypolipidemia, dietary fat is eventually

almost quantitatively absorbed during biliary phospholipid deficiency. In the present

study, we applied these two in vivo models to investigate the relationship between

biliary phospholipid secretion rate and the size and composition of lipoproteins

produced by the intestine.

Our data indicate that in the absence of biliary phospholipid secretion, significantly

larger lipoproteins are secreted into lymph. The secretion of large lymphatic lipo-

proteins could be deduced from several independent observations: from particle size

determination by dynamic light scattering techniques, from the relatively increased

triglyceride and decreased phospholipid and cholesterol concentrations in lymph of

Mdr2 -/- mice, and from the calculated lymphatic lipoprotein core-to-surface ratio. Our

results on altered intestinal chylomicron formation during biliary phospholipid

scarcity in mice are in line with those of Ahn et al., who reported on decreased

lymphatic PL and TG output and an increased lymphatic TG-to-PL ratio in zinc-

deficient rats, possibly due to limited supply of biliary phospholipids to the entero-

cytes during zinc deficiency(1). The lipoproteins secreted by Mdr2 -/- mice were

continuously larger during active fat absorption, but strikingly, in non-fasted mice, the

size difference was only significant during the first two hours of lymph cannulation.

Since mice had access to chow ad libitum in the night prior to the lymph cannulation

experiments, this phenomenon could refer to a delay in intestinal chylomicron

formation or transport into lymph (and subsequently into the plasma compartment)

in Mdr2 -/- mice, as previously postulated(26). Possibly, the large chylomicrons

assembled during intraluminal phospholipid deficiency enter the lymph more slowly

than during sufficient intestinal phospholipid availability. This speculation, combined

86

Chapter 4


with the fact that there is no quantitative lipid malabsorption in Mdr2 -/- mice, supports

the concept that the amount of intraluminal bile phospholipid is important for the rate

of, but not for net intestinal absorption of dietary lipid.

Remarkably, cholesterol was virtually absent in intestinal lipoproteins secreted by

Mdr2 -/- mice, possibly related to the fact that biliary cholesterol secretion is strongly

reduced in these animals. Voshol et al. previously reported on reduced intestinal

cholesterol absorption in Mdr2 -/- mice(25; 26), which was postulated to result from

increased intestinal de novo cholesterol synthesis combined with accelerated

enterocyte desquamation due to exposure to detergent lipid-free bile. However,

Kruit et al. recently reported that fractional cholesterol absorption is unimpaired in

Mdr2 -/- mice(12). The application of different methods for quantifying cholesterol

absorption, i.e., the plasma dual isotope method and the fecal dual isotope method,

respectively, probably explains the discrepancy between these studies. The reduc-

tion in plasma high-density lipoprotein (HDL) cholesterol levels noted in both reports

is conceivably related to altered chylomicron composition and secretion in Mdr2 -/-

mice, since a major part of HDL is thought to be derived from excess chylomicron

surface material (phospholipid and cholesterol), shed during the lipolytic process.

The intestinal requirement of phospholipid for production of chylomicrons may be

comparable to that of the liver for the assembly and secretion of VLDL. Verkade et al.

demonstrated in choline-deprived rats that during hepatic phosphatidylcholine

scarcity, fewer but larger VLDL particles are secreted from the liver(24; 32). We recently

observed that EFA-deficient mice secreted larger VLDL particles from the liver than

controls(29), possibly secondary to hepatic phosphatidylcholine depletion due to

increased biliary phospholipid secretion rates.

Animal models for increased biliary phospholipid secretion are relatively rare. Since

the profoundly augmented biliary phospholipid secretion rate in EFA-deficient mice

affects hepatic lipoprotein size and is associated with dietary fat malabsorption, we

considered the EFA-deficient mouse an intriguing model to further study the poten-

tially organ-specific effects of phospholipid availability on intestinal lipoprotein

production. Our data demonstrate that EFA deficiency in mice is associated with lym-

phatic secretion of considerably smaller chylomicrons compared to EFA-sufficient

controls, as determined by three different particle size estimation techniques. Since

EFA-deficient mice secrete larger hepatic VLDL particles than EFA-sufficient controls,

secretion of smaller lipoproteins apparently is not an intrinsic feature of EFA defi-

ciency. Rather, EFA deficiency differentially affects lipoprotein size in liver and intes-

tine, with phospholipid availability as the major determinant of lipoprotein size.

87



Our data are in accordance with studies by Amate et al., who demonstrated that

dietary EFA-rich phospholipid administration to piglets resulted in production of

lymphatic lipoproteins with significantly smaller diameters compared to piglets fed

EFA-rich triglycerides(2).

Our present data do not allow conclusions on quantitative lipid absorption due to the

highly variable lymph flow rates in these studies. Possibly, continuous enteral lipid

administration could overcome this limitation of our murine lymph cannulation model

in future experiments.

Although the Mdr2 -/- and EFA-deficient mouse models correspond well regarding

postprandial hypolipidemia, they profoundly differ with respect to overall intestinal

lipid absorption as determined by fat balance. Whilst EFA deficiency in our mouse

model profoundly decreased EFA levels in acyl chains of biliary phospholipid, as

described by Bennett-Clark(3; 4) in EFA-deficient rats, we previously demonstrated that

EFA-deficient Mdr2 -/- mice have similar fat malabsorption as EFA-deficient Mdr2 +/+

mice(30). Since EFA-sufficient Mdr2 -/- mice do not malabsorb dietary fat, this indicates

that neither absence of biliary PL secretion, nor EFA depletion of biliary phospholipid

quantitatively affect intestinal fat absorption. In the present study we chose not to

investigate EFA-deficient Mdr2 -/- mice, since these animals do not differ in their

absence of biliary phospholipid secretion compared with EFA-sufficient Mdr2 -/- mice.

Baseline lymphatic chylomicrons from EFA-sufficient were significantly larger than

those from Mdr2 +/+ mice (Figures 1, 3 and 5b), which can be attributed to the fact

that the EFAD and EFAS diets were high-fat diets (34 energy% fat) and the Mdr2 +/+

and Mdr2 -/- mice were fed standard chow (14 energy% fat).

Hayashi and Tso demonstrated that the number of secreted intestinal lipoproteins

remains relatively constant during active lipid absorption, while lipoprotein size

increases(9). Obviously, in Mdr2-deficiency, an increased TG-to-PL ratio is a highly

favorable means to maintain triglyceride packaging into chylomicrons during lack of

surface coat material. In addition, up to 20% of lipoprotein phospholipid is thought to

be derived from de novo synthesis, and it could be speculated that in biliary phos-

pholipid deficiency, intestinal phospholipids also partially compensates for biliary

phospholipids for chylomicron surface coating.

Not only biliary but also enterocytic phospholipids are markedly EFA-depleted

during EFA deficiency(10; 11; 16). The rapid intestinal cellular turnover rate of enterocytes,

which is even faster during EFA deficiency(5), renders the membranes of the intestin-

al mucosa particularly sensitive to altered intraluminal fatty acid availability. Structural

membrane modifications, such as an increased degree of phospholipid fatty acid

ÿ

Chapter 4


saturation, affect the fluidity of intestinal membranes and may subsequently result in

functional alterations of membrane enzymes and transporters, thus impairing intra-

cellular events involved in chylomicron formation or secretion.

The deviant size and composition of lipoproteins secreted during altered biliary phos-

pholipid secretion may not only affect the rate of plasma appearance of dietary

triglycerides, but also intravascular chylomicron metabolism. Smaller chylomicrons

have less affinity for lipoprotein lipase (LPL) than large ones, and there is compe-

tition between small chylomicrons and VLDL for available LPL on the capillary

endothelium(31). Plasma clearance of large chylomicrons is substantially more rapid

than that of small particles(18). Additionally, altered acyl chain composition of EFA-

depleted biliary phospholipids on the chylomicron surface affects chylomicron

clearance(19); chylomicrons with saturated phospholipids are cleared more slowly

than chylomicrons with a high content of PUFA in surface phospholipids. Thus, by

affecting lipoprotein size and clearance rates, biliary phospholipids may affect

plasma and hepatic lipid levels(29).

Our data are compatible with the concept that the size of intestinal chylomicrons,

secreted into lymph under basal conditions or during active lipid absorption, is

inversely related to the quantity of biliary phospholipid secretion, and that altered

lymphatic lipoprotein size is not necessarily related to net dietary fat absorption as

determined by fat balance.

REFERENCES1. Ahn J and Koo SI. Intraduodenal phosphatidylcholine infusion restores the lymphatic absorption of vitamin A and oleic acidin zinc-deficient rats. Nutritional Biochemistry 6: 604-612, 1995.2. Amate L, Gil A and Ramirez M. Dietary long-chain PUFA in the form of TAG or phospholipids influence lymph lipoprotein sizeand composition in piglets. Lipids 37: 975-980, 2002.3. Bennett Clark S, Ekkers TE, Singh A, Balint JA, Holt PR and Rodgers JB. Fat absorption in EFA deficiency: a model experi-mental approach to studies of the mechanism of fat malabsorption of unknown etiology. JLR 14: 581-588, 1973.4. Bennett Clark S. Chylomicron composition during duodenal triglyceride and lecithin infusion. AJP 235: E183-E190, 1978.5. Bull LN, van Eijk MJT and Pawlikowska L. PFIC types 1, 2, and 3. Gut 42: 766-767, 1998.6. Davidson NO, Drewek MJ, Gordon JI and Elovson J. Rat intestinal apolipoprotein B gene expression. Evidence for integrat-ed regulation by bile salt, fatty acid, and phospholipid flux. J Clin Invest 82: 300-308, 1988.7. Enser M and Bartley W. The effect of EFA deficiency on the fatty acid composition of the total lipid of the intestine. BiochemJ 85: 607-614, 1962.8. Groen AK, Van Wijland MJ, Frederiks WM, Smit JJ, Schinkel AH and Oude Elferink RP. Regulation of protein secretion intobile: studies in mice with a disrupted mdr2 p-glycoprotein gene. Gastroenterology 109: 1997-2006, 1995.9. Hayashi H, Fujimoto K, Cardelli JA, Nutting DF, Bergstedt S and Tso P. Fat feeding increases size, but not number, of chy-lomicrons produced by small intestine. Am J Physiol 259: G709-G719, 1990.10. Hoffman DF and Uauy R. Essentiality of dietary omega-3 fatty acids for premature infants: plasma and red blood cell fattyacid composition. Lipids 27: 886-895, 1992.11. Kalivianakis M and Verkade HJ. The mechanism of fat malabsorption in cystic fibrosis patients. Nutrition 15: 167-169, 1999.

89



12. Kruit JK, Plosch T, Havinga R, Boverhof R, Groot PH, Groen AK and Kuipers F. Increased fecal neutral sterol loss upon liverX receptor activation is independent of biliary sterol secretion in mice. Gastroenterology 128: 147-156, 2005.13. Mansbach CM. The origin of chylomicron PC in the rat. J Clin Invest 60: 411-420, 1977.14. Mathur SN, Born E, Murthy S and Field FJ. Phosphatidylcholine increases the secretion of triacylglycerol-rich lipoproteinsby CaCo-2 cells. Biochem J 314: 569-575, 1996.15. Minich DM, Voshol PJ, Havinga R, Stellaard F, Kuipers F, Vonk RJ and Verkade HJ. Biliary phospholipid secretion is notrequired for intestinal absorption and plasma status of linoleic acid in mice. Biochimica et biophysica acta 1441: 14-22, 1999.16. Otte JB, De Ville De Goyet J, Reding R, Hausleithner V, Sokal E, Chardot C and Debande B. Sequential treatment of biliaryatresia with Kasai portoenterostomy and liver transplantation: a review. Hepatology 20: 41S-48S, 1994.17. Patton GM, Bennett Clark S, Fasulo JM and Robins SJ. Utilization of individual lecithins in intestinal lipoprotein formation inthe rat. J Clin Invest 73: 231-240, 1984.18. Quarfordt SH and Goodman DS. Heterogeneity in the rate of plasma clearance of chylomicrons of different size. BiochimBiophys Acta 116: 382-385, 1966.19. Robins SJ, Fasulo JM and Patton GM. Effect of different molecular species of PC on the clearance of emulsion particle lipids.J Lipid Res 29: 1195-1203, 1988.20. Scow RO, Stein Y and Stein O. Incorporation of dietary lecithin and lysolecithin into lymph chylomicrons in the rat. J BiolChem 242: 4919-4924, 1967.21. Smit JJ, Schinkel AH, Oude Elferink RP, Groen AK, Wagenaar E, van Deemter L, Mol CA, Ottenhoff R, van der Lugt NM, vanRoon MA. Homozygous disruption of the murine mdr2 Pgp gene leads to a complete absence of PL from bile and to liver dis-ease. Cell 75: 451-462, 1993.22. Snipes RL. The effects of EFA deficiency on the ultrastructure and functional capacity of the jejunal epithelium. Lab Invest18: 179-189, 1968.23. Tso P, Kendrick H, Balint JA and Simmonds WJ. Role of biliary phosphatidylcholine in the absorption and transport of dietarytriolein in the rat. Gastroenterology 80: 60-65, 1981.24. Verkade HJ, Fast DG, Rusinol AE, Scraba DG and Vance DE. Impaired biosynthesis of PC causes a decrease in the num-ber of very low density lipoprotein particles in the Golgi but not in the ER of rat liver. J Biol Chem 268: 24990-24996, 1993.25. Voshol PJ, Havinga R, Wolters H, Ottenhoff R, Princen HM, Oude Elferink RP, Groen AK and Kuipers F. Reduced plasmacholesterol and increased fecal sterol loss in mdr 2 P-glycoprotein-deficient mice. Gastroenterology 114: 1024-1034, 1998.26. Voshol PJ, Minich DM, Havinga R, OudeElferink RPJ, Verkade HJ, Groen AK and Kuipers F. Postprandial chylomicron for-mation and fat absorption in multidrug resistance gene 2 p-glycoprotein-deficient mice. Gastroenterology 118: 173-182, 2000.27. Wang DQ and Carey MC. Measurement of intestinal cholesterol absorption by plasma and fecal dual isotope ratio, massbalance, and lymph fistula methods in the mouse: An analysis of direct versus indirect methodologies. J Lipid Res .: 2003.28. Wang DQ, Paigen B, Carey MC. Genetic factors at the enterocyte level account for variations in intestinal cholesterol absorp-tion efficiency among inbred strains of mice. JLR42: 1820-30, 2001.29. Werner A, Havinga R, Bos T, Bloks VW, Kuipers F and Verkade HJ. EFA deficiency in mice is associated with hepatic steato-sis and secretion of large VLDL. AJP: 2005.30. Werner A, Minich DM, Havinga R, Bloks V, Van Goor H, Kuipers F and Verkade HJ. Fat malabsorption in essential fatty acid-deficient mice is not due to impaired bile formation. Am J Physiol Gastrointest Liver Physiol 283: G900-G908, 2002.31. Xiang SQ, Cianflone K, Kalant D, Sniderman AD. Differential binding of TG-rich lipoproteins to LPL. JLR 40: 1655-63, 1999.32. Yao ZM and Vance DE. The active synthesis of phosphatidylcholine is required for very low density lipoprotein secretionfrom rat hepatocytes. J Biol Chem 263: 2998-3004, 1988.

90

Chapter 4


No indications for altered essentialfatty acid metabolism in two

murine models for cystic fibrosis5

A. WernerM.E.J. BongersM.J.C. BijveldsH.R. de JongeH.J. Verkade

J Lipid Res. 2004; 45(12): 2277-2286


ABSTRACT

Background: A deficiency of essential fatty acids (EFA) is frequently described in

cystic fibrosis (CF), but whether this is a primary consequence of altered EFA

metabolism or a secondary phenomenon, is unclear. It was suggested that defective

long-chain polyunsaturated fatty acid (LCPUFA) synthesis contributes to CF pheno-

type. To establish whether CFTR dysfunction affects LCPUFA synthesis, we

quantified EFA metabolism in cftr -/-CAM and cftr +/+CAM mice.

Methods: Effects of intestinal phenotype, diet, age and genetic background on EFA

status were evaluated in cftr -/-CAM mice, dF508/dF508 mice and littermate controls.

EFA metabolism was measured by 13C stable isotope methodology in vivo. EFA

status was determined by gas chromatography in tissues of cftr -/-CAM mice,

dF508/dF508 mice, littermate controls and C57Bl/6 wildtypes, fed chow or liquid diet.

Results: After enteral administration of 13C-EFA, arachidonic acid (AA) and docosa-

hexaenoic acid (DHA) were equally 13C-enriched in cftr -/-CAM and cftr +/+CAM mice,

indicating similar EFA elongation/desaturation rates. LA, ALA, AA and DHA

concentrations were equal in pancreas, lung and jejunum of chow-fed cftr -/-CAM and

dF508/dF508 mice and controls. LCPUFA levels were also equal in liquid diet-

weaned cftr -/-CAM mice and littermate controls, but consistently higher than in age- and

diet-matched C57Bl/6 wildtypes.

Conclusions: Cftr -/-CAM mice adequately absorb and metabolize EFA, indicating that

CFTR dysfunction does not impair LCPUFA synthesis. A membrane EFA imbalance

is not inextricably linked to CF genotype. EFA status in murine CF models is

strongly determined by genetic background.

92

Chapter 5


INTRODUCTION

A deficiency of essential fatty acids (EFA) or their long-chain polyunsaturated

metabolites (LCPUFA) has frequently been reported in CF patients(1-4) and has

formerly been attributed to fat malabsorption due to pancreatic insufficiency. Current

high-fat, hypercaloric nutritional strategies and improved pancreas enzyme

replacement therapies can usually maintain patients in optimal nutritional status,

thus normalizing EFA status in many CF patients(5). Yet, several reports still indicate

the occurrence of EFA deficiency in CF(6-8). Although some authors have suggested

that residual fat malabsorption and increased EFA turnover in CF may compromise

EFA status(9; 10), the exact pathophysiology of EFA deficiency in CF patients has not

been elucidated.

A direct link between CFTR dysfunction and EFA metabolism has been postulated by

Gilljam et al.(11), and Bhura-Bandali et al.(12) described impaired EFA incorporation into

phospholipids in human pancreatic CF cells. In cftr -/-UNC mice, Freedman et al.(13)

reported a profound membrane fatty acid imbalance, characterized by increased

concentrations of arachidonic acid (AA) and decreased concentrations of docosa-

hexaenoic acid (DHA) in membrane phospholipids of organs typically affected in CF,

such as pancreas, lung and intestine. Oral supplementation with DHA, but not with

its precursor ALA, corrected this lipid imbalance and was reported to reverse certain

pathological features of the disease. These studies suggested that CFTR exerts

control over LCPUFA synthesis from EFA, and that impaired EFA processing

primarily contributes to CF pathology(13; 14). However, it has not been elucidated

whether perturbed EFA status in CF is a primary result of CFTR malfunction or

secondary to fat malabsorption or increased turnover.

Several CF mouse models have been developed in the past decade, including total

null mice with no detectable CFTR production(15-17) as well as mice with the delta F508

mutation (dF508/dF508 mice), which have low-level residual CFTR activity(18). Similar

to CF patients, CF mouse models display significant phenotypic variability, particu-

larly concerning the severity of gastrointestinal symptoms such as intestinal obstruc-

tion and fat malabsorption.

To assess the effect of CFTR on LCPUFA synthesis, we quantified conversion of EFA

into LCPUFA in vivo in cftr -/-CAM mice and littermate controls. In addition, we analyzed

fecal fatty acid excretion and membrane fatty acid composition in tissues of cftr -/-CAM

mice(17) and homozygous dF508 mice(18), and of their respective littermate controls.

These particular CF models have been demonstrated to differ in intestinal phenotype,

with fat malabsorption present in cftr -/-CAM mice but absent in dF508/dF508 mice(19).

93

No indications for altered EFA metabolism in two murine models for cystic fibrosis


Furthermore, we determined the effects of age, diet and genetic background on EFA

status in these mouse models and in C57Bl/6 wildtype mice. Our results indicate that

cftr -/-CAM mice adequately absorb, elongate and desaturate intragastrically adminis-

tered EFA, and that a membrane fatty acid imbalance in CF-affected tissues is not

inherent to CF genotype in mouse models with and without fat malabsorption.

Rather, EFA status in CF mice is strongly determined by genetic background, diet

and age.

METHODS

AnimalsC57Bl/6/129 cftr -/-tm1CAM mice and cftr +/+tm1CAM littermates(17), homozygous dF508 mice

and sex-matched littermate controls (N/N) of FVB/129 background(18) and wildtype

C57Bl/6 mice were accommodated at the breeding colony at the Erasmus Medical

Center, Rotterdam, the Netherlands. Southern blotting of tail-clip DNA was performed

to verify the genotype of individual animals(20). Mice were housed in a light-controlled

(lights on 6 AM to 6 PM) and temperature-controlled (21°C) facility and were allowed

tap water and standard laboratory chow (Hope Farms BV Woerden, the Netherlands)

or liquid diet (Peptamen) ad libitum from the time of weaning. The Ethical Committee

for Animal Experiments in Rotterdam approved of the experimental protocols.

Experimental dietsThe standard laboratory chow contained 6 weight% fat and 14 energy% fat, and had

the following fatty acid composition: 18.2 mol% palmitic acid (C16:0), 7.0 mol%

stearic acid (C18:0), 25.8 mol% oleic acid (C18:1n-9), 39.1 mol% linoleic acid

(C18:2n-6), 3.5 mol% alpha-linolenic acid (C18:3n-3), 0.3 mol% arachidonic acid

(C20:4n-6) and 0.05 mol% docosahexaenoic acid (C22:6n-3) (Hope Farms BV,

Woerden, the Netherlands). The Peptamen liquid diet (Nestle Clinical Nutrition,

Brussels, Belgium) contained 3.7 g fat/100 ml (33 energy%) and had the following

fatty acid composition: 16.4 mol% palmitic acid (C16:0), 6.7 mol% stearic acid

(C18:0), 22.2 mol% oleic acid (C18:1n-9), 43.6 mol% linoleic acid (C18:2n-6),

4.6 mol% alpha-linolenic acid (C18:3n-3), 0.1 mol% arachidonic acid (C20:4n-6) and

0.08 mol% docosahexaenoic acid (C22:6n-3).

Experimental proceduresCftr -/-CAM mice and cftr +/+CAM littermates (n=5-6 per group) were fed standard labora-

tory chow from weaning. At 3 months of age, mice were anesthetized with isoflurane

94

Chapter 5


and a baseline blood sample was obtained by tail bleeding. Subsequently, a 100 ml

lipid bolus containing uniformly labeled 13C-LA and 13C-ALA was slowly administered

by intragastric gavage, for determination of in vivo conversion of EFA into LCPUFA

and partitioning to different organs. The lipid bolus was composed of olive oil mixed

with U-13C-LA (0.40 mg) and U-13C-ALA (0.40 mg) (Martek Biosciences Corporation,

Columbia, MD, USA). U-13C-LA and U-13C-ALA were 99% 13C-enriched, with a

chemical purity exceeding 97%. At 24 hours after bolus administration, a large blood

sample was obtained by cardiac puncture and pancreas, liver, lungs and intestine

were removed and stored at -80°C until further analysis. Intestine and lungs were

flushed with ice-cold 0.9% (w/w) NaCl before storage. Blood was collected in

heparinized vials and plasma and erythrocytes were separated by centrifugation.

Erythrocyte membrane lipids were hydrolyzed and methylated for fatty acid analysis

the same day(21) to prevent fatty acid oxidation, and plasma was stored at -80°C.

To establish the effect of fat malabsorption, diet, age and genetic background on

body EFA status, homozygous dF508 mice of FVB/129 background, sex-matched

N/N littermates and cftr -/-CAM and cftr +/+CAM mice (n=5-6 per group) were fed standard

laboratory chow from weaning. At the age of three months, mice were anesthetized

with isoflurane and sacrificed by means of cardiac puncture. Lung, pancreas and

jejunum were removed and samples of each were immediately stored at -80°C for

fatty acid and protein analysis. Fecal fatty acid excretion was quantified by gas-

chromatographic analysis of feces aliquots obtained after a 72h fat balance.

A separate group of cftr -/-CAM and cftr +/+CAM mice and C57Bl/6 wildtypes (n=6 per

group) were weaned at 23 days of age, and subsequently put on Peptamen liquid

diet ad libitum for 7 days. At postnatal day 30, mice were anaesthetized and blood,

pancreas, lung, jejunum, ileum, and liver samples were obtained. Jejunum, ileum

and lungs were flushed with ice-cold saline and ileal mucosa was separated from

submucosal layers by scraping with a glass microscope slide on an ice-cooled glass

plate. Pancreatic cell suspensions were prepared by mechanical dissociation and

addition of collagenase as described by Bruzzone et al.(22). Lung tissue was flushed

with Krebs-Henseleit buffer (KHB), pH 7.4, containing 0.5% BSA to rinse off con-

taminating blood. Lung tissue was then finely cut and suspended in 10 ml of

oxygenated KHB containing 1000 units of collagenase, 2000 units of DNAse and 0.5

units of thermolysin, and incubated for 30 min at 37°C. The lung cell suspension was

then sedimented through KHB containing 4% BSA and washed once in KHB.

All organ samples were stored at -80°C until further analysis.

95



Analytical techniques13C enrichment analysis

Gas chromatography combustion isotope ratio mass spectrometry (GC-C-IRMS;

DeltaPlusXL, Thermo Finnigan, Bremen, Germany) was used to measure 13C enrich-

ment of LA and ALA and their metabolites. The GC-C-IRMS was equipped with a

50mx0.22mm BPX70 capillary column and the injector temperature was set at 275°C

with splitless injection. The gas chromatograph oven was programmed from an

initial temperature of 50°C to a final temperature of 250°C in 3 steps (50°C, held 1 min

isotherm; 50-100°C, ramp 7°C/min; 100-225°C, ramp 10°C/min; 225-250°C, ramp

25°C/min, held 10 min). Helium was used as a carrier gas with a constant flow rate

of 0.5 ml per minute. 12CO2+ and 13CO2

+ ions were measured at m/z 44 and 45.

Correction for 17O was achieved by measurement of 18O abundance at m/z 46.

Fatty acid analysis

Fatty acid profiles were determined by hydrolyzing, methylating and extracting total

plasma lipids and erythrocyte membrane lipids as described by Muskiet et al.(21).

For fatty acid analysis of liver, intestine, pancreas and lung tissue, samples were

mechanically homogenized in 0.9% NaCl and lipids were extracted from aliquots of

tissue homogenate as described by Bligh and Dyer(23). The lipid extract was partly

methylated in toto for GC analysis, and partly fractionated into phospholipids,

cholesterol esters, triacylglycerols, diacylglycerols, monoacylglycerols and free fatty

acids using thin-layer chromatography (TLC) (20x20 cm, Silica gel 60 F254, Merck)

with hexane/diethyl ether/acetic acid (80:20:1, v/v/v) as solvent. TLC plates were

dried and colored by iodine, and PL and TG spots were scraped. Of these scrapings,

fatty acid methyl esters were prepared as mentioned above. To account for losses

during lipid extraction, heptadecanoic acid (C17:0) was added to all samples as

internal standard prior to Bligh & Dyer procedures. BHT was added as antioxidant.

Aliquots of chow diet and feces were freeze-dried and homogenized, after which

lipids were hydrolyzed, methylated and extracted for fatty acid analysis. Similarly,

fatty acid composition of Peptamen liquid diet was determined after dissolution in

chloroform/methanol (2/1 v/v).

Fatty acid methyl esters were separated and quantified by gas liquid chroma-

tography (GLC) on a Hewlett Packard gas chromatograph model 6890, with a


detector as described previously(24). We verified purity of AA and DHA peaks as

separated by GLC, using a gas chromatography-mass spectrometer (GC-MS;

Finnigan MAT SSQ7000), alternately equipped with a 50mx0.2mm Ultra 1 capillary

96

Chapter 5


column or a 50mx0.22mm BPX70 capillary column (SGE, Weiterstadt, Germany).

Both methylesters and pentafluorbenzylbromide (PFB-Br) derivatives of tissue fatty

acids were analyzed, no indications for impurity of AA or DHA peaks were detected.

Protein analysis

Total protein contents of tissue homogenates were determined with Folin phenol

reagent as described by Lowry et al.(25) Standard Pierce BSA was used as reference.

Calculations13C abundance was expressed as d13C-PDB value, i.e. the difference between the

sample value and baseline compared to Pee Dee Belemnite limestone. d13C-PDB

values were converted to atom% 13C values. Enrichment (atom% excess) was

calculated by subtracting baseline 13C abundance from all enriched values.

Relative concentrations (mol%) of individual fatty acids in plasma, erythrocytes, liver,

intestine, pancreas and lung were calculated by summation of all fatty acid peak

areas and subsequent expression of areas of individual fatty acids as a percentage

of this amount. Fatty acid contents were quantified by relating the areas of their

chromatogram peaks to that of the internal standard heptadecanoic acid (C17:0).

StatisticsAll results are presented as means ± S.D. for the number of animals indicated. Data

were statistically analyzed using Student's t-test or, for comparison of more than two

groups, ANOVA-test with post-hoc Bonferroni correction. Statistical significance of

differences between means was accepted at p<0.05. Analyses were performed

using SPSS for Windows software (SPSS, Chicago, IL).

RESULTS

In vivo conversion of 13C-labeled EFA into LCPUFALCPUFA in specific tissues originate either from the diet or from endogenous

synthesis by elongation and desaturation of EFA. As CFTR dysfunction has been

postulated to affect EFA tissue incorporation or rate of metabolism(12; 13; 26; 27), we quan-

tified in vivo the appearance in different organs of ingested 13C-labeled EFA, and their

conversion into LCPUFA. At 24 hours after intragastric administration of 13C-LA and13C-ALA, 13C enrichment of LA and ALA could be demonstrated in all analyzed tissues

of cftr -/-CAM mice and cftr +/+CAM controls. 13C-LA and 13C-ALA concentrations were not

significantly different between cftr -/-CAM and cftr +/+CAM mice (Figure 1).

97



Similarly, 13C-AA levels did not significantly differ between cftr -/-CAM mice and littermate

controls. 13C enrichment of DHA was below detection limit in liver triglycerides (not

shown) and in lung PL, but 13C-DHA in jejunum, pancreas and liver phospholipids

was similar in cftr -/-CAM mice and controls. The ratios between 13C-LA and 13C-AA and

between 13C-ALA and 13C-DHA in liver, pancreas, lung, and intestine of cftr -/-CAM and

cftr +/+CAM mice were highly comparable, suggesting adequate rates of LA and ALA

elongation and desaturation in this CF mouse model.

Fatty acid composition of feces and tissue homogenatesTo determine the effects of intestinal phenotype on EFA status, we analyzed fatty acid

composition of feces and of CF-affected organs in CF mouse models with and

without fat malabsorption. Figure 2 shows the daily fecal excretion of the main dietary

fatty acids in dF508/dF508 and cftr -/-CAM mice and their respective controls. Fecal fatty

acid excretion was similar in homozygous dF508 mice and controls, in contrast to

cftr -/-CAM mice which secreted significantly more fatty acids in feces than littermate

controls, confirming the presence of fat malabsorption in this CF mouse model.

Malabsorption of saturated fatty acids was slightly more pronounced than that of

unsaturated fatty acids.

98

Chapter 5

jejunum PL 13C -LA and 13C-AA

LA AA AA/LA

100

x A

TE

0

2

4

6

8

10

13C-ALA and 13C-DHA

0

1

2

3

4

5

ALA DHA DHA/ALA

100

x A

TE

cftr +/+

cftr -/-

13C-ALA and 13C-DHA

0

1

2

3

4

5

100

x A

TE

ALA DHA DHA/ALA

pancreas PL13C-LA and 13C-AA

100

x A

TE

0

2

4

6

8

10

LA AA AA/LA

lung PL13C-LA and 13C-AA

0

2

4

6

8

10

100

x A

TE

LA AA AA/LA

13C-ALA and 13C-DHA

0

1

2

100

x A

TE

ALA DHA DHA/ALA

nd nd

13C-ALA and 13C-DHA

ALA DHA DHA/ALA

liver PL 13C -LA and 13C-AA

AA AA/LA

0

5

10

15

100

x A

TE

LA

100

x A

TE

0

2

4

6

8

10

cftr +/+

cftr -/-


LA AA AA/LA

100

x A

TE

0

2

4

6

8

10

13C-ALA and 13C-DHA

0

1

2

3

4

5

ALA DHA DHA/ALA

100

x A

TE

cftr +/+

cftr -/-


LA AA AA/LA

100

x A

TE

0

2

4

6

8

10

0

2

4

6

8

10

13C-ALA and 13C-DHA

0

1

2

3

4

5

0

1

2

3

4

5

ALA DHA DHA/ALA

100

x A

TE

cftr +/+

cftr -/-cftr +/+

cftr -/-

13C-ALA and 13C-DHA

0

1

2

3

4

5

100

x A

TE

ALA DHA DHA/ALA


100

x A

TE

0

2

4

6

8

10

LA AA AA/LA

13C-ALA and 13C-DHA

0

1

2

3

4

5

0

1

2

3

4

5

100

x A

TE

ALA DHA DHA/ALA


100

x A

TE

0

2

4

6

8

10

0

2

4

6

8

10

LA AA AA/LA


0

2

4

6

8

10

100

x A

TE

LA AA AA/LA

13C-ALA and 13C-DHA

0

1

2

100

x A

TE

ALA DHA DHA/ALA

nd nd


0

2

4

6

8

10

0

2

4

6

8

10

100

x A

TE

LA AA AA/LA

13C-ALA and 13C-DHA

0

1

2

0

1

2

100

x A

TE

ALA DHA DHA/ALA

nd ndnd nd

13C-ALA and 13C-DHA

ALA DHA DHA/ALA


AA AA/LA

0

5

10

15

100

x A

TE

LA

100

x A

TE

0

2

4

6

8

10

cftr +/+

cftr -/-

13C-ALA and 13C-DHA

ALA DHA DHA/ALA


AA AA/LA

0

5

10

15

100

x A

TE

LA

100

x A

TE

0

2

4

6

8

10

cftr +/+

cftr -/-cftr +/+

cftr -/-cftr +/+

cftr -/-

Figure 1: 13C enrichment of linoleic acid (LA, C18:2n-6), arachidonic acid (AA, C20:4n-6), alpha-linolenic acid (ALA, C18:3n-3)and docosahexaenoic acid (DHA, C22:6n-3) in phospholipids of pancreas, lung, intestine and liver of cftr+/+ mice (grey bars)and cftr-/- mice (white bars) at 24h after intragastric administration of 13C-LA and 13C-ALA. Individual fatty acid 13C enrichmentwas calculated from the difference between the sample value and baseline compared to Pee Dee Belemnite limestone (d13C-PDB), and is expressed as 100 x atom % excess. Data represent means ± S.D. of 6 mice per group. No significant differencesin 13C enrichment were detected between cftr+/+ and cftr-/- mice, indicating normal conversion of EFA into LCPUFA.

1


Figure 3 shows LA, AA, ALA and DHA concentrations in tissue homogenates of

pancreas, lung and jejunum in homozygous dF508 mice and cftr -/-CAM mice compared

to their respective littermate controls. No significant differences were observed in

relative concentrations of either these EFA and LCPUFA, nor of saturated and non-

essential fatty acids (data not shown).

The fatty acid composition of the total lipid fraction depends on the fatty acid com-

position of the various lipid classes present (i.e., triglycerides, diglycerides, mono-

glycerides, phospholipids, cholesterol esters, free fatty acids), and the proportions of

these lipid classes may vary considerably between organs. Since phospholipids (PL)

may be a more adequate indicator for determination of EFA status, we specifically

analyzed fatty acid composition of tissue PL fractions of CF-affected organs. Figure

4a shows relative fatty acid concentrations in the PL fraction of pancreas, lung and

99


fecal fatty acid excretionm

mol

/day

cftr+/+cftr-/-

PA SA OA LA DHA0.00

0.01

0.02

0.03

0.04

0.05

*

**

*

*0.00

0.01

0.02

0.03

0.04

0.05

N/NdF508

PA SA OA LA DHA

fecal fatty acid excretionm

mol

/day

cftr+/+cftr-/-

PA SA OA LA DHA0.00

0.01

0.02

0.03

0.04

0.05

*

**

*

*

cftr+/+cftr-/-cftr+/+cftr+/+cftr-/-cftr-/-

PA SA OA LA DHAPA SA OA LA DHA0.00

0.01

0.02

0.03

0.04

0.05

0.00

0.01

0.02

0.03

0.04

0.05

*

**

*

*0.00

0.01

0.02

0.03

0.04

0.05

N/NdF508

PA SA OA LA DHA0.00

0.01

0.02

0.03

0.04

0.05

0.00

0.01

0.02

0.03

0.04

0.05

N/NdF508N/NN/NdF508dF508

PA SA OA LA DHAPA SA OA LA DHA

Figure 2: Relative concentrations of palmitic acid (PA, C16:0), stearic acid (SA, C18:0), oleic acid (OA, C18:1n-9), linoleic acid(LA, C18:2n-6) and docosahexaenoic acid (DHA, C22:6n-3) in fecal lipid extracts of homozygous dF508 mice and cftr-/- miceand their respective littermate controls. Fecal fat excretion was quantified by means of a 72h fecal fat balance. Individual fattyacid concentrations are expressed as mmol of fatty acid excreted per day. Data represent means ± S.D. of 5-6 mice per group.No significant differences were detected for any of the fatty acids between homozygous dF508 mice and controls, but dailyfecal fatty acid excretion was significantly higher in cftr-/- mice than in cftr+/+ controls (p<0.05).

2

lung total lipid

lung total lipid

jejunum total lipid

jejunum total lipid

cftr +/+

cftr -/-N/N?F508/?F508

pancreas total lipid

pancreastotal lipid

0

5

10

15

LA AA ALA DHA

0

10

20

30

LA AA ALA DHA

0

10

20

30

LA AA ALA DHA

mo

l%m

ol%

mo

l%

0

10

20

30

LA AA ALA DHA

0

5

10

15

LA AA ALA DHA

0

10

20

30

LA AA ALA DHA

mo

l%m

ol%

mo

l%

lung total lipid

lung total lipid

lung total lipid

lung total lipid

jejunum total lipid

jejunum total lipid

jejunum total lipid

jejunum total lipid

cftr +/+

cftr -/-N/N?F508/?F508

cftr +/+

cftr -/-cftr +/+

cftr -/-cftr +/+

cftr -/-N/N?F508/?F508N/N?F508/?F508


pancreastotal lipid


pancreastotal lipid

0

5

10

15

LA AA ALA DHA0

5

10

15

0

5

10

15

LA AA ALA DHALA AA ALA DHA

0

10

20

30

LA AA ALA DHA0

10

20

30

0

10

20

30


0

10

20

30

LA AA ALA DHA0

10

20

30

LA AA ALA DHA0

10

20

30

0

10

20

30


mo

l%m

ol%

mo

l%m

ol%

mo

l%m

ol%

0

10

20

30

LA AA ALA DHA

0

5

10

15

LA AA ALA DHA

0

10

20

30

LA AA ALA DHA

mo

l%m

ol%

mo

l%

0

10

20

30

LA AA ALA DHA0

10

20

30

0

10

20

30

0

10

20

30


0

5

10

15

LA AA ALA DHA0

5

10

15

LA AA ALA DHA0

5

10

15

0

5

10

15

0

5

10

15


0

10

20

30

LA AA ALA DHA0

10

20

30

0

10

20

30

0

10

20

30


mo

l%m

ol%

mo

l%m

ol%

mo

l%m

ol%

Figure 3: Relative concentrations of linoleic acid(LA, C18:2n-6), arachidonic acid (AA, C20:4n-6),alpha-linolenic acid (ALA, C18:3n-3) and docosa-hexaenoic acid (DHA, C22:6n-3) in total lipid extractsof homozygous dF508 mice and cftr-/- mice and theirlittermate controls. Fatty acid concentrations areexpressed as mol% of total fatty acids. Data repre-sent means±S.D. of 5-6 mice per group. No signifi-cant differences were detected for any fatty acidbetween dF508 or cftr-/- mice and littermate controls.

3


jejunum of the two CF mouse models and their respective controls. DHA

concentrations in pancreas PL were 25% lower in dF508/dF508 mice compared to

N/N controls (p<0.05) and in lung PL, there was a small significant increase of AA

in dF508/dF508 compared to N/N mice (5.3±0.7 vs. 4.3±0.5, respectively; p<0.05).

In jejunum PL of dF508/dF508 mice, however, AA was 37% decreased (3.1±1.3 vs.

5.0±0.7, p<0.05) and ALA was 15% increased (0.48±0.04 vs. 0.41±0.04, p<0.01)

compared to N/N controls. Similar differences were not observed in tissues of cftr -/-

CAM mice, in which LA, AA, ALA or DHA concentrations were consistently similar

compared to cftr +/+CAM littermates.

The severity of CF phenotype has been implicated in the EFA status of CF patients(7).

Body weight is an important clinical parameter related to severity of CF symptoms(28),

exemplified by a consistently lower weight of cftr -/-CAM mice compared to littermate

controls (26.5±4.4 g vs. 30.7±3.8 g, respectively, p<0.05). In contrast, homozygous

dF508 mice, displaying normal fat absorption as a consequence of milder gastro-

intestinal pathology, show normal weight gain (22.4±1.9 g vs. 23.0±1.4 g, NS).

100

Chapter 5

lung phospholipid

lung phospholipid

jejunum phospholipid


cftr +/+

cftr -/-N/N?F508/?F508

pancreas phospholipid

pancreasphospholipid

0

10

20

30

LA AA ALA DHA

0

10

20

30

LA AA ALA DHA

mol

%m

ol%

mo

l%

0

10

20

30

LA AA ALA DHA

mo

l%

0

10

20

30

LA AA ALA DHA

mo

l%

LA AA ALA DHA0

4

8

10

6

2

LA AA ALA DHA

mo

l%

0

6

8

10

4

2

*

*

**

lung phospholipid

lung phospholipid

lung phospholipid

lung phospholipid





cftr +/+

cftr -/-N/N?F508/?F508

cftr +/+

cftr -/-cftr +/+

cftr -/-cftr +/+

cftr -/-N/N?F508/?F508N/N?F508/?F508





0

10

20

30

LA AA ALA DHA0

10

20

30

0

10

20

30


0

10

20

30

LA AA ALA DHA0

10

20

30

0

10

20

30

LA AA ALA DHA

mol

%m

ol%

mo

l%m

ol%

mo

l%m

ol%

0

10

20

30

LA AA ALA DHA

mo

l%

0

10

20

30

0

10

20

30


mo

l%

0

10

20

30

LA AA ALA DHA

mo

l%

0

10

20

30

0

10

20

30


mo

l%

LA AA ALA DHA0

4

8

10

6

2

LA AA ALA DHALA AA ALA DHA0

4

8

10

6

2

0

4

8

10

6

2

LA AA ALA DHA

mo

l%

0

6

8

10

4

2


mo

l%

0

6

8

10

4

2

0

6

8

10

4

2

*

*

**

Figure 4a: Relative concentrations of linoleic acid (LA, C18:2n-6), arachidonic acid (AA, C20:4n-6), alpha-linolenic acid (ALA,C18:3n-3) and docosahexaenoic acid (DHA, C22:6n-3) in purified PL extracts of homozygous dF508 mice and cftr -/- mice andtheir littermate controls. Fatty acid concentrations are expressed as mol% of total fatty acids. Data represent means±S.D. of 5-6 mice per group, *p<0.05.

4a


To investigate the possible influence of nutritional status on EFA levels, EFA molar

percentages were related to bodyweight for each individual mouse (Figure 4b).

Neither in cftr -/-CAM nor in dF508/dF508 mice or their corresponding controls, signifi-

cant correlations between relative EFA or LCPUFA concentrations and body weight

could be identified in pancreas (Figure 4b), lung or intestinal PL (data not shown).

In addition to relative fatty acid concentrations, absolute concentrations in tissues

were determined and expressed per mg protein. Figure 4c shows that absolute LA,

ALA, AA and DHA contents were similar in pancreas tissue of the two CF mouse

models compared to controls. Similarly, absolute fatty acid concentrations in lung or

jejunum were not significantly different between CF and control mice (not shown).

101


mol

% fa

tty a

cid

mol

% fa

tty a

cid

LA (18:2n-6) pancreas PL

0

10

20

30

0 10 20 30 40bodyweight (g)

ALA (18:3n-3) pancreas PL

0.00.20.40.60.81.0


mol

% fa

tty a

cid

DHA (22:6n-3) pancreas PL

012345


mol

% fa

tty a

cid

AA (20:4n-6) pancreas PL

0

5

10

15


+/+ -/-

mol

% fa

tty a

cid

mol

% fa

tty a

cid


0

10

20

30



0

10

20

30

0

10

20

0

10

20

30

0 10 20 30 400 10 20 30 40bodyweight (g)


0.00.20.40.60.81.0


mol

% fa

tty a

cid


0.00.20.40.60.81.0

0.00.20.40.60.81.0

0 10 20 30 400 10 20 30 4010 20 30 40bodyweight (g)

mol

% fa

tty a

cid


012345


mol

% fa

tty a

cid


012345

012345

0 10 20 30 400 10 20 30 40bodyweight (g)

mol

% fa

tty a

cid


0

5

10

15



0

5

10

15

0

5

10

15

0 10 20 30 400 10 20 30 40bodyweight (g)

+/+ -/-+/++/+ -/--/-Figure 4b: Relative concentrations (mol%) of linoleic acid (LA, C18:2n-6), arachidonic acid (AA, C20:4n-6), alpha-linolenic acid(ALA, C18:3n-3) and docosahexaenoic acid (DHA, C22:6n-3) in pancreas PL related to bodyweight of homozygous dF508 mice(open circles) and cftr-/- mice (open squares) and littermate controls (N/N, closed circles; cftr+/+, closed squares). No correla-tions were detected between body weight and fatty acid concentrations. Data represent means±S.D. of 5-6 mice per group.

Figure 4c: Absolute concentrations (nmol) of LA,AA, ALA and DHA acid in pancreas PL per mg pan-creas protein for homozygous dF508 mice and cftr-

/- mice and littermate controls. No significant differ-ences in absolute fatty acid concentrations per mgprotein were detected between homozygous dF508mice and cftr-/- mice and their littermate controls.Data represent means±S.D. of 5-6 mice per group.

4b

nmol

/ mg

prot

ein

nmol

/ mg

prot

ein

nmol

/ mg

prot

ein

nmol

/ mg

prot

ein

AA pancreas per mg protein

0

10

20

30

?F508N/N cftr +/+ cftr -/-

DHA pancreas per mg protein

0

5

10


LA pancreas per mg protein

0

50

100


ALA pancreas per mg protein

0

1

2


nmol

/ mg

prot

ein

nmol

/ mg

prot

ein

nmol

/ mg

prot

ein

nmol

/ mg

prot

ein

nmol

/ mg

prot

ein

nmol

/ mg

prot

ein

nmol

/ mg

prot

ein

nmol

/ mg

prot

ein


0

10

20

30



0

10

20

30

0

10

20

30

?F508N/N cftr +/+cftr +/+ cftr -/-cftr -/-

DHA pancreas per mg protein

0

5

10

?F508N/N cftr +/+ cftr -/-0

5

10

0

5

0

5

10



0

50

100



0

50

100

0

50

100



0

1

2



0

1

2


4c


Membrane EFA concentrations appeared unaffected in the presently used CF mouse

models compared to littermate controls. Yet, in 1-month-old cftr -/-UNC mice weaned on

a liquid diet (Peptamen), a profound membrane lipid imbalance was reported(13). This

suggests that dietary composition, caloric intake and/or age affect EFA status. To test

this hypothesis, fatty acid profiles were determined in tissues of 1-month-old cftr -/-CAM

and cftr +/+CAM mice, after weaning on Peptamen liquid diet for 7 days. Furthermore,

to assess the role of genetic background, EFA profiles were determined in tissues of

age-matched, liquid diet-weaned C57Bl/6 wildtype mice.

Analysis of fatty acid composition of chow pellets and Peptamen revealed relatively

small differences in EFA and LCPUFA contents, with Peptamen containing less AA

and more DHA than standard chow (Figure 5). Although Peptamen is frequently used

in CF mouse models to prevent intestinal obstruction and to improve nutritional

status, 1-month-old cftr -/-CAM mice weaned on Peptamen still had a significantly lower

body mass than their cftr +/+CAM littermates (11.4±2.2g vs. 13.6±1.1g, p<0.05).

PUFA concentrations in pancreas, lung and jejunum of 1-month-old, Peptamen-

weaned cftr -/-CAM and cftr +/+CAM mice significantly differed from those of adult chow-fed

cftr -/-CAM and cftr +/+CAM mice (Figure 6a; age effect). When compared at the same age,

however, LA, ALA, AA and DHA concentrations in pancreas, lung, intestine and

plasma PL were not significantly different between cftr -/-CAM mice and littermates,

neither at the age of 1 month after Peptamen-weaning (Figure 6b), nor at adult age

during chow-feeding. Yet, AA and DHA concentrations were consistently higher than

in age- and diet-matched C57Bl/6 wildtype mice for all tissues studied (p<0.01).

Similarly, LA concentrations in pancreas, lung and jejunum PL of cftr -/-CAM and

cftr +/+CAM mice were significantly higher than in C57Bl/6 wildtype controls, but not sig-

nificantly different between cftr -/-CAM and cftr +/+CAM mice. ALA levels were low in all

102

Chapter 5

fatty acid composition mouse diets

C22:00.0

0.2

0.4

0.6

0.8

1.0

C20:4n-6 C22:6n-3

* #

*C20:0 C22:5n-3

*

chowpeptamen liquid diet

mol

% o

f tot

al fa

tty a

cids

C16:00

10

20

30

40

50

C18:2n-6

*

C18:0 C18:1n-9

#

C18:3n-3

*


C22:00.0

0.2

0.4

0.6

0.8

1.0

C20:4n-6 C22:6n-3

* #

*C20:0 C22:5n-3

*


mol

% o

f tot

al fa

tty a

cids

C16:00

10

20

30

40

50

C18:2n-6

*

C18:0 C18:1n-9

#

C18:3n-3


C22:00.0

0.2

0.4

0.6

0.8

1.0

C20:4n-6 C22:6n-3

* #

*C20:0 C22:5n-3

*C22:0

0.0

0.2

0.4

0.6

0.8

1.0

0.0

0.2

0.4

0.6

0.8

1.0

C20:4n-6C20:4n-6 C22:6n-3C22:6n-3

* #

*C20:0C20:0 C22:5n-3

*C22:5n-3*

C22:5n-3*


mol

% o

f tot

al fa

tty a

cids

C16:00

10

20

30

40

50

C18:2n-6

*

C18:0 C18:1n-9

#

C18:3n-3

chowpeptamen liquid dietchowchowpeptamen liquid dietpeptamen liquid diet

mol

% o

f tot

al fa

tty a

cids

C16:00

10

20

30

40

50

0

10

20

30

40

50

C18:2n-6

*

C18:2n-6

*

C18:0C18:0 C18:1n-9

#

C18:1n-9

#

C18:3n-3C18:3n-3

*

Figure 5: Fatty acid composition of standard laboratory chow and Peptamen elemental liquid diet. Data represent means ±S.D. of triple aliquot analyses of each diet. *p<0.05 for linoleic acid (C18:2n-6), arachidonic acid (C20:4n-6), eicosapentaenoicacid (C22:5n-3) and docosahexaenoic acid (C22:6n-3) and #p<0.001 for oleic acid (C18:1n-9) and behenic acid (C22:0).

5


103


mo

l%

0

2

4

6

8

10 lung phospholipid

LA AA ALA DHA

*

mo

l%

0

5

10

15

20

25 pancreasphospholipid

cftr +/+ Peptamencftr -/- Peptamencftr +/+chowcftr -/- chow

LA AA ALA DHA

*

mo

l%

0

10

20

30 jejunum phospholipid

LA AA ALA

**

DHA

*

mo

l%

0

2

4

6

8

10

0

2

4

6

8

10 lung phospholipid

LA AA ALA DHA

*

LA AA ALA DHA

*

mo

l%

0

5

10

15

20

25

0

5

10

15

20

25 pancreasphospholipid



LA AA ALA DHA

*

LA AA ALA DHA

**

mo

l%

0

10

20

30

0

10

20

30 jejunum phospholipid

LA AA ALA

**

DHA

*

LA AA ALA

**

DHA

**

Figure 6a: Relative concentrations of linoleic acid (LA, C18:2n-6), arachidonic acid (AA, C20:4n-6), alpha-linolenic acid (ALA,C18:3n-3) and docosahexaenoic acid (DHA, C22:6n-3) in purified PL extracts of pancreas, lung and intestine of 1-month-oldPeptamen-fed cftr+/+ mice (light grey bars) and cftr-/- mice (white bars), adult chow-fed cftr+/+ mice (black bars) and cftr-/- mice(dark grey bars). Individual fatty acid concentrations are expressed as mol% of total fatty acids. Data represent means±S.D. of6 mice per group. *p<0.05 for DHA of pancreas and jejunum PL, for LA of lung and jejunum PL and for AA of jejunum PL fromPeptamen-fed cftr+/+ and cftr-/- mice compared to adult chow-fed cftr+/+ and cftr-/- mice. No significant differences were detectedbetween cftr+/+ and cftr-/- littermates.

6a

tissues analyzed, and although there was a tendency for lower ALA values in

wildtype C57Bl/6 mice compared to cftr -/-CAM mice, this only reached significance

in plasma and ileum (p<0.05 each). Similarly, fatty acid analyses of erythrocyte-

and ileum phospholipids revealed consistently different fatty acid concentrations

in C57Bl/6 mice when compared to cftr -/-CAM and cftr +/+CAM mice (data not shown).

Other LCPUFA of the n-3 and n-6 series (C20:5n-3, C22:5n-3, C22:4n-6) also only

differed between C57Bl/6 mice on one hand, and cftr -/-CAM and cftr +/+CAM mice on

the other hand, and not between cftr -/-CAM mice and cftr +/+CAM littermates (not

shown).


DISCUSSION

We aimed to establish whether perturbed EFA metabolism and altered membrane

EFA composition in CF-affected organs are inextricably linked to CF. Our present

study in two murine models for CF shows no disturbance in EFA metabolism nor in

membrane fatty acid composition, indicating that a membrane EFA imbalance is not

an intrinsic characteristic of CF genotype in mice. By inference, our data indicate that

the altered EFA compositions reported in CF are a secondary phenomenon,

possibly related to inflammation or malnutrition.

Freedman et al.(13) reported markedly increased membrane AA- and decreased DHA-

concentrations in CF-affected organs of a subset of cftr -/-UNC mice compared to non-

littermate C57Bl/6 controls. Oral supplementation with DHA, but not with its

precursor ALA, corrected this membrane EFA imbalance and was reported to

alleviate certain phenotypic manifestations of the disease. The authors suggested a

causative relation between impaired capacity for conversion of EFA into LCPUFA and

CF symptoms. Using 13C-labeled EFA, we quantified rates of EFA elongation, desat-

uration and tissue incorporation in vivo in cftr -/-CAM and cftr +/+CAM mice. After intra-

gastric administration of 13C-EFA, levels of 13C-AA and 13C-DHA in jejunum, pancreas

104

Chapter 5

mo

l%m

ol%

LA AA ALA DHA

0

2

4

6

8

10

LA AA ALA DHA

lung phospholipid

# ##

0

5

10

15

20

25

cftr +/+

cftr -/-C57/Bl/6


*

***

mo

l%m

ol%

mo

l%m

ol%

LA AA ALA DHA

0

2

4

6

8

10

LA AA ALA DHA

lung phospholipid

# ##


0

2

4

6

8

10

0

2

4

6

8

10


lung phospholipid

# ##

0

5

10

15

20

25

cftr +/+

cftr -/-C57/Bl/6


*

***

0

5

10

15

20

25

0

5

10

15

20

25

cftr +/+

cftr -/-C57/Bl/6

cftr +/+

cftr -/-C57/Bl/6


*

***

6bm

ol%

0

10

20

30

LA AA ALA DHA


#

##

LA AA ALA DHA02468

1012

mo

l%

plasma phospholipid

*

**

*

mo

l%m

ol%

0

10

20

30

LA AA ALA DHA


#

##

0

10

20

30



#

##

LA AA ALA DHA02468

1012

mo

l%

plasma phospholipid

*

**

*

LA AA ALA DHA02468

1012

02468

1012

mo

l%

plasma phospholipid

*

**

*

Figure 6b: Relative concentrations of LA (C18:2n-6), AA (C20:4n-6), ALA (C18:3n-3) and DHA (C22:6n-3) in purified PL extractsof pancreas, lung and intestine of cftr+/+ mice (grey bars), cftr-/- mice (white bars) and wildtype C57/Bl/6/129 mice (black bars).All mice were weaned on Peptamen liquid diet from post-natal day 23 and fatty acid analyses were performed at post-natal day30. Fatty acid concentrations are expressed as mol% of total fatty acids. Data represent means±SD of 6 mice per group.*p<0.05 for LA and DHA and **p<0.005 for AA in pancreas PL of wildtype C57/Bl/6/129 mice vs. cftr+/+ and cftr-/- mice.#p<0.001 for LA, AA and DHA in lung and intestinal PL of C57/Bl/6/129 mice vs.to cftr+/+ and cftr-/- mice. No significantdifferences were detected between cftr+/+ and cftr-/- mice.


and liver PL were equal in cftr -/-CAM and cftr +/+CAM mice, indicating that in this CF mouse

model, EFA elongation and desaturation is unaffected and that impaired LCPUFA

synthesis is no inextricable feature of CF phenotype.

The phenotypic manifestations of CF are highly variable in patients as well as in

different murine models for CF. We analyzed EFA status in two CF mouse models:

homozygous dF508 mice with the dF508 exon 10 insertional mutation(18), expressing

a mild phenotype without fat malabsorption, and cftr -/-CAM (University of Cambridge)

mice, in which exon 10 replacement results in complete absence of CFTR activity

and a severe gastrointestinal phenotype, including fat malabsorption(17). For

comparison, we used sex-matched littermates as controls. Quantification of fecal

fatty acid excretion demonstrated that cftr -/-CAM mice indeed malabsorbed dietary fatty

acids. Yet, in neither of the two murine CF models we found indications for major

membrane EFA alterations in CF-affected organs as compared to littermate controls.

The slight and inconsistent alterations of AA levels that we did measure in lung and

jejunum, and the marginally decreased DHA levels in pancreas, were found only in

dF508/dF508 mice and not in cftr -/-CAM mice, despite the fact that the more severe

phenotype of cftr -/-CAM mice would be expected to correlate with a higher incidence of

membrane lipid imbalances(29). Only when cftr -/-CAM and cftr +/+CAM mice were compared

to wildtype controls of different (C57Bl/6) genetic background, pronounced

differences in membrane fatty acid composition became apparent.

The discrepancies between our observations and those of Freedman et al. are

unlikely to be explained by differences in preparative steps prior to GC injection. The

discrepancies could however be related to the age difference between mice in our

study (3 months) and in Freedman's study (1 month)(13). In contrast to the chow-fed

adult mice used by us, 1-month-old liquid-diet weaned cftr -/-UNC mice displayed a

profound lipid imbalance in CF-affected tissues, compared to C57Bl/6 mice.

Theoretically, a conditional essentiality of dietary LCPUFA during early life may result

in transiently low LCPUFA levels in young mice, which may resolve when EFA

metabolizing capacity reaches maturity at adult age. Young cftr -/- mice might be more

vulnerable than wildtype controls for such a transient deficiency of LCPUFA due to

impaired fat absorption in CF. However, comparison of membrane fatty acids of

1-month-old, liquid diet-fed cftr +/+CAM and cftr -/-CAM mice with those of 3-month-old,

chow-fed mice indicated that the former actually had higher relative levels of EFA and

LCPUFA. Similar to the 3-month-old mice, no differences in fatty acid composition

were detected between 1-month-old cftr -/-CAM mice and cftr +/+CAM littermates,

suggesting that differences in fatty acid levels between 1- and 3-month-old mice is

more likely related to the different diets, or to an age-dependent effect unrelated to

CFTR malfunction.

105



The different diets fed to cftr -/-UNC mice and cftr -/-CAM mice could theoretically account

for the inconsistency regarding EFA levels in these two models. Both cftr -/-CAM mice

and cftr -/-UNC mice display a severe phenotype characterized by fat malabsorption,

goblet cell hyperplasia and failure to thrive, although cftr -/-UNC mice are more

severely affected. When weaned on a chow-based diet, mortality due to intestinal

obstruction is considerable in cftr -/-UNC mice during the first weeks of life. Weaning on

a complete elemental liquid diet, such as Peptamen, significantly improves survival

rates, but CF mice fed Peptamen remain considerably smaller when compared to

normal littermates. To meet daily caloric needs, adult mice have to consume up to 15

ml of Peptamen per day(28), and lower intake may result in malnutrition. Striking

similarities have been described between Peptamen-fed cftr -/-UNC mice and a mal-

nourished CF mouse model regarding pulmonary cytokine profiles(30), suggesting

that malnutrition secondary to liquid diet feeding may contribute to symptoms in

Peptamen-fed CF mice(29). Relative EFA concentrations differ only slightly between

chow and Peptamen, with Peptamen containing relatively less AA and more DHA

than solid chow. Cftr -/-UNC mice fed Peptamen, however, had high levels of AA and low

levels of DHA, which makes the fatty acid composition of the liquid diet an unlikely

contributor to the observed membrane EFA imbalance in these mice.

Yet, quantitative absorption studies would be required to fully exclude differences in

net enteral uptake of EFA from chow or from Peptamen.

The discrepancy between our results and those of Freedman et al. may also be due

to variations inherent to the use of different mouse models for CF. To date, over 10

different murine CF models have been characterized, which can be categorized into

mutants in which CFTR expression is simply disrupted (i.e., cftr -/-1HGU, cftr -/-HSC,

cftr -/-BAY, cftr -/-UNC and cftr -/-CAM mice(15-17; 31; 32)) and mutants that model specific clinical

mutations such as the dF508 mutation in cftr -/-EUR and cftr -/-1KTH mice(18; 33). Within the

group with CFTR gene disruption, the potential to produce CFTR mRNA ranges from

no detectable CFTR mRNA in absolute null mice (cftr -/-CAM, cftr -/-CAM, cftr -/-HSC) to

mutants in which up to 10% of CFTR mRNA production is retained (cftr -/-1HGU).

Generally, mice with lowest residual CFTR activity display the most severe pheno-

type, but phenotypic differences can also result from the different genetic back-

grounds onto which CFTR mutations have been introduced. The UNC mutation has

been crossed into three different strains, i.e., C57Bl/6/129, B6D2/129 and

BALB/C/129 mice, while the CAM mutation has been outcrossed to a C57Bl/6/129

population. Whereas we used sex-matched littermates as controls for cftr -/-CAM mice

to evaluate EFA status, Freedman et al. used non-littermate, wildtype C57Bl/6 mice.

Our present data indicate that genetic background and age have an overriding effect

106

Chapter 5


on EFA status in general and on DHA and AA levels in particular, so any meaningful

comparisons of EFA status between CF mice and controls should take these

confounding factors into account.

In addition to the specific type of CFTR mutation and to environmental influences,

phenotypic variability between CF patients and mouse models is thought to be

related to independently segregating disease-modifying genes. Proteins encoded by

genes other than the CFTR gene may partially substitute for mutant CFTR, and indi-

vidual variability in levels of tissue expression and functional activity for these other

proteins may explain the inter-individual phenotypic differences between patients or

mice with identical CFTR mutations(34; 35). Several candidate modifier genes have been

postulated to account for the wide spectrum of lung disease severity in patients

homozygous for the dF508 mutation(36). Rozmahel et al. demonstrated in mice the

presence of a CFTR-independent locus that modulated severity of gastrointestinal

disease(32), and Zielenski et al. identified a similar modifier gene for meconium ileus

on human chromosome 19(37). Similarly, the expression of liver disease has been

described to be modulated by independently inherited modifier genes. This again

underlines the prerequisite of using littermate controls in murine models for CF.

Our findings of normal membrane fatty acid composition in two CF mouse models

correspond to results described by Dombrowsky et al., who found normal levels of

DHA and even decreased levels of AA in phospholipid species of standard diet-fed

adult cftr -/-1HGU mice(38). As in our study, the differences in EFA levels were very small,

and inconsistent between phospholipid classes. The fact that HGU mice have 10%

residual CFTR mRNA makes conclusions regarding the role of CFTR in EFA

metabolism difficult, yet, both Dombrowsky's and our results underline the variability

in membrane PL composition between different CF mouse models. Strandvik et al.

described essential fatty acid deficiency in plasma phospholipids of CF patients(7),

but differences were small and AA levels were normal in all patients. The most

pronounced EFA alterations were found in patients with severe mutations (i.e., dF508

and 394delTT), and although no correlations were reported with other genotypes,

a relation with fat malabsorption cannot be excluded.

In summary, from in vivo analyses of LCPUFA synthesis in a mouse model for CF, we

conclude that impaired LCPUFA synthesis or imbalanced membrane fatty acid

composition are no inextricable features of CF phenotype. Fat malabsorption does

not have a strong effect on EFA status in CF mice. Extrapolating these conclusions

to CF patients may implicate that sufficient oral EFA intake could effectively prevent

107



compromised EFA status in CF. For studying essential fatty acid metabolism in

murine CF models and inferring observations to the human condition, meticulous

verification of mouse background strains and the use of littermate controls is of

crucial importance.

AcknowledgementsThe authors would like to thank Rick Havinga, Henk Elzinga and Ingrid Martini for

their technical expertise and assistance in the experiments described in this article

and Dr. Frans Stellaard for his help with the stable isotope studies.

REFERENCES1. Lloyd-Still JD, Johnson SB, Holman RT. EFA status in CF and effects of safflower oil supplementation. AJCN 1981; 34: 1-7.2. Biggemann B, Laryea MD, Schuster A, Griese M, Reinhardt D, Bremer HJ. Status of plasma and erythrocyte fatty acids andvitamin A and E in young children with CF. Scand J Gastroenterol 1988;.Suppl 143:135-41: 135-141.3. Strandvik B.. Relation between EFA metabolism and gastrointestinal symptoms in CF. Acta Paediatr Scand 1989; Suppl363:58-65.4. Lepage G, Levy E, Ronco N, Smith L, Galeano N, Roy CC. Direct transesterification of plasma fatty acids for the diagnosisof essential fatty acid deficiency in CF. J Lipid Res 1989; 30: 1483-1490.5. Parsons HG, O'Loughlin EV, Forbes D, Cooper D, Gall DG. . Supplemental calories improve EFA deficiency in CF patients.Pediatr Res 1988; 24: 353-356.6. Roulet M, Frascarolo P, Rappaz I, Pilet M. EFA deficiency in well nourished CF patients. Eur J Pediatr 1997;156:952-956.7. Strandvik B, Gronowitz E, Enlund F, Martinsson T, Wahlstrom J. Essential fatty acid deficiency in relation to genotype inpatients with CF. J Pediatr 2001; 139: 650-655.8. Benabdeslam H, Garcia I, Bellon G, Gilly R, Revol A. Biochemical assessment of the nutritional status of CF patients treatedwith pancreatic enzyme extracts. Am J Clin Nutr 1998; 67: 912-918.9. Portal BC, Richard MJ, Faure HS, Hadjian AJ, Favier AE. Altered antioxidant status and increased lipid peroxidation in chil-dren with CF. Am J Clin Nutr 1995; 61: 843-847.10. Rogiers V, Dab I, Michotte Y, Vercruysse A, Crokaert R, Vis HL. Abnormal fatty acid turnover in the phospholipids of the redblood cell membranes of CF patients. Pediatr Res 1984; 18: 704-709.11. Gilljam H, Strandvik B, Ellin A, Wiman LG. Increased mole fraction of arachidonic acid in bronchial phospholipids in patientswith CF. Scand J Clin Lab Invest 1986; 46: 511-518.12. Bhura-Bandali FN, Suh M, Man SF, Clandinin MT. The dF508 mutation in the cystic fibrosis transmembrane conductanceregulator alters control of essential fatty acid utilization in epithelial cells. J Nutr 2000; 130: 2870-2875.13. Freedman SD, Katz MH, Parker EM, Laposata M, Urman MY, Alvarez JG. A membrane lipid imbalance plays a role in thephenotypic expression of CF in cftr -/- mice. Proc Natl Acad Sci U S A 1999; 96: 13995-14000.14. Freedman SD, Blanco PG, Zaman MM, Shea JC, Ollero M, Hopper IK, Weed DA, Gelrud A, Regan MM, Laposata M, AlvarezJG, O'Sullivan BP. Association of CF with abnormalities in fatty acid metabolism. N.Engl.J.Med. 2004; 350: 560-569.15. Snouwaert JN, Brigman KK, Latour AM, Malouf NN, Boucher RC, Smithies O, Koller BH. An animal model for CF made bygene targeting. Science 1992; 257: 1083-1088.16. O'Neal WK, Hasty P, McCray PB, Jr., Casey B, Rivera-Perez J, Welsh MJ, Beaudet AL, Bradley A. A severe phenotype inmice with a duplication of exon 3 in the CF locus. Hum Mol Genet 1993; 2: 1561-1569.17. Ratcliff R, Evans MJ, Cuthbert AW, MacVinish LJ, Foster D, Anderson JR, Colledge WH. Production of a severe CF mutationin mice by gene targeting. Nat Genet 1993; 4: 35-41.18. van Doorninck JH, French PJ, Verbeek E, Peters RH, Morreau H, Bijman J, Scholte BJ. A mouse model for the CF deltaF508 mutation. EMBO J 1995; 14: 4403-4411.19. Bijvelds M, Hulzebos C, Bronsveld H, Havinga R, Stellaard F, Sinaasappel M, De Jonge H, Verkade HJ. Digestive DiseaseWeek 2003. Orlando, Florida, USA. Abstracts. Gastroenterology 124: A1-102.20. Kerem B, Rommens JM, Buchanan JA, Markiewicz D, Cox TK, Chakravarti A, Buchwald M, Tsui LC. Identification of the CFgene: genetic analysis. Science 1989; 245: 1073-1080.21. Muskiet FA, van Doormaal JJ, Martini IA, Wolthers BG, van der Slik W. Capillary gas chromatographic profiling of total long-chain fatty acids and cholesterol in biological materials. J Chromatogr 1983; 278: 231-244.

108

Chapter 5


22. Bruzzone R, Halban PA, Gjinovci A, Trimble ER. A new, rapid, method for preparation of dispersed pancreatic acini. BiochemJ 1985; 226: 621-624.23. Bligh EG, Dyer WJ. A rapid method for total lipid extraction and purification. Can J Biochem Physiol 1959; 37: 911-917.24. Werner A, Minich DM, Havinga R, Bloks V, Van Goor H, Kuipers F, Verkade HJ. Fat malabsorption in EFA-deficient mice isnot due to impaired bile formation. Am J Physiol Gastrointest Liver Physiol 2002; 283: G900-G908.25. Lowry OH, Rosebrough NJ, Farr AL, Randal RJ. Protein measurement with the Folin phenol reagent. JBS 1951;193:265-75.26. Kang JX, Man SF, Brown NE, Labrecque PA, Clandinin MT. The chloride channel blocker anthracene 9-carboxylate inhibitsfatty acid incorporation into phospholipid in cultured human airway epithelial cells. Biochem J 1992; 285: 725-729.27. Freedman SD, Blanco PG, Shea JC, Alvarez JG. Analysis of lipid abnormalities in CF mice. Methods Mol Med 2002; 70:517-24.: 517-524.28. Eckman EA, Cotton CU, Kube DM, Davis PB. Dietary changes improve survival of CFTR S489X homozygous mutant mice.Am J Physiol 1995; 269:L625-30.29. Davidson DJ, Rolfe M. Mouse models of CF. Trends Genet 2001; 17:S29-37.30. Yu H, Nasr SZ, Deretic V. Innate lung defenses and compromised Pseudomonas Aeruginosa clearance in the malnourishedmouse model of respiratory infections in CF. Infect Immun 2000; 68: 2142-2147.31. Dorin JR, Dickinson P, Alton EW, Smith SN, Geddes DM, Stevenson BJ, Kimber WL, Fleming S, Clarke AR, Hooper ML. CFin the mouse by targeted insertional mutagenesis. Nature 1992; 359: 211-21532. Rozmahel R, Wilschanski M, Matin A, Plyte S, Oliver M, Auerbach W, Moore A, Forstner J, Durie P, Nadeau J, Bear C, TsuiLC. Modulation of disease severity in CFTR deficient mice by a secondary genetic factor. Nat Genet 1996; 12: 280-287.33. Zeiher BG, Eichwald E, Zabner J, Smith JJ, Puga AP, McCray PB, Jr., Capecchi MR, Welsh MJ, Thomas KR. A mouse modelfor the dF508 allele of CF. J Clin Invest 1995; 96: 2051-2064.34. Drumm ML. 2001. Modifier genes and variation in CF. Respir Res 2: 125-128.35. Zielenski J. Genotype and phenotype in CF. Respiration 2000; 67: 117-133.36. Merlo CA, Boyle MP. Modifier genes in CF lung disease. J Lab Clin Med 2003; 141: 237-241.37. Zielinski J, Corey M, Rozmahel R, Markiewicz D, Aznarez I, Casals T, Larriba S, Mercier B, Cutting GR, Krebsova A, MacekM, Langfelder-Schwind E, Marshall B, De Celie-Germana J, Claustres M, Palecio A, Bal J, Nowakowska A, Ferec C, Estivill X,Durie P, Tsui LC. Detection of CF modifier locus for meconium ileus on human chromosome 19q13. Nat Genet 1999;22:128-938. Dombrowsky H, Clark GT, Rau GA, Bernhard W, Postle AD. Molecular species compositions of lung and pancreas phos-pholipids in the cftr(tm1HGU/tm1HGU) CF mouse. Pediatr Res 2003; 53: 447-454.

109



110


A. WernerR. Havinga F. Kuipers

H.J. Verkade

Treatment of essential fatty aciddeficiency with dietary triglyceridesor phospholipids in a murine model

of extrahepatic cholestasis6Am J Physiol Gastrointest Liver Physiol. 2004; 286(5): G822-832


ABSTRACT

Background: Essential fatty acid (EFA) deficiency during cholestasis is mainly due to

malabsorption of dietary EFA(1). Theoretically, dietary phospholipids (PL) may have a

higher bioavailability than dietary triglycerides (TG) during cholestasis.

We developed murine models for EFA deficiency with and without extrahepatic

cholestasis, and compared the efficacy of oral supplementation of EFA as PL or TG.

Methods: EFA deficiency was induced in mice by feeding a high-fat EFA-deficient

(EFAD) diet. After three weeks on this diet, bile duct ligation (BDL) was performed in

a subgroup of mice to establish extrahepatic cholestasis. Cholestatic and non-

cholestatic EFA-deficient mice continued on the EFAD diet (controls), or were

supplemented for three weeks with EFA-rich TG or EFA-rich PL. Fatty acid

composition was determined in plasma, erythrocytes, liver and brain.

Results: After four weeks of EFAD diet, induction of EFA deficiency was confirmed by

a six-fold increased triene/tetraene ratio (T/T-ratio) in erythrocytes of non-cholestatic

and cholestatic mice (p<0.001). EFA-rich TG and EFA-rich PL were equally effective

in preventing further increase of the erythrocyte T/T-ratio, which was observed in

cholestatic and non-cholestatic non-supplemented mice (12- and 16-fold the initial

value, respectively). In cholestatic mice, EFA-rich PL was superior to EFA-rich TG in

decreasing T/T-ratios of liver triglycerides and phospholipids (each p<0.05), and in

increasing brain phospholipid concentrations of the long-chain polyunsaturated fatty

acids (LCPUFA) docosahexaenoic acid and arachidonic acid (each p<0.05).

Conclusions: Oral EFA supplementation in the form of PL is more effective than in the

form of TG in increasing LCPUFA concentrations in liver and brain of cholestatic

EFA-deficient mice.

112

Chapter 6


113

Treatment of EFA deficiency with dietary TG or PL in a murine model of extrahepatic cholestasis

INTRODUCTION

Essential fatty acids (EFA) and their long-chain polyunsaturated metabolites have

been recognized to play a role in growth, development of the central nervous

system, eicosanoid production and control of lipid homeostasis. Due to the inability

of mammalian cells to synthesize EFA de novo, adequate EFA levels depend

entirely on sufficient dietary intake and absorption. Conditions leading to fat mal-

absorption, such as cholestasis, have been associated with a high incidence of

essential fatty acid deficiency (EFAD)(2-4). In a rat model of cholestasis, we recently

demonstrated that impaired intestinal absorption of EFA is the main contributor to

EFA deficiency in cholestatic conditions, rather than altered post-absorptive EFA

metabolism(1). So far, no adequate oral strategies for prevention or treatment of EFA

deficiency during cholestasis have been developed. In regular diets, 90% of EFA is

present as acyl esters in triglycerides, and only 10% is present as acyl esters in phos-

pholipids and cholesterol esters(5). Under physiological conditions, however, the

intestine receives significant amounts of EFA as biliary phospholipids, which contain

up to 40 mol% of EFA (mostly linoleic acid), esterified at the sn-2 position.

Triglycerides and phospholipids have different intestinal absorption mechanisms.

Due to their hydrophobic nature, triglycerides are insoluble in the aqueous intestinal

lumen, and products of triglyceride lipolysis, i.e., free fatty acids and mono-

glycerides, highly depend on bile components for solubilization into mixed micelles.

After absorption by the enterocyte, fatty acids and monoglycerides are re-esterified

into triglycerides and secreted in lymph as the major core components of chylo-

microns. Phospholipids, on the other hand, are relatively independent of bile

components for intestinal absorption. Phospholipids have a higher tendency than

triglycerides to interact with water and can associate into liquid crystals (bilayers)(6),

which have been suggested to play a role in luminal lipid solubilization under bile-

deficient conditions(7). Phospholipids can be absorbed intact or after partial digestion

to lysophospholipids and free fatty acids(8-10). Phospholipids of luminal origin are pre-

dominantly used by enterocytes for assembly of the surface coat of chylomicrons

secreted into lymph. In conditions of cholestasis, i.e. during decreased or absent bile

formation, triglyceride absorption is severely impaired. Phospholipids can facilitate

dietary lipid absorption under bile deficient conditions by enhancing luminal lipid

solubilization and by providing surface components for lipoprotein assembly(6;11).

Additionally, phospholipids have been postulated to have greater post-absorptive

bioavailability(12-16). Based on these characteristics, we hypothesized that phospho-

lipids could be a better vehicle for oral EFA supplementation during cholestasis than

triglycerides. In this study, we compared the efficacy of EFA-rich PL and of EFA-rich


TG for treatment of EFA deficiency under cholestatic conditions in mice. For this pur-

pose, we developed a model for both EFA deficiency and cholestasis by feeding

mice an EFA-deficient (EFAD) diet followed by bile duct ligation. Subsequent oral

supplementation with EFA either as TG or PL demonstrated that the latter resulted in

higher concentrations of EFA-derived LCPUFA in target organs as brain and liver.


AnimalsWildtype mice with a free virus breed (FVB) background were obtained from Harlan

(Horst, the Netherlands). Male mice (bodyweight 25-35 g) were housed in a light-

controlled (lights on 6 AM-6 PM) and temperature-controlled (21°C) facility and were

allowed tap water and chow (Hope Farms B.V. 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.

Experimental dietsThe EFA-deficient (EFAD) diet contained 20 energy% protein, 46 energy% carbo-

hydrate and 34 energy% fat, respectively, and had the following fatty acid composi-

tion: 41.4 mol% palmitic acid (C16:0), 47.9 mol% stearic acid (C18:0), 7.7 mol% oleic

acid (C18:1n-9) and 3 mol% linoleic acid (C18:2n-6). An isocaloric EFA-sufficient

(EFAS) diet was used as control diet, containing 20 energy% protein, 43 energy%

carbohydrate and 37 energy% fat with 32.1 mol% C16:0, 5.5% C18:0, 32.2 mol%

C18:1n-9 and 30.2% C18:2n-6 (custom synthesis, diet numbers 4141.08 (EFAD) and

4141.07 (EFAS), Hope Farms BV, Woerden, the Netherlands). For EFA supplementa-

tion, EFAD chow pellets were finely pulverized and homogeneously mixed with either

triglyceride oil or phospholipid oil (lecithin), dissolved in water en ethanol. Both TG

and PL oil were purified from crude soybean oil and had the following fatty acid

profile (weight%): 53.5% C18:2, 7.6% C18:3, 22.8% C18:1, 10.7% C16:0, 4.0% C18:0

and <0.5% of C14:0, C16:1, C20:0, C20:1, C22:0 and C24:0 each (TG); and 56.1%

C18:2, 8.2% 18:3, 13.2% C18:1, 16.1% C16:0, 4.5% C18:0 and <0.5% of C14:0,

C16:1, C20:0, C20:1, C22:0 and C24:0 each (PL), respectively.

We aimed to supplement the EFA-deficient mice with 2.5 mg LA per day. For the TG-

oil, the percentage of fatty acid mass relative to the glycerol backbone mass is

95.5%, of which 53.5% is linoleic acid. For the PL-oil, the fatty acid mass vs. the mass

of the glycerol-phosphate-choline backbone is 72.3%, of which 56.1% is linoleic acid.

To obtain equimolar amounts of LA in the two EFA-supplemented diets, we added

114

Chapter 6


115


5.6 g TG (=0.006 mol TG) per kg chow and 5.8 g PL (=0.007 mol PL), resulting in

0.003 mol LA per kg of chow for both oils. Based on average daily chow intake of 3

g (measured in pilot experiments with EFAD mice), this resulted in 2.5 mg of LA

supplementation per day. The PL and TG soybean oils were a generous gift from

Unimills BV, Zwijndrecht, the Netherlands.

Experimental proceduresInduction of EFA deficiency in mice and oral administration of EFA-rich TG or PL

Mice were fed standard laboratory chow containing 6 weight% fat (RMH-B, Hope

Farms BV, Woerden, the Netherlands) from weaning. Before starting the EFAD diet, a

blood sample was obtained by tail bleeding under halothane anesthesia for determi-

nation of baseline EFA status. Blood was collected in micro-hematocrit tubes

containing heparin, and plasma and erythrocytes were separated by centrifugation

at 2400 rpm for 10 min (Eppendorf Centrifuge, Eppendorf, Germany). Plasma and

erythrocyte samples were hydrolyzed and methylated the same day(17) for gas-

chromatographic analysis of fatty acid profiles. All mice were then fed a high-fat

(16 weight% fat), EFAD diet for four weeks. Subsequently, mice were randomly

assigned to EFAD diet supplemented with either EFA-rich TG or EFA-rich PL, or

continued on the EFAD diet for three weeks (n=5 per group). At weekly intervals

mice were weighed, chow containers were weighed to monitor food intake and blood

samples were taken by tail bleeding for determination of EFA status.

Bile duct ligation in EFAD mice and oral administration of EFA-rich PL or TG

A separate group of mice was fed the EFAD diet for three weeks, after which their bile

ducts were ligated by placing three sutures proximal to the gallbladder under

halothane/NO2 anesthesia. Animals were allowed to recover from surgery for one

week during which the EFAD diet was continued. Subsequently, i.e. after four weeks

on EFAD diet, mice were randomly assigned to EFAD diet enriched with either EFA-

rich TG or EFA-rich PL, or continued on the EFAD diet for three weeks. A separate

control group of non-bile duct ligated control mice received an EFA sufficient (EFAS)

diet for seven weeks (n=6 for each dietary group). During the entire experiment,

bodyweight, chow ingestion, plasma liver enzymes and EFA status were determined

at weekly intervals. After three weeks of supplementation, mice were terminated

under anesthesia by heart puncture, by which a large blood sample (0.6 - 1.0 ml) was

obtained, and liver and brain were removed and stored at -80°C for fatty acid

analysis. A schematic overview of the experimental design is depicted in Figure 1.


Analytical techniquesFatty acid status was analyzed by hydrolyzing, methylating and extracting total

plasma lipids and erythrocyte membrane lipids as described by Muskiet et al.(17). For

fatty acid analysis of brain and liver, tissue samples were mechanically homogenized

in 0.9% NaCl and lipids were extracted from aliquots of tissue homogenate as

described by Bligh and Dyer(18). Lipid extracts were fractionated into phospholipids,

cholesterol esters, triacylglycerols, diacylglycerols, monoacylglycerols and free fatty

acids using thin-layer chromatography (TLC) (20x20 cm, Silica gel 60 F254, Merck),

with hexane/diethyl ether/acetic acid (80:20:1, v/v/v) as solvent. TLC plates were

dried and colored by iodine, PL and TG spots were scraped off and methylated(17).

To account for losses during lipid extraction, heptadecanoic acid (C17:0, Sigma

Chemical Company, St. Louis, MO, USA) was added to all samples as internal

standard prior to extraction. Butylated hydroxytoluene was added as antioxidant.

Chow pellets were freeze-dried and mechanically homogenized and from aliquots of

each diet, lipids were hydrolyzed, methylated and extracted as described above.

Fatty acid methyl esters were separated and quantified by gas liquid chromato-

graphy on a Hewlett Packard gas chromatograph model 6890, equipped with a


detector, using program conditions as described previously(19). Individual fatty acid

methyl esters were quantified by relating areas of their chromatogram peaks to that

116

Chapter 6

EFAD diet

EFAD diet + EFA-rich PL

EFAD diet

EFAD diet + EFA rich TG

bile duct ligation

CHOLESTASIS

wk 1 wk 2 wk 3 wk 4 wk 5 wk 6 wk 7

RBC plasma

liverbrain

RBC plasma

RBCplasma

RBC plasma

RBCplasma

RBCplasma

EFAS diet

EFAD diet


EFAD diet


EFAD diet


EFAD diet


bile duct ligation

CHOLESTASIS

wk 1 wk 2 wk 3 wk 4 wk 5 wk 6 wk 7wk 1 wk 2 wk 3 wk 4 wk 5 wk 6 wk 7

RBC plasma

liverbrain

RBC plasma

RBCplasma

RBC plasma

RBCplasma

RBCplasma

RBC plasma

liverbrain

RBC plasma

RBCplasma

RBC plasma

RBCplasma

RBCplasma

EFAS diet

Figure 1: Experimental design of EFA supplementation to cholestatic EFA-deficient mice. EFA deficiency was induced by feed-ing the EFAD diet for 3 weeks, followed by surgical bile duct ligation. After one week of recovery, mice were assigned to EFADdiet enriched with TG or PL, or continued on the EFAD diet for 3 weeks. A group of non-bile duct ligated control mice receivedan EFA sufficient (EFAS) diet for 7 weeks (n=6 for each dietary group). Blood samples were taken weekly and liver and brainwere removed for fatty acid analysis at the end of the experiment.

1


117


of the internal standard C17:0. Plasma aspartate aminotransferase (AST), alanine

aminotransferase (ALT), alkaline phosphatase (AP) and cholesterol concentrations

were determined using routine clinical procedures. Plasma total bile salt levels were

determined according to Mashige et al.(20). Liver and brain histology was examined on

frozen tissue sections after Oil-Red-O (ORO) staining, which colors neutral lipids

(mainly triglycerides), and on paraformaldehyde-fixed, paraffin-embedded tissue

sections after hematoxylin-eosin (HE) staining by standard procedures.

Calculations and statisticsRelative concentrations (mol%) of plasma, RBC, liver and brain fatty acids were cal-

culated by summation of all fatty acid peak areas and expression of individual fatty

acid areas as a percentage of this amount. EFA status was evaluated by comparing

mol% of individual EFA and LCPUFA, and by calculating markers for EFA deficiency

such as the triene/tetraene-ratio in different body compartments(21). All results are pre-

sented as means ± S.D. for the number of animals indicated. Data were statistically

analyzed using Student's t-test or, for comparison of more than two groups, ANOVA-

test with post-hoc Bonferroni correction. Analyses were performed using SPSS for

Windows software (SPSS, Chicago, IL).

RESULTS

EFA deficiency in non-cholestatic miceIn a previous study(19), we developed a murine model for diet-induced EFA deficiency

by feeding mice an EFAD diet for 8 weeks, which resulted in a pronounced dietary

fat malabsorption. In the present study, feeding the EFAD diet for 4 weeks did not

decrease bodyweight, but mice stopped growing (baseline: 32.4±2.9 g; at 4 weeks:

34.1±3.3 g, NS). After 3 weeks of EFA supplementation, no differences in body-

weight were found between mice fed EFA-rich TG (33.6±3.9 g) or EFA-rich PL

(33.1±1.9 g, NS). Chow intake was 3.5±0.5 g/day in both dietary groups throughout

the experimental period, indicating an average daily LA intake of 3 mg. Since EFA

deficiency has been reported to have species-specific effects on bile formation(19;22),

we measured the effects of dietary EFA deficiency and -supplementation on plasma

bile salts. EFA deficiency was associated with increasing plasma bile salt levels

(21±4 µM at baseline to 40±7 µM after 4 weeks of EFAD diet). Three weeks of sup-

plementation with TG or PL tended to reverse the increased plasma bile salt levels

(32±10 and 38±6 µM, respectively), but differences were not statistically significant.


Figure 2 shows the molar percentages of relevant PUFA in erythrocyte membrane

lipids of non-cholestatic mice during EFAD diet feeding (control), EFA-rich PL

supplementation or EFA-rich TG supplementation.

The classic biochemical parameter describing EFA status is the triene/tetraene-ratio

(T/T-ratio), i.e., the molar ratio between the non-essential fatty acid eicosatrienoic acid

(C20:3n-9) and arachidonic acid (C20:4n-6). Development of EFA deficiency is

associated with an elevated T/T-ratio. Compared to baseline (0.02±0.01), the T/T-

ratio in erythrocytes of non-cholestatic mice increased more than 6-fold (p<0.001)

after four weeks, and 16-fold after another three weeks of EFAD diet (p<0.001, Figure

2a). EFA supplementation, either with EFA-rich PL (0.09±0.02) or EFA-rich TG

(0.09±0.03) completely prevented this further increase in T/T-ratio (p<0.001 for either

PL or TG vs. EFAD) and tended to decrease the ratio compared to four-week values.

Molar percentages of linoleic acid (LA, Figure 2b), and of its corresponding long-

chain polyunsaturated metabolite arachidonic acid (AA, Figure 2c) were decreased

after four weeks on EFAD diet (both p<0.001). After another three weeks on this diet,

LA levels had further decreased (p<0.001) while AA levels remained at a stable low

level. Supplementation with either EFA-rich TG or PL reversed the decreased LA and

AA concentrations with similar efficacy (p<0.005) (Figure 2b, 2c).

In control mice fed the high-fat EFA-sufficient (EFAS) diet for seven weeks, LA and

118

Chapter 6

AA (20:4n-6) LA (18:2n-6) T/T ratio

weeksweeksweeks

#

mo

l% o

f to

tal f

atty

aci

ds

0.0

0.1

0.2

0.3

0.4

0 2 6 84

TG

PL

EFAD

*

* *

0.0

5.0

10.0

15.0

20.0

25.0

0 2 6 84

TG

PL

EFAD

*#

mo

l% o

f to

tal f

atty

aci

ds

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0 2 6 84

TG

PL

EFAD

*

#

* *m

ol%

of t

ota

l fat

ty a

cid

s

AA (20:4n-6) LA (18:2n-6) T/T ratio AA (20:4n-6) LA (18:2n-6) T/T ratio

weeksweeksweeks weeksweeksweeks

#

mo

l% o

f to

tal f

atty

aci

ds

0.0

0.1

0.2

0.3

0.4

0 2 6 84

TG

PL

EFAD

*

* *

mo

l% o

f to

tal f

atty

aci

ds

0.0

0.1

0.2

0.3

0.4

0.0

0.1

0.2

0.3

0.4

0 2 6 840 2 6 84

TG

PL

EFAD

TG

PL

EFAD

*

* *

0.0

5.0

10.0

15.0

20.0

25.0

0 2 6 84

TG

PL

EFAD

*#

mo

l% o

f to

tal f

atty

aci

ds

0.0

5.0

10.0

15.0

20.0

25.0

0.0

5.0

10.0

15.0

20.0

25.0

0 2 6 840 2 6 840 2 6 84

TG

PL

EFAD

TG

PL

EFAD

**##

mo

l% o

f to

tal f

atty

aci

ds

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0 2 6 84

TG

PL

EFAD

*

#

* *m

ol%

of t

ota

l fat

ty a

cid

s

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0.0

2.0

4.0

6.0

8.0

10.0

12.0

0 2 6 840 2 6 84

TG

PL

EFAD

TG

PL

EFAD

**

##

* *m

ol%

of t

ota

l fat

ty a

cid

s

Figure 2: Triene/tetraene ratio (T/T-ratio), linoleic acid (LA) and arachidonic acid (AA) in red blood cells (RBC) of non-cholesta-tic mice fed EFAD diet for 4 weeks, and EFA-rich PL, EFA-rich TG or no EFA for 3 weeks. Fatty acid concentrations areexpressed as mol% of total fatty acids. Data represent means±SD of 5 mice per group. * p<0.001 for week 0 vs. week 4 for allfatty acids, and ** p<0.005 for the TT-ratio and LA in week 4 vs. week 7. # p<0.001 for EFAD vs. PL or TG In week 7. Nosignificant differences were found between PL- or TG-fed mice.

2a 2b 2c


AA levels remained stable (LA: 8.9±0.6, AA 15.7±0.6; not depicted in the figure).

Molar percentages of the n-3 EFA ALA were low in all dietary groups and did not

change significantly during four or seven weeks of EFAD diet, nor during three weeks

of EFA supplementation (data not shown).

EFA deficiency in cholestatic miceUnder physiological (non-cholestatic) conditions, mice fed a high-fat, EFA-sufficient

(EFAS) diet for 7 weeks gradually increased in bodyweight from 25.8±0.6 g to

40.2±2.2 g (Figure 3). Development of EFA deficiency and cholestasis interrupted

physiological growth in mice. Non-supplemented and TG-supplemented EFAD mice

significantly lost weight during the seven-week experiment (from 25.8±1.5 g at base-

line to 22.7±1.2 g and 23.2±1.5 g, respectively; each p<0.01). Suppletion of EFA

with PL, however, prevented this weight loss completely, resulting in a stable body

weight throughout the experiment (PL: 27.0±2.5 g, p<0.05 for PL vs. TG and EFAS).

Plasma bile salt concentrations significantly increased from 21±4 µM at baseline to

46±11 µM after three weeks of EFAD diet. In the fourth week of bile duct ligation,

plasma bile salt concentrations had strongly increased in all groups. However, EFA

supplementation in the form of either PL or TG profoundly mitigated the increase in

plasma bile salt concentration compared to non-supplemented mice (PL: 1242±163

µM, TG: 1104±252 M, EFAD: 1983±510 µM; p<0.05 for either TG or PL vs. EFAD in

week 7; p<0.005 for all groups in week 3 vs. week 7). As expected, alanine amino-

119


Figure 3: Body weight in bile duct-ligated mice fedEFAD diet or EFAD diet supplemented with PL or TG,and of non-cholestatic EFA-sufficient control mice.Data represent means±SD of 5 mice per group.*p<0.01 for TG, EFAD and EFAS in week 0 vs.7.#p<0.05 for PL vs. TG, EFAD and EFAS in week 7.For PL-fed mice, bodyweight was not significantlydifferent in week 7 compared to baseline.

Bodyweight

(gra

m)

non-cholestatic EFAS control

weeks

0

10

20

30

40

0 3 4 5 6 71 2 8

TG

PL

EFAD

*

#

#

Bodyweight

(gra

m)


weeks

0

10

20

30

40

0 3 4 5 6 71 2 8

TG

PL

EFAD

*

#

#

Bodyweight

(gra

m)


weeks

0

10

20

30

40

0

10

20

30

40

0 3 4 5 6 71 2 80 3 4 5 6 71 2 8

TG

PL

EFAD

TG

PL

EFAD

*

#

#

3


transferase (ALT), aspartate aminotransferase (AST) and alkaline phosphatase (AP)

activity in plasma strongly increased after induction of extrahepatic cholestasis

(Table 1). No profound differences between the three experimental groups were

noted in these parameters. Plasma cholesterol concentrations had remained fairly

constant after four weeks of EFAD diet / one week of BDL. After another three weeks

of cholestasis, however, plasma cholesterol levels were significantly higher in PL- and

TG-supplemented mice compared to non-supplemented mice.

The T/T-ratio (C20:3n-9/C20:4n-6) in total plasma lipid increased during development

of EFA deficiency and cholestasis from 0.01±0.00 at baseline to 0.51±0.28 in week

four (p<0.001, Figure 4). Upon supplementation with PL or TG, the T/T-ratio

decreased to 0.32±0.17 and 0.36±0.03, respectively, whereas in non-supplemented

mice the T/T-ratio further increased to 0.61±0.20. Plasma molar percentages of LA

and AA were significantly lower in week four compared to baseline values (p<0.001).

A further decrease in LA levels was partially prevented only after supplementation

with PL and not with TG (p<0.05 for PL vs. EFAD at week 7; TG vs. EFAD NS).

Since total plasma lipid fatty acid composition in non-fasted animals is strongly deter-

mined by postprandial dietary TG, we also analyzed fatty acid profiles of the isolated

plasma PL fraction of the terminal blood sample in week seven (Figure 5). The T/T-

ratio tended to be lower and LA, AA and DHA-concentrations higher in TG- and PL-

fed mice compared to non-supplemented mice, but differences were not statistically

significant. Fatty acid concentrations were not significantly different between the two

supplementation groups.

120

Chapter 6

6.4 ± 2.7 **9.4 ± 0.5 ##9.2 ± 2.8 ##3.3 ± 1.73.2 ± 1.2Cholesterol (mM)

1167 ± 3851520 ± 278 #1885 ± 1097 #895 ± 228 *81 ± 31AP(U/l)

406 ± 192253 ± 60261 ± 82354 ± 199 *34 ± 14ALT (U/l)

739 ± 288451 ± 116549 ± 23659 ± 323 *48 ± 30AST (U/l)

EFADTGPLEFADTGPLEFADTGPL

wk 7 (=4 wk after BDL)wk 4 (=1wk after BDL)wk 0

6.4 ± 2.7 **9.4 ± 0.5 ##9.2 ± 2.8 ##3.3 ± 1.73.2 ± 1.2Cholesterol (mM)

1167 ± 3851520 ± 278 #1885 ± 1097 #895 ± 228 *81 ± 31AP(U/l)

406 ± 192253 ± 60261 ± 82354 ± 199 *34 ± 14ALT (U/l)

739 ± 288451 ± 116549 ± 23659 ± 323 *48 ± 30AST (U/l)

EFADTGPLEFADTGPLEFADTGPL

wk 7 (=4 wk after BDL)wk 4 (=1wk after BDL)wk 0

Table 1: Plasma aspartate transaminase (AST, Units/L), alanine transaminase (ALT, Units/L), alkaline phosphatase (AP, Units/L)and cholesterol (mM) at baseline (week 0), after one week of bile duct ligation (week 4) and after 4 weeks of bile duct ligation (week 7) in TG-supplemented, PL-supplemented or non-supplemented mice. Data represent means±SD of 4-6 mice per group.*p<0.001 for AST, ALT and AP concentrations in week 4 compared to baseline. #p<0.001 for TG and PL supplemented micein week 4 vs. week 7. ##p<0.01 for cholesterol levels in TG and PL supplemented mice in week 7 vs. 4. **p<0.005 for cho-lesterol levels in EFAD vs. TG and PL supplemented mice in wk 7.


121


weeks

LA (18:2n-6)

0

10

20

30

40

50

0 2 4 6 8

TG

PL

EFAD

*

#

mo

l% o

f to

tal f

atty

aci

ds

weeks

T/T ratio

0.0

0.2

0.4

0.6

0.8

1.0

0 2 4 6 8

TG

PL

EFAD

*

mo

l% o

f to

tal f

atty

aci

ds

weeks

AA (20:4n-6)

0

5

10

15

0 2 4 6 8

TG

PL

EFAD

*

mo

l% o

f to

tal f

atty

aci

ds

weeks

LA (18:2n-6)

0

10

20

30

40

50

0 2 4 6 8

TG

PL

EFAD

*

#

mo

l% o

f to

tal f

atty

aci

ds

weeks

LA (18:2n-6)

0

10

20

30

40

50

0

10

20

30

40

50

0 2 4 6 80 2 4 6 8

TG

PL

EFAD

TG

PL

EFAD

*

#

mo

l% o

f to

tal f

atty

aci

ds

weeks

T/T ratio

0.0

0.2

0.4

0.6

0.8

1.0

0 2 4 6 8

TG

PL

EFAD

*

mo

l% o

f to

tal f

atty

aci

ds

weeks

T/T ratio

0.0

0.2

0.4

0.6

0.8

1.0

0 2 4 6 8

TG

PL

EFAD

*

mo

l% o

f to

tal f

atty

aci

ds

weeks

AA (20:4n-6)

0

5

10

15

0 2 4 6 8

TG

PL

EFAD

*

mo

l% o

f to

tal f

atty

aci

ds

weeks

AA (20:4n-6)

0

5

10

15

0

5

10

15

0 2 4 6 80 2 4 6 8

TG

PL

EFAD

TG

PL

EFAD

*

mo

l% o

f to

tal f

atty

aci

ds

Figure 4: Triene/tetraene ratio (T/T-ratio), linoleic acid (LA) and arachidonic acid (AA) in plasma of bile duct ligated mice fedEFAD diet for 4 weeks, and EFA-rich PL, EFA-rich TG or no EFA for 3 weeks. Individual fatty acid concentrations are expressedas mol% of total fatty acids. Data represent means ± SD of 4-6 mice per group. * p<0.001 for the TT-ratio, LA and AA in week0 vs. week 4. # p<0.05 for PL vs. EFAD in week 7; LA in TG-fed mice was not significantly different from EFAD mice in week 7.

Plasma PL cholestatic EFAD mice

Figure 5D

Figure 5A

Figure 5C

Figure 5B

T/T ratio

PL TG EFAD

mol

/ m

ol

EFAS 0.0

0.4

0.6

1.0

0.8

0.2

*

DHA (22:6n-3)

PL TG EFAD

mol

%

EFAS

0.0

0.5

1.5

1.0

#

mol

%

EFAS

LA (18:2n-6)

0

5

10

15

20

25

PL TG EFAD

*

PL TG EFAD

mol

%

EFAS

AA (20: 4n-6)

0

6

8

10

12

4

2

**

Plasma PL cholestatic EFAD mice

Figure 5D

Figure 5A

Figure 5C

Figure 5B

T/T ratio

PL TG EFAD

mol

/ m

ol

EFAS 0.0

0.4

0.6

1.0

0.8

0.2

*

T/T ratio

PL TG EFAD

mol

/ m

ol

EFAS 0.0

0.4

0.6

1.0

0.8

0.2

0.0

0.4

0.6

1.0

0.8

0.2

**

DHA (22:6n-3)

PL TG EFAD

mol

%

EFAS

0.0

0.5

1.5

1.0

#

DHA (22:6n-3)

PL TG EFADPL TG EFAD

mol

%

EFAS

0.0

0.5

1.5

1.0

#

mol

%

EFAS

LA (18:2n-6)

0

5

10

15

20

25

PL TG EFAD

*

mol

%

EFAS

LA (18:2n-6)

0

5

10

15

20

25

0

5

10

15

20

25

0

5

10

15

20

25


**

PL TG EFAD

mol

%

EFAS

AA (20: 4n-6)

0

6

8

10

12

4

2

**


mol

%

EFAS

AA (20: 4n-6)

0

6

8

10

12

4

2

0

6

8

10

12

4

2

******

Figure 5: Triene/tetraene ratio (T/T-ratio), linoleic acid (LA), arachidonic acid (AA) and docosahexaenoic acid (DHA) in the plas-ma PL-fraction of cholestatic mice, after 7 weeks of experimental diet feeding. Fatty acid concentrations are expressed as mol%of total fatty acids. Data represent means±SD of 4-6 mice per group. *p<0.001 for the T/T-ratio and LA in PL-, TG- and non-supplemented mice compared to non-cholestatic EFAS mice. **p<0.05 for AA in PL, TG and non-supplemented mice com-pared to non-cholestatic EFAS controls. #p<0.05 for DHA in EFAD vs. EFAS mice in week 7; DHA in TG-fed mice and PL-fedmice was not significantly different from non-cholestatic EFAS mice.

4a 4b 4c

5a 5b

5c 5d


The fatty acid composition of the plasma compartment is largely determined by the

plasma triglyceride fraction, which reflects dietary fatty acid composition to a certain

extent. Fatty acid profiles in erythrocyte membranes are assumed to be a more sta-

ble reflection of overall essential fatty acid status. The T/T-ratio in erythrocytes was

0.02±0.01 at baseline and increased to 0.12±0.06 in week four (i.e., one week after

bile duct ligation, p<0.001; Figure 6). Continuation of the EFAD diet in cholestatic

mice induced a further increase of the erythrocyte T/T-ratio, which was prevented by

supplementation with either PL or TG (Figure 6). Linoleic acid (LA) decreased from

10.5±0.5 mol% at baseline to 6.4±1.7 mol% in week four (Figure 6b).

Supplementation with PL for three weeks resulted in higher LA concentrations com-

pared to the EFAD group (5.3±1.5 vs. 3.5±1.5 mol%, respectively; p<0.005). TG-

supplemented mice had intermediate LA levels at 7 weeks (4.3±0.8; TG vs. EFAD:

NS). AA molar percentages in cholestatic mice remained remarkably stable during

the entire experiment in all dietary groups: 15.3±2.9 at baseline; 15.7±1.0 in week

four (Figure 6c); and 15.3±4.0 (PL), 13.3±4.3 (TG) and 13.2±1.8 (EFAD) in week

seven of the study period. respectively (NS). Surprisingly, when comparing AA

values of cholestatic with those of non-cholestatic EFA-deficient mice (Figure 2c), we

observed that bile duct ligation completely prevented the decline in AA during EFAD

diet feeding. AA levels in cholestatic EFAD mice remained similar to those in non-

cholestatic EFAS mice (15.4±0.5, NS).

122

Chapter 6

EFAS

weeks

T/T ratio

0.0

0.1

0.2

0.3

0.4

0.5

0 2 4 6 8

TG

PL

EFAD

* **

mo

l% o

f to

tal

fatt

y ac

ids

weeks

LA (18:2n-6)

5

10

15

00 2 4 6 8

EFAS

TG

PL

EFAD

#

*

mo

l% o

f to

tal

fatt

y a

cid

s

weeks

AA (20:4n-6)

5

10

15

20

25

00 2 4 6 8

EFAS

TG

PL

EFAD

mo

l% o

f to

tal f

att

y a

cid

s

EFAS

weeks

T/T ratio

0.0

0.1

0.2

0.3

0.4

0.5

0 2 4 6 8

TG

PL

EFAD

* **

mo

l% o

f to

tal

fatt

y ac

ids

EFAS

weeks

T/T ratio

0.0

0.1

0.2

0.3

0.4

0.5

0.0

0.1

0.2

0.3

0.4

0.5

0 2 4 6 80 2 4 6 80 2 4 6 8

TG

PL

EFAD

TG

PL

EFAD

* **

mo

l% o

f to

tal

fatt

y ac

ids

weeks

LA (18:2n-6)

5

10

15

00 2 4 6 8

EFAS

TG

PL

EFAD

#

*

mo

l% o

f to

tal

fatt

y a

cid

s

weeks

LA (18:2n-6)

5

10

15

00 2 4 6 80 2 4 6 8

EFAS

TG

PL

EFAD

#

*

mo

l% o

f to

tal

fatt

y a

cid

s

weeks

AA (20:4n-6)

5

10

15

20

25

00 2 4 6 8

EFAS

TG

PL

EFAD

mo

l% o

f to

tal f

att

y a

cid

s

weeks

AA (20:4n-6)

5

10

15

20

25

00 2 4 6 80 2 4 6 8

EFAS

TG

PL

EFAD

mo

l% o

f to

tal f

att

y a

cid

s

Figure 6:Triene/tetraene ratio (T/T-ratio), linoleic acid (LA) and arachidonic acid (AA) in red blood cells (RBC) of cholestatic miceduring the 7-week period of experimental diet feeding. Individual fatty acid concentrations are expressed as molar percentagesof total fatty acids. Data represent means±SD of 4-6 mice per group. *p<0.001 for the T/T-ratio and for LA in week 0 vs. 4.**p<0.05 for T/T-ratios of EFAD mice compared to PL- and TG-supplemented mice in week 7. #p<0.005 for LA in PL-supple-mented vs. non-supplemented EFAD mice in week 7; LA concentrations in TG-fed mice were not significantly different fromthose of EFAD mice. AA levels were not significantly different between non-cholestatic EFAS mice and any of the cholestaticEFAD mice throughout the experimental period.

6a 6b 6c


Since the liver is the principal site of desaturation and elongation of EFA into their

long-chain polyunsaturated metabolites, we analyzed the fatty acid composition of

liver triglyceride and liver phospholipid fractions. In liver phospholipids (Figure 7), the

T/T-ratio was significantly lower in PL-fed mice than in TG-fed and EFAD mice after

four weeks of cholestasis (p<0.05, Figure 7a). Linoleic acid (LA), arachidonic acid

(AA) and docosahexaenoic acid (DHA) concentrations were not significantly different

between PL-, TG-, and non-supplemented cholestatic mice. Yet, LA- and AA-levels of

liver phospholipids were significantly lower in cholestatic animals than in non-

cholestatic EFAS mice (p<0.001).

Differential effects of TG and PL supplementation were also observed in the liver

triglyceride fraction (Figure 8). Although LA levels were similar in all 3 cholestatic

groups, the T/T-ratio of PL-supplemented mice was comparable to that of non-

cholestatic EFAS mice (Figure 8a), whereas T/T-ratios of TG-supplemented and

EFAD mice were significantly higher (p<0.05). Arachidonic acid was significantly

higher in PL-supplemented mice compared to TG- or non-supplemented mice (PL:

1.4±0.4 mol%; TG: 0.7±0.2 mol%; EFAD: 0.7±0.2 mol%, p<0.05). Similarly, DHA

levels were higher in the PL-group compared to the TG- and EFAD-groups, and even

higher than in EFAS control mice (Figure 8d).

123


Figure 7B

Liver PL cholestatic EFAD mice

Figure 7A

Figure 7C Figure 7D

T/T ratio

mol

/ m

ol

PL TG EFAD0.0

0.5

1.0

EFAS

*

DHA (22:6n-3)

mol

%

0

1

2

3

4

5

PL TG EFAD

EFAS

NS

mol

%

AA (20:4n-6)

0

5

10

15

20

PL TG EFAD

EFAS

#

mol

%LA (18:2n-6)

0

5

10

15

20

25

PL TG EFAD

EFAS #

Figure 7B

Liver PL cholestatic EFAD mice

Figure 7A

Figure 7C Figure 7D

T/T ratio

mol

/ m

ol

PL TG EFAD0.0

0.5

1.0

EFAS

*

T/T ratio

mol

/ m

ol

PL TG EFAD0.0

0.5

1.0

EFAS

*

DHA (22:6n-3)

mol

%

0

1

2

3

4

5

PL TG EFAD

EFAS

NS

DHA (22:6n-3)

mol

%

0

1

2

3

4

5

0

1

2

3

4

5

PL TG EFAD

EFAS

NS

mol

%

AA (20:4n-6)

0

5

10

15

20

PL TG EFAD

EFAS

#

mol

%

AA (20:4n-6)

0

5

10

15

20

0

5

10

15

20


EFAS

#

mol

%LA (18:2n-6)

0

5

10

15

20

25

PL TG EFAD

EFAS #m

ol%

LA (18:2n-6)

0

5

10

15

20

25

0

5

10

15

20

25

PL TG EFAD

EFAS #

Figure 7: T/T-ratio, linoleic acid (LA), arachidonic acid (AA) and docosahexaenoic acid (DHA) in liver PL of cholestatic mice after7 weeks of experimental diet feeding, and of non-cholestatic EFAS controls (dotted line). Fatty acid concentrations are mol% oftotal fatty acids. Data represent means±SD of 4-6 mice per group. *p<0.05 for T/T-ratios of PL-fed and EFAS mice vs. TG-fedand EFAD mice. #p<0.001 for LA and AA of PL-, TG- and non-supplemented cholestatic mice vs. non-cholestatic EFAS mice.

7a 7b

7c 7d


Oil red O staining for neutral lipids on frozen liver sections showed lipid accumula-

tion in periportal (zone 1) but not in perivenous (zone 3) hepatocytes of EFA-deficient

non-cholestatic mice. EFA-sufficient non-cholestatic mice did not show hepatic lipid

accumulation. In paraffin-embedded HE-stained liver sections of cholestatic mice,

extensive bile duct proliferation was observed in all three dietary groups, with exten-

sive hepatocyte damage due to toxic bile accumulation. No accumulation or zonal

distribution of lipid was observed in livers of cholestatic mice. There were no overt

histological differences between the three cholestatic groups (data not shown).

Since the central nervous system is a well-known target organ for LCPUFA and

contains particularly high levels of docosahexaenoic acid, we analyzed fatty acid

profiles of brain tissue samples. PL-supplemented mice had significantly higher AA

and DHA concentrations in phospholipids isolated from brain tissue than TG-fed and

EFAD mice (each p<0.05 for differences between PL vs. TG and EFAD, Figure 9),

and even slightly higher DHA concentrations than non-cholestatic EFAS mice,

124

Chapter 6

Liver TG cholestatic EFAD mice

Figure 8A

Figure 8C

T/T ratio

0. 0

0. 5

1. 0

1. 5

PL TG EFAD

mo

l / m

ol

EFAS

*

mol

%

EFAS

LA (18:2n-6)

0

5

10

15

20

25

PL TG EFA D

**

AA (20:4n-6)

mol

%

0.0

0.5

1.0

1.5

2.0

PL TG E FAD

E FAS

#

DHA (22:6n-3)

0.0

0.5

1.0

PL TG E FAD

mo

l%

EFAS

##

Liver TG cholestatic EFAD mice

Figure 8A

Figure 8C

T/T ratio

0. 0

0. 5

1. 0

1. 5

PL TG EFAD

mo

l / m

ol

EFAS

*T/T ratio

0. 0

0. 5

1. 0

1. 5

0. 0

0. 5

1. 0

1. 5

PL TG EFAD

mo

l / m

ol

EFAS

*

mol

%

EFAS

LA (18:2n-6)

0

5

10

15

20

25

PL TG EFA D

**mol

%

EFAS

LA (18:2n-6)

0

5

10

15

20

25

0

5

10

15

20

25

PL TG EFA DPL TG EFA D

****

AA (20:4n-6)

mol

%

0.0

0.5

1.0

1.5

2.0

PL TG E FAD

E FAS

#

AA (20:4n-6)

mol

%

0.0

0.5

1.0

1.5

2.0

0.0

0.5

1.0

1.5

2.0

PL TG E FADPL TG E FAD

E FAS

#

DHA (22:6n-3)

0.0

0.5

1.0

PL TG E FAD

mo

l%

EFAS

##

DHA (22:6n-3)

0.0

0.5

1.0

0.0

0.5

1.0

PL TG E FAD

mo

l%

EFAS

####

Figure 8: Triene/tetraene ratio (T/T-ratio), linoleic acid (LA), arachidonic acid (AA) and docosahexaenoic acid (DHA) in liver TGof cholestatic mice after 7 weeks of experimental diet feeding. Fatty acid concentrations are expressed as mol% of total fattyacids. Data represent means±SD of 4-6 mice per group. *p<0.05 for the T/T-ratio of PL-fed and EFAS mice (dotted line) vs. TG-fed and EFAD mice. **p<0.001 for LA concentrations in cholestatic mice vs. non-cholestatic EFAS mice. #p<0.05 for AA in PL-fed and EFAS mice vs. TG-fed and EFAD mice. ##p<0.05 for DHA levels in PL-fed mice vs. TG-fed, EFAD and EFAS mice.

8a 8b

8c 8d


although the latter difference was not significant. The third major brain fatty acid,

C22:4n-6, was similarly lower in TG-fed and EFAD mice than in non-cholestaticcontrols (1.69±0.13 mol% (TG), 1.63±0.12 mol% (EFAD), 2.09±0.20 mol% (EFAS),

p<0.01), whereas PL-supplemented mice had C22:4n-6 concentrations comparable

to non-cholestatic EFAS mice (1.83±0.13 mol%, NS; data not shown).

DISCUSSION

We compared the efficacy of EFA-rich triglycerides (TG) and EFA-rich phospholipids

(PL) for correcting EFA deficiency under cholestatic conditions in mice. We

hypothesized that PL would be more effective than TG for oral treatment of EFAdeficiency in cholestatic liver disease, since dietary TG are profoundly malabsorbed

during cholestasis and PL are less dependent on bile for intestinal absorption.

Present results indicate that, indeed, oral supplementation with PL is superior to oralTG in increasing EFA-derived LCPUFA concentrations in brain and liver.

125


LA (18:2n-6)

PL TG EFAD

mol

%

0.0

0.2

0.4

0.6

0.8

1.0

EFAS

**

**

AA (20:4n-6)

PL TG EFAD

mol

%

0

2

4

6

8

10

EFAS

#

DHA (22:6n-3)

PL TG EFAD

mol

%

0

5

10

15

EFAS

#

LA (18:2n-6)

PL TG EFAD

mol

%

0.0

0.2

0.4

0.6

0.8

1.0

EFAS

**

**

LA (18:2n-6)

PL TG EFAD

mol

%

0.0

0.2

0.4

0.6

0.8

1.0

EFAS

**

**

AA (20:4n-6)

PL TG EFAD

mol

%

0

2

4

6

8

10

EFAS

#

AA (20:4n-6)

PL TG EFAD

mol

%

0

2

4

6

8

10

EFAS

#

DHA (22:6n-3)

PL TG EFAD

mol

%

0

5

10

15

EFAS

#

DHA (22:6n-3)

PL TG EFAD

mol

%

0

5

10

15

EFAS

#

Figure 9: Linoleic acid (LA), arachidonic acid (AA) and docosahexaenoic acid (DHA) in brain PL of cholestatic mice after 7weeks of experimental diet feeding, and of non-cholestatic EFAS controls (dotted line). Individual fatty acid concentrations areexpressed as molar percentages of total fatty acids. Data represent means ± SD of 4-6 mice per group. *p<0.05 for T/T-ratiosof PL-fed and TG-fed mice compared to EFAD mice, and for all cholestatic groups compared to non-cholestatic EFAS mice (dot-ted line). **p<0.05 for LA concentrations of EFAD mice compared to PL-fed and EFAS mice. AA and DHA were significantlyhigher in brain PL of PL-fed mice compared to TG-fed and EFAD mice ( #p<0.01).

9a 9b 9c


We used and adapted a murine model for diet-induced EFA deficiency that we devel-

oped and characterized previously(19). Ligation of the common bile duct in mice fed

an EFAD diet for three weeks resulted in acute extrahepatic cholestasis, combined

with a diet-induced EFA deficiency. Both our cholestatic and non-cholestatic mouse

models for EFA deficiency developed the characteristic biochemical hallmarks of

EFA deficiency in plasma and erythrocyte fatty acid profiles(21;23;24). Plasma and

erythrocyte EFA and LCPUFA concentrations strongly decreased, and levels of non-

essential fatty acids such as mead acid (C20:3n-9) and oleic acid concomitantly

increased. The cessation of growth that we observed in EFA-deficient mice is likely

related to impaired dietary fat absorption during EFA deficiency, which had been

described previously (22;25;26).

Superimposing extrahepatic cholestasis on EFA deficiency in mice subsequently

resulted in weight loss. Bile-diverted rats compensate for bile deficiency-induced fat

malabsorption by increasing their chow ingestion(27), in contrast to rats with bile duct

ligation(1). In our bile duct-ligated EFA-deficient mice, chow intake similarly remained

constant, suggesting that two causes for fat malabsorption are present in our mouse

model; bile deficiency and EFA deficiency, the combination of which presumably

induced weight loss. Bile duct ligation profoundly elevated plasma bile salt

concentrations, which were already slightly elevated by prior EFAD diet feeding.

Liver histology revealed extensive bile duct proliferation and parenchymal damage in

cholestatic mice, accompanied by jaundice within days after bile duct closure.

The rapid onset of these biochemical and morphological parameters of EFA

deficiency and cholestasis in our mice provided us with an effective model for

studying EFA deficiency under acute cholestatic conditions.

In cholestatic EFAD mice, oral administration of EFA-rich PL completely restored

DHA and AA levels in the target organs brain and liver, in contrast to EFA-rich TG

which did not improve these parameters relative to continuation of the EFAD diet.

It is well-known that the high levels of DHA and AA in the excitable retinal and synap-

tosomal membranes of the central nervous system are crucial for adequate

membrane reactivity and function of receptor proteins(28-30). The molecular mechanism

underlying the high concentrations of DHA and AA in the brain is unknown. Scott and

Bazan(31) demonstrated that the majority of brain DHA originates from hepatic elon-

gation and desaturation of n-3 fatty acids and subsequent redistribution to the brain.

Additionally, astrocytes and brain endothelial cells are capable of synthesizing LC-

PUFA from EFA, which may constitute a minor but still relevant source of DHA and

AA for the central nervous system(32-35). The plasma compartment appears to be the

main source of brain DHA and AA. LCPUFA are partly present in plasma as lipo-

126

Chapter 6


protein components, but Purdon and Rapoport demonstrated in rats that LCPUFA

esterified within circulating lipoproteins do not enter the brain to a measurable extent,

and that only the unesterified form is incorporated(36;37). Unesterified LCPUFA bound

to albumin are a well recognized plasma source of PUFA for the brain(38), but

concentrations of particularly n-3 fatty acids are low, and it is questionable whether

the free fatty acid compartment accounts for the majority of LCPUFA supply to the

brain(39). LCPUFA are also present in plasma as lyso-PL, bound to albumin(40). Thies

and Bernoud(41;42) reported data suggesting that the brain preferentially absorbs DHA

as sn-2 lyso-PL compared to unesterified DHA. Since dietary PL are partly absorbed

as lyso-PL and partly as intact PL molecules, the highly efficient increase in brain

LCPUFA concentration after PL supplementation in our study could be in line with

these observations, supporting the presumed high bioavailability of dietary PL(12;13;15;16).

Surprisingly, PL-supplementation seemed to increase brain DHA to an even slightly

higher level than in non-cholestatic mice fed the EFAS diet. However, since both

EFAD and EFAS diets contain very low amounts of ALA, prolonged EFAS diet

feeding may result in marginal DHA levels. The TG- and PL-enriched diets equally

provided supplementary ALA, which has the capacity to increase DHA levels, even

compared to EFAS mice. However, comparison between the TG- and the PL-fed

groups indicated that only PL supplementation increased brain DHA.

We did not find indications that EFA deficiency and cholestasis affected activities of

hepatic desaturation and elongation enzymes required for AA and DHA synthesis

from their respective precursors. Ratios between C20:3n-6 and C20:4n-6, estimating

delta-5 desaturase activity, and between C22:5n-3 and C22:6n-3 marking delta-6

desaturase activity, were similar between supplemented and non-supplemented

mice, nor between cholestatic and non-cholestatic EFAD mice. Present observations

are in agreement with previous results in EFAS bile duct- ligated rats, which showed

no indications for altered post-absorptive EFA metabolism during cholestasis(1).

Oral PL more efficiently improved LCPUFA concentrations in brain and in liver than

oral TG. Interestingly, PL and TG reversed parameters of EFA deficiency in plasma

and erythrocytes with equal efficacy. Demarne et al. and Minich et al. demonstrated

in cholestatic rats that absorption of unsaturated fatty acids, and of EFA in particular,

is relatively preserved compared to that of saturated species(1;11). Quantitative

absorption studies would be required to fully exclude differences in net enteral

uptake of TG-EFA and PL-EFA. In addition, the products of TG and PL lipolysis (2-MG

and 1-lyso PL) follow qualitatively different routes within the enterocyte, possibly

resulting in different post-absorptive metabolic pathways of the attached EFA.

127



Although we could not demonstrate differential effects of the TG and PL supplements

in the plasma PL fraction, fatty acid profiles of plasma PL closely corresponded with

fatty acid profiles as measured in liver PL. If PL on the surface of chylomicron

particles were specifically targeted for EFA delivery to the brain, then plasma PL fatty

acid analysis is likely to miss differential effects of oral PL over TG if blood samples

are not taken during or immediately after fat absorption, due to the rapidity of

chylomicron clearance(43).

Our present data are in contradiction with the paradigm that the erythrocyte mem-

brane fatty acid composition is a tentative index for overall body EFA status. Rather,

our data are in line with the recent questioning by several authors of the validity of

extrapolating erythrocyte membrane fatty acid profiles to their status in target organs

for EFA. Korotkova and Strandvik(44) demonstrated that EFA deficiency in rats differ-

entially affects fatty acid profiles of erythrocyte-, serum-, jejunum-, ileum- and colon-

PL. Rioux and Innis(45) reported that in piglets, plasma PL are good indicators for liver

and bile PL-fatty acids, but not for brain fatty acid composition. Our results in mice

support these observations, since, indeed, fatty acid profiles of the isolated plasma

PL fraction closely corresponded with those of liver PL but not of brain PL.

Data suggest that specific channeling of PUFA to target organs as the central

nervous system occurs, at the expense of less critical tissues such as erythrocytes.

Supplementation of EFA as PL not only improved biochemical parameters of EFA

deficiency in cholestatic mice, oral PL also prevented weight loss during EFA

deficiency and cholestasis, whereas oral TG did not. Nishioka et al.(7) recently demon-

strated that enteral infusion of PL-cholesterol liposomes partially corrects lipid mal-

absorption in bile-diverted rats, compatible with a facilitating effect of enteral PL

supplementation on fat uptake during bile deficiency. Although no information on fat

balance is available from the present studies, it is tempting to speculate that EFA-rich

PL supplementation partially corrects fat malabsorption in EFAD cholestatic mice.

A surprising observation was the remarkably stable AA concentration in erythrocytes

of cholestatic EFAD mice, when compared to non-cholestatic EFAD mice. While in

the latter AA levels decreased after starting the EFAD diet, bile duct ligation per se

appeared to maintain AA concentrations at a normal level. Tso et al.(5) reported that

human bile PL provide up to 1.7 g AA to the intestine per day. An average Western

adult diet supplies 1.8 g AA daily(46), indicating that biliary PL secretion into the

intestine provides a significant portion of enteral AA. Our results support the concept

that biliary secretion of AA in the form of PL quantitatively affects overall body AA

homeostasis.

128

Chapter 6


EFA deficiency in pediatric cholestatic patients has been proven difficult to correct by

merely increasing EFA ingestion. It should be realized that in patients, in contrast to

the presented mouse model, cholestasis usually develops gradually, and EFA

deficiency is not due to lack of dietary EFA content. Yet, our present experiments

suggest that oral administration of EFA-rich PL might be an effective treatment

strategy for reversing EFA deficiency in patients with cholestatic liver disease.

The efficacy of oral PL supplementation for treatment of EFA deficiency in pediatric

patients with cholestasis is currently under investigation.

AcknowledgementsThe authors would like to thank H. Helmink and W. de Groot from Unimills

Zwijndrecht (the Netherlands) for their generous gift of the soybean TG and PL oils,

and Ingrid Martini, Juul Baller and Renze Boverhof for their technical expertise and

assistance in the experiments described in this article.

REFERENCES1. Minich DM, Havinga R, Stellaard F, Vonk RJ, Kuipers F, Verkade HJ. Intestinal absorption and postabsorptive metabolism oflinoleic acid in rats with short-term bile duct ligation. Am.J.Physiol.2000.Dec.;279.(6.):G1242.-8. 2000;279:G1242-G1248.2. Burke PA, Ling PR, Forse RA, Lewis DW, Jenkins R, Bistrian BR. Sites of conditional essential fatty acid deficiency in endstage liver disease. JPEN J.Parenter.Enteral Nutr. 2001;25:188-93.3. Robberecht E, Koletzko B, Christophe A. Several mechanisms contribute to the abnormal fatty acid composition of serumphospholipids and cholesterol esters in cholestatic children with extrahepatic biliary atresia. Prostaglandins Leukot.Essent.FattyAcids 1997;56:199-204.4. Socha P, Koletzko B, Swiatkowska E, Pawlowska J, Stolarczyk A, Socha J. EFA metabolism in infants with cholestasis. ActaPaediatr. 1998;87:278-83.5. Tso P. Intestinal lipid absorption. In: Tso P, ed. Physiology of the gastrointestinal tract. New York: Raven Press 1994:1867-907.6. Verkade HJ, Tso P. Biophysics of intestinal luminal lipids. In: Tso P, Mansbach CM, Kuksis A, eds. Intestinal lipid metabolism.New York: Kluwer 2000:1-18.7. Nishioka T, Havinga R, Tazuma S, Stellaard F, Kuipers F, Verkade HJ. Administration of phosphatidylcholine-cholesterol lipo-somes partially reconstitutes fat absorption in chronically bile-diverted rats. Biochim.Biophys.Acta 2004;1636:90-8.8. Scow RO, Stein Y, Stein O. Incorporation of dietary lecithin and lysolecithin into lymph chylomicrons in the rat. J.Biol.Chem.1967;242:4919-24.9. Bloom B, Kiyasu JY, Reinhardt WO, Chaikoff IL. Absorption of plasma phospholipides. Manner of transport from intestinallumen to lacteals. Am.J.Physiol 54 A.D.;177:84-6.10. Mansbach CM. The origin of chylomicron phosphatidylcholine in the rat. J.Clin.Invest 1977;60:411-20.11. Demarne Y, Corring T, Pihet A, Sacquet E. Fat absorption in germ-free and conventional rats artificially deprived of bile secre-tion. Gut 1982;23:49-57.12. Lemaitre-Delaunay D, Pachiaudi C, Laville M, Pousin J, Armstrong M, Lagarde M. Blood compartmental metabolism of DHAin humans after ingestion of a single dose of 13C-DHA in phosphatidylcholine. J.Lipid Res. 1999;40:1867-74.13. Boehm G, Borte M, Boehles H, Mueller H, Kohn G, Moro G. DHA and AA content of serum and RBC membrane PL ofpreterm infants fed breast milk, standard formula or formula suppleted with n3 and n6 LCPUFA. Eur J Pediatr 1996;155:410-6.14. Wijendran V, Huang MC, Diau GY, Boehm G, Nathanielsz PW, Brenna JT. Efficacy of dietary AA provided as triglyceride orphospholipid as substrates for brain AA accretion in baboon neonates. Pediatr.Res. 2002;51:265-72.15. Simopoulos AP, Salem N, Jr. Egg yolk as a source of LCPUFA in infant feeding. Am.J.Clin.Nutr. 1992;55:411-4.16. Carnielli VP, Verlato G, Pederzini F et al. Intestinal absorption of long-chain polyunsaturated fatty acids in preterm infants fedbreast milk or formula. Am.J.Clin.Nutr. 1998;67:97-103.17. Muskiet FA, van Doormaal JJ, Martini IA, Wolthers BG, van der SW. Capillary gas chromatographic profiling of total long-chain fatty acids and cholesterol in biological materials. J.Chromatogr. 1983;278:231-44.

129



18. Bligh EG, Dyer WJ. A rapid method for total lipid extraction and purification. Can.J.Biochem.Physiol. 1959;37:911-7.19. Werner A, Minich DM, Havinga R et al. Fat malabsorption in essential fatty acid-deficient mice is not due to impaired bileformation. Am.J.Physiol Gastrointest.Liver Physiol 2002;283:G900-G908.20. Mashige F, Imai K, Osuga T. A simple and sensitive assay of total serum bile acids. Clin.Chim.Acta 1976;70:79-86.21. Holman RT. The ratio of trienoic:tetraenoic acids in tissue lipids as a measure of EFA requirement. J Nutr 1960;70:405-10.22. Levy E, Garofalo C, Thibault L et al. Intraluminal and intracellular phases of fat absorption are impaired in essential fatty aciddeficiency. Am.J.Physiol. 1992;262:G319-G326.23. Aaes-Jorgensen E, Holman RT. Essential fatty acid deficiency-content of polyenic acids in testes and heart as an indicatorof EFA status. J.Nutr. 1958;65:633-41.24. Yamanaka WK, Clemans GW, Hutchinson ML. Essential fatty acid deficiency in humans. Prog.Lipid Res. 1981;19:187-215.25. Bennett Clark S, Ekkers TE, Singh A, Balint JA, Holt PR, Rodgers JB. Fat absorption in EFA deficiency: a model experi-mental approach to studies of the mechanism of fat malabsorption of unknown etiology. JLR 1973;14:581-8.26. Barnes RH, Miller ES, Burr GO. Fat absorption in essential fatty acid deficiency. J Biol Chem 1941;140:773-8.27. Minich DM, Kalivianakis M, Havinga R et al. Bile diversion in rats leads to a decreased plasma concentration of linoleic acidwhich is not due to decreased net intestinal absorption of dietary linoleic acid. Biochim.Biophys.Acta 1999;1438:111-9.28. Salem N, Jr., Litman B, Kim HY, Gawrisch K. Mechanisms of action of DHA in the nervous system. Lipids 2001;36:945-59.29. Auestad N, Montalto MB, Hall RT et al. Visual acuity, erythrocyte fatty acid composition, and growth in term infants fed for-mulas with long-chain polyunsaturated fatty acids for one year. Ross Pediatric Lipid Study. Pediatr.Res. 1997;41:1-10.30. Carlson SE. Docosahexaenoic acid and arachidonic acid in infant development. Semin.Neonatol. 2001;6:437-49.31. Scott BL, Bazan NG. Membrane docosahexaenoate is supplied to the developing brain and retina by the liver.Proc.Natl.Acad.Sci.U.S.A 1989;86:2903-7.32. Williard DE, Harmon SD, Kaduce TL et al. Docosahexaenoic acid synthesis from n-3 polyunsaturated fatty acids in differ-entiated rat brain astrocytes. J.Lipid Res. 2001;42:1368-76.33. Moore SA, Yoder E, Spector AA. Role of the blood-brain barrier in the formation of long-chain omega-3 and omega-6 fattyacids from essential fatty acid precursors. J.Neurochem. 1990;55:391-402.34. Green P, Yavin E. Elongation, desaturation, and esterification of EFA by fetal rat brain in vivo. J.Lipid Res. 1993;34:2099-107.35. Pawlosky RJ, Ward G, Salem N, Jr. EFA uptake and metabolism in the developing rodent brain. Lipids 1996;31 :S103-7.:S103-S107.36. Purdon D, Arai T, Rapoport S. No evidence for direct incorporation of esterified palmitic acid from plasma into brain lipidsof awake adult rat. J.Lipid Res. 1997;38:526-30.37. Rapoport SI, Chang MC, Spector AA. Delivery and turnover of plasma-derived essential PUFA in mammalian brain. J.LipidRes. 2001;42:678-85.38. Dhopeshwarkar GA, Mead JF. Uptake and transport of fatty acids into the brain and the role of the blood-brain barrier sys-tem. Adv.Lipid Res. 1973;11:109-42.39. Spector AA. Plasma free fatty acid and lipoproteins as sources of PUFA for the brain. J.Mol.Neurosci. 2001;16:159-65.40. Qi K, Hall M, Deckelbaum RJ. LCPUFA accretion in brain. Curr.Opin.Clin.Nutr.Metab Care 2002;5:133-8.41. Thies F, Pillon C, Moliere P, Lagarde M, Lecerf J. Preferential incorporation of sn-2 lysoPC DHA over unesterified DHA in theyoung rat brain. Am.J.Physiol 1994;267:R1273-R1279.42. Bernoud N, Fenart L, Moliere P et al. Preferential transfer of 2-docosahexaenoyl-1-lysophosphatidylcholine through an invitro blood-brain barrier over unesterified docosahexaenoic acid. J.Neurochem. 1999;72:338-45.43. Hultin M, Savonen R, Olivecrona T. Chylomicron metabolism in rats: lipolysis, recirculation of TG-derived fatty acids in plas-ma FFA, and fate of core lipids as analyzed by compartmental modelling. Journal of lipid research 1996;37:1022-36.44. Korotkova M, Strandvik B. Essential fatty acid deficiency affects the fatty acid composition of the rat small intestinal andcolonic mucosa differently. Biochim.Biophys.Acta 2000;1487:319-25.45. Rioux FM, Innis SM, Dyer R, MacKinnon M. Diet-induced changes in liver and bile but not brain fatty acids can be predict-ed from differences in plasma phospholipid fatty acids in formula- and milk-fed piglets. J.Nutr. 1997;127:370-7.46. Brindley DN. Hepatic secretion of lysophosphatidylcholine: A novel transport system for polyunsaturated fatty acids andcholine. J.Nutr.Biochem. 1993;4:442-9.

130

Chapter 6


Oral treatment ofessential fatty acid deficiency with

triglycerides or phospholipids inchildren with end stage liver disease

Submitted

7

A. WernerC.M.A. Bijleveld

I.A. Martini M. van Rijn

J. van der Heiden P.J.J. Sauer

H. J. Verkade


ABSTRACT

Background: Essential fatty acid (EFA) deficiency is frequently observed in children

with end-stage liver disease, mainly due to malabsorption of dietary EFA. Recently,

we demonstrated in cholestatic EFA-deficient mice that oral phospholipids (PL) more

effectively improve EFA-status than oral triglycerides (TG). In this study, we

compared the effects of oral EFA supplementation in the form of PL or TG on EFA

status in children with end-stage liver disease.

Methods: Pediatric candidates for liver transplantation were orally supplemented with

EFA-rich TG or EFA-rich PL for three months. Red blood cell (RBC) fatty acid

composition was determined at baseline and after one, two and three months of EFA

supplementation, as were serum liver enzyme and vitamin A and E concentrations.

Results were compared to data obtained from non-supplemented patients prior to

the study period. Dietary EFA intake was calculated from food diaries at baseline and

after three months of EFA supplementation.

Results: Mead acid (C20:3n-9) significantly increased (p<0.05), and LA and the total

amount of n-6 fatty acids decreased (p<0.005) in RBC of the non-supplemented

group (p<0.01), whereas total n-6 fatty acids remained stable in both the PL and the

TG group. The rate of increase in RBC-LA was significantly higher in PL-supplement-

ed, but not in TG-supplemented children, compared with non-supplemented

children. No significant differences in RBC alpha-linolenic acid (ALA), arachidonic

acid (AA) or docosahexaenoic acid (DHA) were found between TG-, PL-, or non-

supplemented patients. Baseline dietary LA intake was well above the Dutch daily

recommended intake, and similar in the three groups.

Conclusions: Oral EFA supplementation prevents deterioration of EFA status in

children with end-stage liver disease. Present results suggest that EFA supplemen-

tation as PL is slightly more effective for improving RBC EFA than as TG.

132

Chapter 7


INTRODUCTION

Essential fatty acid (EFA) deficiency is a common finding in children with cholestatic

liver disease(1;2). Retrospective analysis of fatty acid profiles determined in our

hospital between 1990 and 1996 revealed that almost 80% of children listed for ortho-

topic liver transplantation for end-stage (cholestatic) liver disease (ESLD) had

indications for compromised essential fatty acid status(3). EFA deficiency during

cholestasis is predominantly due to dietary fat malabsorption(4). Absorption of dietary

fat during enteral bile insufficiency may be reduced from 97% to 40% of the amount

ingested. It has been demonstrated that EFA deficiency not only can result from

dietary fat malabsorption, but reversely, inadequate EFA levels may also impair

absorption of dietary fat(5-8).

EFA deficiency in children with end-stage liver disease (ESLD) has been proven

difficult to treat by merely increasing dietary EFA ingestion(3;9;10). In dietary fat, EFA is

predominantly present in the form of triglycerides (90%), and only 10% as phospho-

lipid and cholesterol ester. The major steps involved in intestinal lipid absorption are

intraluminal hydrolysis, micellar solubilization, translocation across the unstirred

water layer and the microvillous enterocyte membrane, and intracellular lipoprotein

assembly and secretion(5;6;11). The relative importance of each of these processes

differ for the absorption of triglycerides and phospholipids, partly based on physico-

chemical differences. Lipids have been categorized into polar and non-polar lipid

classes according to the nature of their interactions with water(11;12). Polar lipids are

divided into 3 subclasses. Triglycerides are hydrophobic class 1 polar lipid molecules

and are insoluble in the aqueous intestinal lumen, require (partial) hydrolysis and

highly depend on bile for solubilization into micelles. After absorption by the entero-

cyte and subsequent re-esterification, triglycerides are secreted into lymph as core

components of chylomicrons. Phospholipids, on the other hand, are class 2 polar

lipids, which are more hydrophilic and associate into liquid crystals in the intestinal

lumen, which has been suggested to improve luminal fat solubilization during bile

deficiency. Phospholipids are the major surface components for chylomicrons, and

have been postulated to have a higher post-absorptive bioavailability for the target

organs of EFA, liver and brain. In EFA-deficient mice with acute cholestasis, we

recently demonstrated that oral EFA-rich phospholipids more effectively improved

EFA status than oral EFA-rich triglycerides(13). Carnielli et al. demonstrated in preterm

infants that n-3 LCPUFA were more effectively absorbed from infant formulas when

supplied as PL than as TG(14). In this study, PUFA absorption efficacies were

measured by fecal balance techniques, and no data on RBC EFA were provided.

Obviously, in human studies, obtaining material to analyze brain or liver EFA contents

133

Oral treatment of EFA deficiency with TG or PL in children with end-stage liver diease


is not feasible. Yet, considering the favorable outcomes of PL-supplementation on

brain and liver PUFA in mice, equal effects of TG and PL supplementation on RBC

EFA in humans might still advocate preferential EFA supplementation with PL rather

than with TG in cholestatic patients.

In the present study, we compared the efficacy of oral EFA-rich PL and oral EFA-rich

TG supplementation on RBC-EFA status in children with ESLD. Non-supplemented

ESLD patients formed the control group.

SUBJECTS AND METHODS

Patient characteristicsPediatric liver transplantation in the Netherlands is centralized in Groningen. Patients

listed for orthotopic liver transplantation (OLT) visit the University Medical Center

Groningen outpatient clinic on a monthly basis for clinical and biochemical

evaluation during the waiting period for transplantation. Participants for the study

were recruited between January 2000 and January 2004. Since we had previously

observed marginal EFA-status in the majority of children with ESLD(3), children were

randomly assigned to the TG or the PL supplementation group. Participants

consented to ingest an EFA supplement for three months, and were unaware of the

type of supplement they were assigned to. A non-supplemented control group was

composed of data available from patients, prior to the supplementation study. The

study group included 19 pediatric pre-OLT patients (10 TG, 9 PL), 11 male and 8

female, ranging in age from 1 to 16 years.

Study protocol / Experimental proceduresVenous EDTA-anticoagulated blood samples were collected (3-5 ml) during regular

out-patient clinic visits (baseline) and after 1, 2 and 3 months of oral EFA supple-

mentation. EFA status in erythrocytes, plasma vitamin A and E concentrations and

serum liver enzyme activities (AST, ALT, gamma-GT, alkaline phosphatase, bilirubin,

albumin and coagulation parameters) were determined monthly during the study

period, conform our standard procedures for pediatric liver transplant candidates.

Dietary EFA intake (apart from the supplement) was calculated from 2-day

consecutive food diaries, at the start and at the end of the 3-month supplementation

period, by a clinical dietician using the Netherlands Nutrients Table "NEVO" 2001.

A schematic overview of the experimental design is depicted in Figure 1. The Medical

Ethics Committee of the University Medical Center Groningen approved the study

134

Chapter 7


protocol, which included obtaining informed consent from the participating children

and their parents. Figure 2 summarizes the anthropometric data (age, body weight,

male/female) and the underlying liver disease.

Experimental EFA supplementsThe primary TG and PL were purified from crude soybean oil and had the following

fatty acid profile:TG-supplement PL-supplementmol% mol%

linoleic acid (C18:2n-6) 53.5 56.1alpha-linolenic acid (C18:3n-3) 7.6 8.2oleic acid (C18:1n-9) 22.8 13.2palmitic acid (C16:0) 10.7 16.1stearic acid (C18:0) 4.0 4.5myristic acid (C14:0) <0.5 <0.5 palmitoleic acid (C16:1n-7) <0.5 <0.5arachidic acid (C20:0) <0.5 <0.5arachidonic acid (C20:4n-6) <0.5 <0.5behenic acid (C22:0) <0.5 <0.5eicosapentaenoic acid (C20:5n-3) <0.5 <0.5docosahexaenoic acid (C22:6n-3) <0.5 <0.5

Both the PL and the TG oils were a generous gift of Unimills BV, the Netherlands.

We aimed to supplement the cholestatic children with 0.125 mg LA per kg body

weight per day, which is 25% of the daily recommended dietary intake (RDI) as for-

mulated by the Netherlands Nutrition Council (Nederlandse Voedingsraad).

In the TG oil, 4% of molecular mass is accounted for by the glycerol backbone and

96% by fatty acids, of which 54% is linoleic acid (LA). The PL oil (lecithin) was

135


pre-OLT cholestatic patients

RBC-EFA

plasma liver fuctions

plasma vitamin A, E

dietary EFA intake questionnaire

RBC-EFA


plasma vitamin A, E

RBC-EFA


plasma vitamin A, E

RBC-EFA


plasma vitamin A, E


EFA-rich PL

EFA-rich TG

no EFA supplement

month 1 month 2 month 3

t=0 t=1 t=2 t=3

pre-OLT cholestatic patients

RBC-EFA


plasma vitamin A, E


RBC-EFA


plasma vitamin A, E

RBC-EFA


plasma vitamin A, E

RBC-EFA


plasma vitamin A, E


EFA-rich PL

EFA-rich TG

no EFA supplement


t=0 t=1 t=2 t=3

RBC-EFA


plasma vitamin A, E


RBC-EFA


plasma vitamin A, E

RBC-EFA


plasma vitamin A, E

RBC-EFA


plasma vitamin A, E


EFA-rich PL

EFA-rich TG

no EFA supplement


t=0 t=1 t=2 t=3

EFA-rich PL

EFA-rich TG

no EFA supplement

EFA-rich PL

EFA-rich TG

no EFA supplement

month 1 month 2 month 3month 1 month 2 month 3

t=0 t=1 t=2 t=3t=0 t=1 t=2 t=3

Figure 1: Experimental design of EFA supplementation to children with end-stage liver disease. RBC fatty acid profiles, serumliver enzymes and vitamin A and E levels were determined monthly, and dietary EFA intake was calculated from food diaries atbaseline and after 3 months of supplementation.


almost exclusively composed of lecithin (>98%). For the PL oil, the glycerol-

phosphate-choline backbone composes 28% of molecular mass, and 56% of the

remaining 72% fatty acid mass is linoleic acid (LA). To obtain equimolar daily

amounts of linoleic acid (LA) for the two EFA-supplements, the TG-oil was dosed at

0.23 ml per kg bodyweight per day, and the PL-emulsion at 1.5 ml per kg per day,

resulting in 0.4 mmol LA per kg bodyweight for each supplement. The purified PL oil

(lecithin) was administered in the form of a water-in-oil emulsion, prepared by our

University Medical Center Groningen (UMCG) Hospital Pharmacy, based on the high

viscosity of the pure PL oil. The PL emulsion contained per liter: 1.0 g methyl-

parahydroxybenzoate (conservative), 1.0 g saccharine sodium-2-water (sweetener),

150 g PL oil, 20 g caramellose sodium, 10 g polysorbate 80 (emulsion stabilizer), and

1.0 g peach or banana aroma, ad 1.0 liter distilled water. The EFA supplements were

advised to be taken in three daily doses, during or after meals. Based on pilot

studies, daily ingestion of 30 ml of supplement was the maximum tolerable amount

for most children. Prescription of greater volumes of EFA-supplement frequently

resulted in malcompliance, partly due to the moderate palatability of the supplement.

Therefore, the prescribed amount of PL-supplement of 1.5 ml per kg per day was

restricted to a maximum intake of 3 dd 10 ml for children of 20 kg and more.

This maximum LA intale was matched with a maximum daily TG supplementation of

4.6 ml for children over 20 kg. As a result, both TG-and PL-supplemented children on

average received 2.2±0.5 g LA per day extra.

Shelf life investigation of both EFA-supplements, including fatty acid profile analyses

at baseline and at 2-weekly intervals for 3 months, before and after sterilization

(3 hours at 140°C), revealed no alterations in fatty acid composition (data not shown).

The pH of both supplements remained stable at 6.0 for 3 months and no micro-

biological contamination occurred.

Analytical techniquesFatty acid analysis

EDTA-plasma, platelets and erythrocytes (RBC) were separated by centrifugation (10

min at 2500 g). Platelet-poor plasma was stored at -80°C. The buffy coat (WBC) was

removed from the RBC pellet, and RBC were washed thrice with physiological saline.

To prevent fatty acid oxidation, erythrocyte membrane lipids were hydrolyzed and

transmethylated to fatty acid methyl esters in methanol/HCl (5:1 vol:vol) and

extracted in hexane the same day(15). Aliquots of the TG and PL supplements were

dissolved in chloroform/methanol (1:2), and lipids were hydrolyzed, methylated and

extracted for fatty acid analysis as described above. Heptadecanoic acid (C17:0) was

added to all samples as internal standard prior to methylation and extraction

136

Chapter 7


procedures, and BHT was added as antioxidant. Fatty acid methyl esters were

separated and quantified by gas liquid chromatography (GLC) on a Hewlett Packard

gas chromatograph model 6890, with a 50mx0.2mm Ultra 1 capillary column

(Hewlett Packard, Palo Alto, CA) and a FID detector as described previously(8).

Plasma liver enzyme activities and vitamin A and E concentrations were measured

using standard clinical laboratory procedures.

CalculationsRelative concentrations (mol%) of individual fatty acids in RBC membranes and in the

EFA supplements were calculated by summation of all fatty acid peak areas and sub-

sequent expression of individual fatty acid peaks as a percentage of this amount.

Fatty acid contents were quantified by relating the areas of their chromatogram

peaks to that of the internal standard heptadecanoic acid (C17:0). Fatty acid and liver

enzyme concentrations were corrected for inter- and intra-individual variations in time

between blood sampling, by normalization to 30 days between each measuring

point. Occasional missing values from time points of one or two months after

starting EFA supplementation were interpolated. Correlations between fatty acid

concentrations and age have been described previously(16), but since there was no

significant difference in age and bodyweight between the study groups, fatty acid

concentrations were not corrected for these parameters. Intake of dietary EFA was

calculated from food diaries from the participants by a clinical dietician, using the

Netherlands Nutrients Table "NEVO" 2001. Intakes were expressed in g/day, and

compared to the RDI for children as determined by the Netherlands Nutrition Council.

StatisticsAll results are presented as means ± S.D. for the number of patients indicated. Data

were statistically analyzed using Student t-test, or, for comparison of more than two

groups, ANOVA-test with post-hoc Bonferroni correction. Statistical significance of

differences between means was accepted at p<0.05. Analyses were performed

using SPSS for Windows software (SPSS, Chicago, IL).

RESULTS

Patient characteristicsFigure 2 shows age, bodyweight, sex and underlying hepatic disease of the

included ESLD patients. Significant correlations existed between age and weight and

baseline RBC-concentrations of linoleic acid (LA), alpha-linolenic acid (ALA), docosa-

hexaenoic acid (DHA) but not arachidonic acid (AA) (p<0.05, data not shown).

137



The underlying cause for end-stage liver disease was biliary atresia in 59% of

patients, and cirrhosis by various causes (primary sclerosing cholangitis, progressive

familial intrahepatic cholestasis, alpha-1-antitrypsin deficiency, auto-immune

hepatitis, cystic fibrosis) in the remaining 41%. In the PL supplemented group, one

patient failed to complete the three-month supplementation period due to perceived

impalatability of the supplement. In the TG supplemented group, one patient did not

complete the supplementation period due to liver transplantation. No complaints of

nausea or other gastrointestinal symptoms due to the supplements were reported by

any of the patients. The groups did not differ in mean age or bodyweight.

Age: Bodyweight:

Non-supplemented 6.8 ± 5.6 27.9 ± 19.7

TG 7.1 ± 5.8 28.9 ± 20.1

PL 9.6 ± 5.5 35.5 ± 21.9

138

Chapter 7

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

PFIC2

PFIC2

biliary atresia

auto-immune hepatits

biliary atresia

biliary atresia

biliary atresia

biliary atresia

PFIC2

PFIC2

biliary atresia

biliary atresia

tyrosemia type 1

nonsyndromatic bile duct paucity

cystic fibrosis, cirrhosis

biliary atresia

biliary atresia

congenital intrahepatic cholestasis

primary sclerosing cholangitis

acute liverfailure, M. Wilson

biliary atresia

TPN cholestasis


cystic intra+extrahepatic bile ducts

Alagille syndrome

biliary atresia


TG1 (m)

TG2 (m)

TG3 (f)

TG4 (m)

TG5 (m)

TG6 (f)

TG7 (f)

TG8 (m)

TG9 (f)

TG10 (m)

PL1 (m)

PL2 (m)

PL3 (f)

PL4 (m)

PL5 (m)

PL6 (f)

PL7 (f)

PL8 (f)

PL9 (m)

X1 (m)

X2 (m)

X3 (m)

X4 (m)

X5 (f)

X6 (f)

X7 (f)

X8 (m)

X9 (f)

X10 (m)

X11 (f)

X12 (f)

X13 (m)

X14 (f)

X15 (m)

X16 (m)

X17 (f)

X18 (f)

X19 (f)

X20 (m)

X21 (f)

X22 (m)

X23 (f)

X24 (m)

X25 (m) biliary atresia

3

1

2

2

9

10

13

16

16

7

2

13

11

6

16

14

2

2

9

1

10

6

16

15

13

10

14

3

1

5

14

13

1

3

9

1

0

0

11

1

age

auto-immune hepatits13 53

14

8

15

11

35

35

50

60

auto-immune hepatits13 53

52

22

15

50

27

19

73

53

14

15

35

10

27

19

73

39

40

26

50

16

9

18

45

66

12

15

31

5

5

6

30

7

weightbiliary atresia2 10

biliary atresia2 10

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

biliary atresia

PFIC2

PFIC2

biliary atresia


biliary atresia

biliary atresia

biliary atresia

biliary atresia

PFIC2

PFIC2

biliary atresia

biliary atresia

tyrosemia type 1

nonsyndromatic bile duct paucity

cystic fibrosis, cirrhosis

biliary atresia

biliary atresia

congenital intrahepatic cholestasis


acute liverfailure, M. Wilson

biliary atresia

TPN cholestasis


cystic intra+extrahepatic bile ducts

Alagille syndrome

biliary atresia


TG1 (m)

TG2 (m)

TG3 (f)

TG4 (m)

TG5 (m)

TG6 (f)

TG7 (f)

TG8 (m)

TG9 (f)

TG10 (m)

PL1 (m)

PL2 (m)

PL3 (f)

PL4 (m)

PL5 (m)

PL6 (f)

PL7 (f)

PL8 (f)

PL9 (m)

X1 (m)

X2 (m)

X3 (m)

X4 (m)

X5 (f)

X6 (f)

X7 (f)

X8 (m)

X9 (f)

X10 (m)

X11 (f)

X12 (f)

X13 (m)

X14 (f)

X15 (m)

X16 (m)

X17 (f)

X18 (f)

X19 (f)

X20 (m)

X21 (f)

X22 (m)

X23 (f)

X24 (m)

X25 (m) biliary atresia

3

1

2

2

9

10

13

16

16

7

2

13

11

6

16

14

2

2

9

1

10

6

16

15

13

10

14

3

1

5

14

13

1

3

9

1

0

0

11

1

age

auto-immune hepatits13 53 auto-immune hepatits13 53

14

8

15

11

35

35

50

60

auto-immune hepatits13 53 auto-immune hepatits13 53

52

22

15

50

27

19

73

53

14

15

35

10

27

19

73

39

40

26

50

16

9

18

45

66

12

15

31

5

5

6

30

7

weightbiliary atresia2 10 biliary atresia2 10

biliary atresia2 10 biliary atresia2 10

Figure 2: Age, bodyweight, sex and underlying hepatic disease ofESLD patients not supplemented with EFA or supplemented withEFA as TG or PL.


Fatty acid composition of erythrocyte membranesFigure 3 shows the effects of non-supplementation (control) and EFA supplementa-

tion as TG or PL on RBC linoleic acid (LA, C18:2n-6), alpha-linolenic acid (ALA,

C18:3n-3) and their respective metabolites, arachidonic acid (AA, C20:4n-6) and

docosahexaenoic acid (DHA, 22:6n-3), during the study period. Significant differ-

ences in concentrations of linoleic acid (LA) could not be detected between the three

groups at baseline, nor any time point during supplementation, but within each

group, remarkable changes were observed. RBC-LA concentration decreased in the

non-supplemented group over the 12 week study period (-0.9 mol%, p<0.05), and

was stable in the TG group (+0.3 mol%, NS). Yet, the increase in LA was signifi-

cantly higher in the PL supplemented group (+1.2 mol%) than in the non-

supplemented group (p<0.005). Accordingly, the calculated monthly change in

RBC-LA concentration was significantly higher in the PL- compared with the non-

supplemented group (+0.45% vs. -0.39%, p<0.01; TG group, +0.09 mol%, NS).

Concentrations of ALA and DHA slightly increased in both the TG- and PL-

supplemented groups, and decreased in the non-supplemented patients, but the

differences were not statistically significant. Arachidonic acid (AA) concentrations

remained remarkably stable over time within and between the three groups (Fig. 3d).

139


3a

3c

AA (20:4n-6)

mol

% o

f tot

al fa

tty a

cids

0

5

10

15

0 4 8 12time (weeks)

LA (18:2n-6)

mol

% o

f tot

al fa

tty a

cids

0

5

10

15

0 4 8 12

*

*

time (weeks)

PLnon

TG

ALA (18:3n-3)

time (weeks)

mol

% o

f tot

al fa

tty a

cids

0 4 8 120.0

0.1

0.2

0.3

0.4

0.5

DHA (22:6n-3)

mol

% o

f tot

al fa

tty a

cids


0

1

2

3

4

5

3b

3d

delta LA (t3-t0)

-3.0

-2.0

-1.0

0.0

1.0

2.0

3.0

mo

l%

TGPL

no supplement

*

3a

3c

AA (20:4n-6)

mol

% o

f tot

al fa

tty a

cids

0

5

10

15


LA (18:2n-6)

mol

% o

f tot

al fa

tty a

cids

0

5

10

15

0 4 8 12

*

*

time (weeks)

PLnon

TG

ALA (18:3n-3)

time (weeks)

mol

% o

f tot

al fa

tty a

cids

0 4 8 120.0

0.1

0.2

0.3

0.4

0.5

DHA (22:6n-3)

mol

% o

f tot

al fa

tty a

cids


0

1

2

3

4

5

3b

3d

delta LA (t3-t0)

-3.0

-2.0

-1.0

0.0

1.0

2.0

3.0

mo

l%

TGPL

no supplement

*

AA (20:4n-6)

mol

% o

f tot

al fa

tty a

cids

0

5

10

15


AA (20:4n-6)

mol

% o

f tot

al fa

tty a

cids

0

5

10

15


LA (18:2n-6)

mol

% o

f tot

al fa

tty a

cids

0

5

10

15

0 4 8 12

*

*

time (weeks)

LA (18:2n-6)

mol

% o

f tot

al fa

tty a

cids

0

5

10

15

0

5

10

15

0 4 8 120 4 8 12

*

*

time (weeks)

PLnon

TGPLnonnon

TG

ALA (18:3n-3)

time (weeks)

mol

% o

f tot

al fa

tty a

cids

0 4 8 120.0

0.1

0.2

0.3

0.4

0.5

DHA (22:6n-3)

mol

% o

f tot

al fa

tty a

cids


0

1

2

3

4

5

3b

3d

ALA (18:3n-3)

time (weeks)

mol

% o

f tot

al fa

tty a

cids

0 4 8 120.0

0.1

0.2

0.3

0.4

0.5ALA (18:3n-3)

time (weeks)

mol

% o

f tot

al fa

tty a

cids

0 4 8 120.0

0.1

0.2

0.3

0.4

0.5

DHA (22:6n-3)

mol

% o

f tot

al fa

tty a

cids


0

1

2

3

4

5

3b

3d

delta LA (t3-t0)

-3.0

-2.0

-1.0

0.0

1.0

2.0

3.0

mo

l%

TGPL

no supplement

*delta LA (t3-t0)

-3.0

-2.0

-1.0

0.0

1.0

2.0

3.0

-3.0

-2.0

-1.0

0.0

1.0

2.0

3.0

mo

l%

TGPL

no supplementTGTGPLPL

no supplementno supplement

*

Figure 3: Concentrations of linoleic acid, alpha linolenic acid, arachidonic acid and docosahexaenoic acid in red blood cells ofchildren with end-stage liver disease, recieving EFA-rich PL (open circles), EFA-rich TG (grey circles) or no EFA supplement(black triangles) for 3 months. Fatty acid concentrations are in mol% of total fatty acids. Inserted is the increase in RBC LA frombaseline until 3-months of suppletion. Data represent means±SD of 9-25 patients per group. *p<0.05


No significant correlation existed between baseline RBC-LA level and monthly

change in RBC-LA concentration in any of the groups (Figure 4).

Figure 5 shows markers for EFA-deficiency, such as the sum of n-6 fatty acids (n-6

deficiency), C20:3n-9 (also called mead acid, indicating n-6 deficiency) and the ratio

between C22:5n-6 and C20:4n-6 (n-3 deficiency). Total n-6 fatty acids significantly

decreased and mead acid significantly increased in the non-supplemented group,

indicating development of n-6 and n-3 deficiency, whereas in both supplemented

groups these parameters stabilized. A mild deficiency of n-3 fatty acids, defined by a

C22:5n-6 to C20:4n-6 ratio > 0.068(16), was and remained present in all groups.

Estimation of desaturase activityFor estimation of hepatic desaturase activities in chronic cholestatic patients, we

calculated the ratio of C20:4n-6 to C20:3n-6 (delta-5-desaturase), C22:5n-6 to

C24:4n-6 (delta-6-desaturase) and C22:6n-3 to C22:5n-3 (delta-6-desaturase).

Based on these ratios, the estimated desaturase activities were not significantly dif-

ferent between supplemented or non-supplemented patients (Figure 6), and did not

change over time during the three-month study period (data not shown).

140

Chapter 7

del

ta m

ol%

LA

per

mo

nth

del

ta m

ol%

LA

per

mo

nth

del

ta m

ol%

LA

per

mo

nth

non-supplemented

-2.0

-1.0

0.0

1.0

2.0

5 10 15 20

R2=0.0577

mol% LA

-2.0

-1.0

0.0

1.0

2.0

5 10 15 20

R2=0.0026

PL

mol% LA

-2.0

-1.0

0.0

1.0

2.0

5 10 15 20

R2=0.004

TG

mol% LA

del

ta m

ol%

LA

per

mo

nth

del

ta m

ol%

LA

per

mo

nth

del

ta m

ol%

LA

per

mo

nth

del

ta m

ol%

LA

per

mo

nth

del

ta m

ol%

LA

per

mo

nth

del

ta m

ol%

LA

per

mo

nth

non-supplemented

-2.0

-1.0

0.0

1.0

2.0

5 10 15 20

R2=0.0577

mol% LA

non-supplemented

-2.0

-1.0

0.0

1.0

2.0

5 10 15 20

R2=0.0577

mol% LA

-2.0

-1.0

0.0

1.0

2.0

5 10 15 20

R2=0.0026

PL

mol% LA

-2.0

-1.0

0.0

1.0

2.0

5 10 15 20

R2=0.0026

PL

mol% LA

-2.0

-1.0

0.0

1.0

2.0

5 10 15 20

R2=0.004

TG

mol% LA

-2.0

-1.0

0.0

1.0

2.0

5 10 15 20

R2=0.004

TG

mol% LA

Figure 4: Monthly change in RBC-LA was not signif-icantly correlated to baseline RBC LA levels.


Dietary fat and LA intakeBasal dietary intake of linoleic acid (LA), saturated fatty acids (SAFA), mono-

unsaturated fatty acids (MUFA) and polyunsaturated fatty acids (PUFA) were not

significantly different between PL, TG or non-supplemented cholestatic children

(Figure 7). When expressed per kg bodyweight, dietary LA intake was well above the

RDI for children (0.5 g/kg), as formulated by the Netherlands Nutrition Council (TG:

0.8±0.5, PL: 0.7±0.5, non-supplemented: 0.8±0.6, NS). Average extra LA intake via

the supplement was 2.2±0.5 g per day for each supplementation group (NS), i.e.,

0.082 and 0.087 g extra LA per kg bodyweight per day for the TG- and the PL-group,

respectively (NS).

Liver enzymesSerum markers for cholestasis and liver failure did not differ between supplemented

and non-supplemented children at baseline (Figure 8), and did not alter significantly

141


n-6 fatty acids

mol

% o

f tot

alFA

25

30

35

0 0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45Mead acid (20:3n-9)

mol

% o

f tot

alFA

*

t0 t3 t0 t3t0 t3t0 t3 t0 t3t0 t3

*

0.00

0.02

0.04

0.06

0.08

0.10

0.1222:5n-6 / 20:4n-6

mol

/ mol

t0 t3 t0 t3t0 t3

TG

PL

no supplement

n-6 fatty acidsm

ol%

of t

otal

FA

25

30

35

0 0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45Mead acid (20:3n-9)

mol

% o

f tot

alFA

*

t0 t3 t0 t3t0 t3t0 t3t0 t3 t0 t3t0 t3t0 t3t0 t3t0 t3 t0 t3t0 t3t0 t3t0 t3 t0 t3t0 t3t0 t3t0 t3

*

0.00

0.02

0.04

0.06

0.08

0.10

0.1222:5n-6 / 20:4n-6

mol

/ mol

t0 t3 t0 t3t0 t30.00

0.02

0.04

0.06

0.08

0.10

0.12

0.00

0.02

0.04

0.06

0.08

0.10

0.1222:5n-6 / 20:4n-6

mol

/ mol

t0 t3 t0 t3t0 t3t0 t3t0 t3 t0 t3t0 t3t0 t3t0 t3

TG

PL

no supplement

TGTG

PLPL


Figure 5: Total n-6 fatty acids, Mead acid (C20:3n-9) and the ratio of C22:5n-6 tp C20:4n-6) in red blood cells (RBC) of childrenwith end-stage liver disease fed EFA-rich PL (white bars), EFA-rich TG (grey bars) or no EFA (black bars) for 3 months.Individual fatty acid concentrations are expressed as molar percentages of total fatty acids. Data represent means ± SD of 9-25 patients per group. *p<0.05

desaturase activity

mo

l / m

ol

0

2

4

6

8

10

20:4n-6 20:3n-6

22:6n-3 22:5n-3

22:5n-6 22:4n-6

0.0

0.1

0.2

0.3

0.4

0.5

0.6TG

PL

no supplement

desaturase activity

mo

l / m

ol

0

2

4

6

8

10

20:4n-6 20:3n-6

22:6n-3 22:5n-3

22:5n-6 22:4n-6

0.0

0.1

0.2

0.3

0.4

0.5

0.6

mo

l / m

ol

0

2

4

6

8

10

0

2

4

6

8

10

20:4n-6 20:3n-620:4n-6 20:3n-6

22:6n-3 22:5n-322:6n-3 22:5n-3

22:5n-6 22:4n-622:5n-6 22:4n-6

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.0

0.1

0.2

0.3

0.4

0.5

0.6TG

PL

no supplement

TGTG

PLPL


Figure 6: Activities of hepatic desaturases in chron-ic cholestatic patients as estimated by the ratios ofC20:4n-6 to C20:3n-6 (delta-5-desaturase),C22:5n-6 to C24:4n-6 (delta-6-desaturase) andC22:6n-3 to C22:5n-3 (delta-6-desaturase). Ratioswere not significantly different between supplement-ed or non-supplemented patients and did notchange over time during the 3-month study period

5a 5b 5c

6


during the three-month study period (not shown). Similarly, plasma vitamin A and

vitamin E concentrations remained at a stable level during the study, and did not

differ significantly between the three groups (Figure 9).

DISCUSSION

We compared the efficacy of oral phospholipids (PL) and oral triglycerides (TG) as

vehicles for EFA, for prevention and correction of EFA deficiency in children with end-

stage liver disease (ESLD). We hypothesized that PL would be more effective than

TG for oral EFA supplementation in cholestatic liver disease, since dietary TG are

markedly malabsorbed during cholestasis, and intestinal PL absorption is relatively

bile-independent. In addition, PL have been postulated to have a higher post-

absorptive bioavailability for target organs of EFA and their long-chain poly-

unsaturated metabolites, such as brain and liver(14;17;18). Our results show a subtle

superiority of oral PL compared to TG on monthly increase of erythrocyte-LA.

142

Chapter 7

Baseline dietary intake

0

10

20

30

40

50

LA SAFA MUFA PUFA

gra

m /d

ay

TG

PL

no supplement

Baseline dietary intake

0

10

20

30

40

50

0

10

20

30

40

50

LA SAFA MUFA PUFA

gra

m /d

ay

TG

PL

no supplementTGTG

PLPL


Figure 7: Basal dietary intake of linoleic acid (LA),saturated fatty acids (SAFA), monounsaturated fattyacids (MUFA) and polyunsaturated fatty acids(PUFA), calculated in grams per day from 2-day con-secutive food diaries, were not significantly differentbetween PL, TG or non-supplemented cholestaticchildren.

AP (U/l)

LDH (U/l)

ASAT (U/l)

ALAT (U/l)

bili tot (umol/l)

alb (g/l)

gGT (U/l)

PT (sec)

APTT (sec)

fibrinogeen (g/l)

AT (%)

570 ± 271

304 ± 96

133 ± 109

90 ± 63

127 ± 157

34.9 ± 6.9

196 ± 200

18 ± 7

37 ± 10

2.7 ± 0.8

82 ± 29

729 ± 259

312 ± 61

176 ±256

107 ± 91

97 ± 109

32.2 ± 6.3

238 ± 176

16 ± 5

34 ± 7

2.7 ± 0.6

85 ± 36

non TG PL

753 ± 307

288 ± 63

125 ± 73

85 ± 48

137 ± 164

34.6 ± 7.9

206 ± 194

17 ± 5

36 ± 5

2.6 ± 0.7

85 ± 41

AP (U/l)

LDH (U/l)

ASAT (U/l)

ALAT (U/l)

bili tot (umol/l)

alb (g/l)

gGT (U/l)

PT (sec)

APTT (sec)

fibrinogeen (g/l)

AT (%)

570 ± 271

304 ± 96

133 ± 109

90 ± 63

127 ± 157

34.9 ± 6.9

196 ± 200

18 ± 7

37 ± 10

2.7 ± 0.8

82 ± 29

729 ± 259

312 ± 61

176 ±256

107 ± 91

97 ± 109

32.2 ± 6.3

238 ± 176

16 ± 5

34 ± 7

2.7 ± 0.6

85 ± 36

non TG PLnon TG PL

753 ± 307

288 ± 63

125 ± 73

85 ± 48

137 ± 164

34.6 ± 7.9

206 ± 194

17 ± 5

36 ± 5

2.6 ± 0.7

85 ± 41

vit A

0.0

0.5

1.0

1.5

2.0

TG PLnon

vit E

0

10

20

30

40

50

TG PLnon

vit A

0.0

0.5

1.0

1.5

2.0

TGTG PLPLnonnon

vit E

0

10

20

30

40

50

0

10

20

30

40

50

TGTG PLPLnonnon

Figure 8: Serum markers for cholestasis and liver failure did not dif-fer between supplemented and non-supplemented children.

Figure 9: Plasma vitamin A and vitamin E concen-trations remained at a stable level during the 3-month study period, and did not differ significantlybetween TG-, PL-, and non-supplemented children.

TG

PL

no supplement

TG

PL

no supplement

TGTG

PLPL



Between 1982 and 2000, 180 liver transplantations were performed in Groningen, in

136 children(29). A high incidence of EFA deficiency in these children with ESLD has

frequently been described, but due to the aspecificity of symptoms and the pro-

longed subclinical course, EFA deficiency is a biochemical rather than a clinical

diagnosis. Still, no consensus exists regarding validated "gold standard" cut-off

values that clearly define EFA deficiency in children. Previous studies in our hospital

demonstrated that 75-80% of chronic cholestatic children had various indications for

EFA deficiency, i.e., n-3 deficiency, n-6 deficiency or both(3). In the present study we

included patients irrespective of the EFA status at base line. One could argue that

supplementation would only be indicated for patients with a biochemically marginal

or deficient EFA status. However, biochemical analysis of RBC-EFA of included

patients indicated marginal EFA status in all groups at baseline. Furthermore,

baseline concentration of RBC-LA was not significantly correlated with its monthly

increase of decrease (with or without supplementation), which in retrospect seems to

validate random inclusion for supplementation.

Theoretically, chronic cholestasis may impair activities of hepatic desaturation and

elongation enzymes, which synthesize AA and DHA from their respective precursors

LA and ALA. Previous studies in short-term cholestatic rats showed no indications for

altered desataration/elongation capacity(4). In this study in children with chronic

cholestasis, ratios between C20:3n-6 and C20:4n-6, marking delta-5 desaturase

activity, and between C22:5n-3 and C22:6n-3 as a marker for delta-6 desaturase

activity, were not significantly different between TG-, PL-, or non-supplemented

cholestatic children. These estimates do not support a major difference in EFA

metabolism among the groups. It is not quite clear, however, how these values relate

to non-diseased controls. Also, the contribution of dietary LCPUFA intake on these

ratios is unclear, preventing strong conclusions on the competence of EFA

metabolism in these children. Liver enzymes and plasma fat-soluble vitamin

concentrations remained stable during the three-month study period and did not

differ significantly between groups.

Characteristic biochemical hallmarks of EFA deficiency(19;20) in RBC fatty acid profiles,

such as decreased total n-6 fatty acids and increased C20:3n-9 concentrations,

significantly deteriorated over time in the non-supplemented patients (Figure 5). The

TG and PL supplements did not differentially affect LA, ALA, DHA or AA levels in RBC

membranes after three months of supplementation. Interestingly, however, the PL

supplement clearly induced an positive monthly increase in RBC LA, in contrast to

the TG supplement, despite the fact that overall LA ingestion in the former tended to

143



be lower (TG vs. PL, 0.88 vs. 0.79 g/kg bodyweight/day, respectively). Thus, subtle

changes in favor of EFA supplementation in the form of PL can be derived from the

results, although it needs to be restated that the changes did not materialize into

significant differences upon group-wise comparisons over the three-month study

period. Several explanations are possible for this paradox. Firstly, considering the

long half-life of RBC, the supplementation period may have been too short to observe

profound differences between the groups in RBC fatty acids. It can not be excluded

that upon prolonged supplementation, the rate of LA increase in the PL-group could

have resulted in significantly higher RBC LA contents. Secondly, the dosage of sup-

plemented LA could have been too low. Originally, we aimed to supply the patients

with an extra 50% LA of the recommended daily intake (RDI) for children. However,

the high viscosity of the PL supplement (soybean lecithin) required emulsification

into an oil-in-water solution that resulted in a strong increase in the volume of sup-

plement to be ingested daily. The maximum ingestible volume of supplement for the

children was 30 ml per day, and therefore we decreased the dose of extra LA from

50% to 25% of RDI (i.e., 0.125 g LA/kg/day), and restricted the volume of supplement

to a daily maximum of 30 ml for children above 20 kg. Average daily dietary LA inges-

tion (apart from the supplements) of the cholestatic children was well above the RDI

of 0.5 g/kg/day, presumably due to clinical dietary counseling. Since baseline dietary

LA intake was 0.75±0.35 g/kg/day, and 11 of 19 supplemented patients weighed

more than 20 kg, the actual extra supply of LA in the cholestatic patients as TG/PL

supplement eventually was 15±7% of dietary intake. Possibly, this amount may have

been insufficient to result in a pronounced increase in RBC-EFA. Yet, even at this rel-

atively low supplementation dose, a positive effect of PL-supplementation could be

observed. Strandvik et al. supplemented cystic fibrosis patients with even smaller

doses of LA (30-50 mg/kg/day)(21), which significantly improved serum EFA status in

these patients. However, supplementation in this study was intravenous, and the

effects were evaluated after one year of supplementation.

Several authors have questioned the concept that the erythrocyte membrane fatty

acid composition represents a reliable reflection of overall body EFA status. Due to

their long half-life and short-term independence of post-prandial plasma fatty acid

concentrations, erythrocytes are regarded as a stable and easily accessible com-

partment for evaluation of body EFA status. However, studies by Korotkova et al.(22)

demonstrated that phospholipid fatty acid profiles of erythrocytes, serum, jejunum,

ileum and colon were differentially affected by EFA deficiency in rats. Similarly, Rioux

et al.(23) reported that in piglets, the plasma phospholipid fatty acid profile ade-

quately reflects that of liver and bile, but not of brain phospholipids.

In EFA-deficient cholestatic mice, we previously observed that oral EFA-PL adminis-

144

Chapter 7


tration improved LCPUFA concentrations in brain and liver more efficiently than EFA-

TG, but remarkably, EFA concentrations in erythrocytes were equally improved by TG

and PL(13). These results suggest that after absorption, specific targeting occurs of

LCPUFA to organs as the central nervous system, liver or gut, at the expense of less

critical tissues such as erythrocytes. The brain has been postulated to preferentially

absorb LCPUFA as lyso-PL, in contrast to unesterified LCPUFA bound to albumin, or

LCPUFA esterified into lipoproteins(24-28). Since dietary PL are partly absorbed as lyso-

PL and partly as intact PL molecules, the highly efficient increase in brain LCPUFA

levels in mice after PL supplementation supports the postulated high bioavailability

of enteral PL for LCPUFA target organs(14;17;18).

We conclude that oral EFA supplementation in the form of PL is slightly more

effective than as TG for prevention or correction of EFA deficiency in children with

end-stage liver disease.

REFERENCES1. Robberecht E, Koletzko B, Christophe A. Several mechanisms contribute to the abnormal fatty acid composition of serum PLand cholesterol esters in cholestatic children with extrahepatic biliary atresia. Prostaglandins Leukot.Essent.Fatty Acids1997;56:199-204.2. Socha P, Koletzko B, Swiatkowska E, Pawlowska J, Stolarczyk A, Socha J. Essential fatty acid metabolism in infants withcholestasis. Acta Paediatr. 1998;87:278-83.3. Sealy MJ, Muskiet FAJ, Martini IA et al. Essentiele vetzuurdeficientie bij pediatrische patienten. Tijdschrift voorKindergeneeskunde 1997;65:144-50.4. Minich DM, Havinga R, Stellaard F, Vonk RJ, Kuipers F, Verkade HJ. Intestinal absorption and postabsorptive metabolism oflinoleic acid in rats with short-term bile duct ligation. Am.J.Physiol.2000.Dec.;279.(6.):G1242.-8. 2000;279:G1242-G1248.5. Tso P. Intestinal lipid absorption. In: Tso P, ed. Physiology of the gastrointestinal tract. New York: Raven Press 1994:1867-907.6. Carey MC, Hernell O. Digestion and absorption of fat. Seminars in gastrointestinal disease 1992;3:189-208.7. Levy E, Garofalo C, Thibault L et al. Intraluminal and intracellular phases of fat absorption are impaired in essential fatty aciddeficiency. Am.J.Physiol. 1992;262:G319-G326.8. Werner A, Minich DM, Havinga R et al. Fat malabsorption in essential fatty acid-deficient mice is not due to impaired bile for-mation. Am.J.Physiol 2002;283:G900-G908.9. Miyano T, Yamashiro Y, Shimizu T, Arai T, Suruga T, Hayasawa H. EFA deficiency in congenital biliary atresia: successful treat-ment to reverse deficiency. J.Pediatr.Surg. 1986;21:277-81.10. Yamashiro Y, Ohtsuka Y, Shimizu N et al. Effetcs of ursodeoxycholic acid treatment on essential fatty acid deficiency inpatients with biliary atresia. Pediatr.Surg. 1994;29:425-8.11. Carey MC, Small DM, Bliss CM. Lipid digestion and absorption. Ann.Rev.Physiol. 1983;45:651-77.12. Carey MC, Small DM. The characteristics of mixed micellar solutions with particular reference to bile. Am.J.Med.1970;49:590-608.:590-608.13. Werner A, Havinga R, Kuipers F, Verkade HJ. Treatment of EFA deficiency with dietary triglycerides or phospholipids in amurine model of extrahepatic cholestasis. Am.J.Physiol Gastrointest.Liver Physiol 2004;286:G822-G832.14. Carnielli VP, Verlato G, Pederzini F et al. Intestinal absorption of LCPUFA in preterm infants fed breast milk or formula.Am.J.Clin.Nutr. 1998;67:97-103.15. Muskiet FA, van Doormaal JJ, Martini IA, Wolthers BG, van der SW. Capillary gas chromatographic profiling of total long-chain fatty acids and cholesterol in biological materials. J.Chromatogr. 1983;278:231-44.16. Fokkema MR, Smit EN, Martini IA, Woltil HA, Boersma ER, Muskiet FAJ. Assessment of EFA and w3-fatty acid status bymeasurement of RBC 20:3w9 (Mead acid), 22:5w6/22:6w3 and 22:5w6/22:4w6. Prostaglandins Leukot.Essent.Fatty Acids2002;67:345-56.

145



17. Boehm G, Borte M, Boehles HJ, Mueller H, Kohn G, Moro G. DHA and AA content of serum and red blood cell membranephospholipids of preterm infants fed breast milk, standard formula or formula supplemented with n-3 and n-6 long-chain polyun-saturated fatty acids. Eur J Pediatr 1996;155:410-6.18. Wijendran V, Huang MC, Diau GY, Boehm G, Nathanielsz PW, Brenna JT. Efficacy of dietary AA provided as triglyceride orphospholipid as substrates for brain AA accretion in baboon neonates. Pediatr.Res. 2002;51:265-72.19. Yamanaka WK, Clemans GW, Hutchinson ML. Essential fatty acid deficiency in humans. Prog.Lipid Res. 1981;19:187-215.20. Holman RT. The ratio of trienoic:tetraenoic acids in tissue lipids as a measure of EFA requirement. J Nutr 1960;70:405-10.21. Strandvik B, Hultcrantz R. Liver function and morphology during long-term fatty acid supplementation in CF. Liver1994;14:32-6.22. Korotkova M, Strandvik B. EFA deficiency affects the fatty acid composition of the rat small intestinal and colonic mucosadifferently. Biochim.Biophys.Acta 2000;1487:319-25.23. Rioux FM, Innis SM, Dyer R, MacKinnon M. Diet-induced changes in liver and bile but not brain fatty acids can be predict-ed from differences in plasma phospholipid fatty acids in formula- and milk-fed piglets. J.Nutr. 1997;127:370-7.24. Thies F, Pillon C, Moliere P, Lagarde M, Lecerf J. Preferential incorporation of sn-2 lysoPC DHA over unesterified DHA in theyoung rat brain. Am.J.Physiol 1994;267:R1273-R1279.25. Bernoud N, Fenart L, Moliere P et al. Preferential transfer of 2-docosahexaenoyl-1-lysophosphatidylcholine through an invitro blood-brain barrier over unesterified docosahexaenoic acid. J.Neurochem. 1999;72:338-45.26. Spector AA. Plasma free fatty acid and lipoproteins as sources of PUFA for the brain. J.Mol.Neurosci. 2001;16:159-65.27. Dhopeshwarkar GA, Mead JF. Uptake and transport of fatty acids into the brain and the role of the blood-brain barrier sys-tem. Adv.Lipid Res. 1973;11:109-42.28. Qi K, Hall M, Deckelbaum RJ. LCPUFA accretion in brain. Curr.Opin.Clin.Nutr.Metab Care 2002;5:133-8.29.Sieders E, Peeters PM, TenVergert EM, deJong KP, Porte RJ, Zwaveling JH, Bijleveld CMA, Slooff MJ. Prognostic factors forlong-term actual patient survival after orthotopic liver transplantation in children. Transplantation 2000;70:1448-53

146

Chapter 7


General disussion and summary8proefschrift_def_v010605def.qxp 2-6-2005 1:31 Pagina 147

SUMMARY

This thesis focuses on the role of phospholipids (PL) in absorption and metabolism

of essential fatty acids (EFA), under physiological and bile-deficient conditions. EFA

cannot be synthesized de novo by the body, but are crucial for normal function and

development, either as such or after metabolization into long-chain polyunsaturated

fatty acids (LCPUFA). EFA deficiency is associated with dietary fat malabsorption,

growth retardation, steatosis and impaired neurological development, although

symptoms can be rather aspecific and only become apparent after an extended

subclinical course. Children have limited adipose tissue stores of EFA and high EFA

requirements during growth and development. These characteristics make children

particularly dependent on adequate supply and absorption of EFA via the diet.

Dietary EFA are predominantly esterified into triglycerides (TG), which are

profoundly malabsorbed during impaired enteral lipolysis or solubilization, as in

cystic fibrosis or cholestasis, respectively. EFA esterified into PL are more easily

absorbed under these conditions, since the polar PL molecules do not require bile

for solu-bilization in an aqueous environment, and can partially be absorbed intact,

without (phopspho-)lipolysis(6; 22). Additionally, PL have a high post-absorptive

availability for the body(8; 10; 11).

We aimed to characterize the impact of EFA deficiency on intestinal and liver function

and to develop a dietary treatment strategy for prevention or correction of EFA

deficiency in susceptible conditions, like cholestasis and cystic fibrosis.

Fat malabsorption in EFAD mice is not due to impairedbile formationIn chapter 2, we investigated whether EFA deficiency-induced fat malabsorption in

mice is mediated by impaired bile production. Bile salts and bile PL facilitate dietary

fat absorption by enabling intraluminal solubilization and by providing surface coat

material for CM formation, respectively. In rats, EFA deficiency does not affect lipo-

lysis, but decreases bile secretion(17; 20). In mice, however, we demonstrated that bile

production was profoundly increased during EFA deficiency, and bile salt composi-

tion was unchanged, excluding impaired bile production as a cause for EFA

deficiency-induced fat malabsorption. To specify the role of biliary PL secretion in fat

malabsorption, we compared the effects of EFA deficiency in normal and in geneti-

cally modified mice that secrete PL-free bile (Mdr2 -/- mice). If alterations in secreted

bile PL were involved in EFA deficiency-associated fat malabsorption, fat absorption

in Mdr2 -/- mice should be unaffected by EFA depletion. However, fat absorption and

148

Chapter 8


bile flow were equally affected by EFA deficiency in Mdr2 -/- and wildtype mice.

Our data indicate that fat malabsorption during EFA deficiency in mice is not due to

decreased bile production but, by inference, is more likely related to impaired intra-

cellular processing of dietary fat in enterocytes, possibly due to EFA depletion of

plasma- and/or microsomal membranes.

EFA deficiency in mice is associated with steatosis and secretionof large VLDLIn addition to fat malabsorption and increased bile secretion, EFA deficiency in mice

is associated with hypotriglyceridemia and hepatic steatosis. In chapter 3 we

evaluated whether impaired hepatic VLDL secretion contributes to these metabolic

consequences of EFA deficiency in mice. Both in vivo and in vitro, EFA deficiency

increased hepatic TG levels, but hepatic VLDL-TG secretion rate was quantitatively

normal. Interestingly, EFA deficiency induced hepatic production of remarkably large

VLDL particles compared to the non-deficient condition.

EFA are suppressors of lipogenesis via down-regulation of the transcription factor

SREBP1c, therefore, hepatic synthesis of non-EFA can increase during EFA

deficiency. The observed hepatic accumulation of TG in EFA-deficient mice may

additionally be related to increased VLDL clearance. Indeed, hepatic expression of

apoAV and apoCII, genes involved in VLDL catabolism, was increased. Although HL

and LPL activities were normal in EFA-deficient mice, large VLDL particles may be

subject to increased clearance rates. As EFA depletion affects the physicochemical

and biological characteristics of PL bilayer membranes, it may similarly influence PL

monolayers on the surface of nascent lipoproteins. Both lipoprotein size, i.e.,

curvature of the lipoprotein surface, and low EFA content can disturb the physical

VLDL surface structure, increasing accessibility to lipases or affinity for apoC-II or

apoA-V. Steatosis and hypotriglyceridemia in EFA-deficient mice could be a com-

bined result of increased hepatic lipogenesis, unimpaired hepatic VLDL-TG secretion

and production of large VLDL particles that may be subject to rapid clearance.

We hypothesize that the effects of EFA deficiency on VLDL particle size are related to

PL availability for lipoprotein assembly. Increased biliary PL secretion in EFA-deficient

mice (chapter 2) may limit hepatic PL availability for VLDL assembly, inducing

secretion of large lipoprotein particles.

Lymphatic CM size is inversely related to biliary PL secretionin miceWe tested the hypothesis that PL availability determines lipoprotein size in chapter 4,

by investigating whether the size of intestinal lipoproteins is also inversely related to

149

General discussion and summary


PL availability. Since biliary PL secretion is absent in Mdr2 -/- mice and strongly

increased in EFA-deficient mice, we determined lymphatic chylomicron (CM) size

after mesenteric lymph duct cannulation in these two mouse models. CM were

considerably larger during intestinal lack of biliary PL (Mdr2 -/- mice), whereas hyper-

secretion of bile PL into the intestine (EFAD-deficient mice) induced secretion of

smaller CM into lymph. These observations confirmed our hypothesis that PL

availability is a major determinant of lipoprotein size. Since EFA-deficient mice

secrete larger hepatic VLDL than EFA-sufficient controls, secretion of small lipo-

proteins is apparently not an intrinsic feature but rather an organ-specific feature of

EFA deficiency. Similar to the metabolism of VLDL particles, altered CM size and fatty

acid composition affect intravascular processing(13). Large CM may enter the lymph

slower, but are likely to be subsequently metabolized with high efficacy, since large

CM have a greater affinity for lipases and are cleared more rapidly than small parti-

cles. In addition, EFA-rich CM are cleared faster than EFA-depleted CM. This could

partially explain the observation that, although post-prandial plasma appearance of

ingested lipid is delayed both in EFA-deficient and bile-PL-deficient (Mdr2 -/-) mice,

only EFA-deficient mice have net dietary fat malabsorption. This supports the

concept that intraluminal PL are not crucially important for quantitative intestinal

absorption of dietary lipids, but all the more for their absorption kinetics and post-

absorptive metabolism.

No indications for altered EFA metabolism in two murine modelsfor CFBijvelds et al.(5) demonstrated in mouse models for cystic fibrosis (CF) that cftr -/-CAM

mice have a profound dietary fat malabsorption, which could result in EFA

deficiency. Indeed, deficiencies of EFA or LCPUFA are still reported in CF patients,

despite hypercaloric nutrition and pancreas enzyme replacement therapy. It has

been suggested that CFTR malfunction directly affects LCPUFA synthesis,

increasing pro-inflammatory n-6 and decreasing anti-inflammatory n-3 LCPUFA

levels in CF-affected organs, which would contribute to CF symptoms.

Supplementation with n-3 LCPUFA, but not with the precursor ALA, was postulated

to alleviate phenotypic manifestations of the disease.

To determine whether the CF condition merits specific supplementation of EFA

and/or LCPUFA, we analyzed EFA status in two mouse models for CF, with and

without fat malabsorption. In both CF models, we demonstrated that organ LCPUFA

profiles were highly similar to those of healthy sex-matched littermates. In vivo

conversion of 13C-EFA into AA and DHA was unimpaired in CF mice compared to

littermate controls, indicating that impaired LCPUFA synthesis is not an intrinsic

150

Chapter 8


feature of CF phenotype, and that fat malabsorption does not strongly affect EFA

status in CF mice.

Many CF mouse models are available, which display great phenotypic variability,

similar to the situation in CF patients. Apart from the specific CFTR mutation and

environmental influences, phenotypic variability in CF is related to independently

inherited disease-modifying genes, encoding proteins that may partially compensate

for effects of CFTR dysfunction. We demonstrated that diet and age, but primarily

genetic background is an overriding determinant of EFA status in CF mice. In the

studies suggesting that LCPUFA synthesis is controlled by CFTR, thus affecting CF

pathology, the CF mice were compared to control mice of a different mouse strain.

For any meaningful comparison of EFA status between CF mouse models, and

particularly for inferring observations to CF patients, meticulous verification of mouse

genetic backgrounds and use of littermate controls are a prerequisite.

We conclude that CFTR dysfunction does not impair LCPUFA synthesis, and altered

PUFA levels in CF are more likely secondary to inflammation or malnutrition.

Extrapolating these conclusions to CF patients would implicate that sufficient oral

EFA intake should effectively prevent EFA or LCPUFA deficiency in CF. Since fat

malabsorption did not strongly affect EFA status in CF mice, specific EFA supple-

mentation does not seem required in this condition, and we further concentrated on

developing EFA supplementation strategies for patients with cholestatic liver disease.

Oral treatment of EFA deficiency with TG or PL in cholestatic conditionsEFA deficiency is common during cholestasis, due to malabsorption of dietary EFA

which are predominantly esterified into hydrophobic TG molecules. The amphiphilic

PL are more soluble than TG in the aqueous intestinal lumen and more readily

absorbed during bile deficiency. PL are absorbed intact or after digestion to lyso-PL.

In chapter 5, we demonstrated in EFA-deficient mice with acute cholestasis that EFA

supplementation with oral TG or PL was equally effective in preventing decrease of

EFA concentrations in RBC, yet in brain and liver, PL were highly superior to TG and

significantly improved EFA-derived LCPUFA levels. In addition, oral PL prevented

weight loss during EFA deficiency and cholestasis, in contrast to oral TG, supporting

the concept that enteral PL have a facilitating effect on lipid absorption during

bile deficiency.

In chapter 6, we compared the efficacy of oral EFA supplementation as PL and TG

for treatment of EFA deficiency in children with chronic cholestasis, who have a high

incidence of compromised EFA status(16). Three-month supplementation with EFA as

TG or PL prevented the deterioration in RBC LA, mead acid and total n-6 fatty acids

as observed in non-supplemented children. Although we could not directly demon-

151



strate differential effects of TG or PL supplementation on RBC EFA upon group-wise

comparisons, oral PL clearly induced a positive monthly increase in RBC LA

compared to non-supplemented children, whereas oral TG did not.

Interestingly, in cholestatic mice, the differential effects of TG and PL EFA

supplementation during cholestasis were not evident in RBC, but the beneficial

effects of oral PL were highly evident in EFA target organs liver and brain.

In line with Korotkova and Rioux, we question the paradigm that RBC fatty acid

profiles are a tentative index for overall body EFA status. Due to their long half-life and

short-term independence of post-prandial plasma fatty acid levels, RBC may be a

stable and easily accessible compartment for evaluation of body EFA status. Yet it

seems that specific channeling of EFA occurs to target organs (brain, liver, intestine),

at the expense of less critical tissues as RBC. Apart from international consensus on

validated cut-off values defining biochemical EFA-deficiency, there would be great

merit in developing sensitive functional tests to assess deficiency, sufficiency and

requirements of EFA and LCPUFA in patients and animal models(7; 21).

The mechanism underlying the remarkably efficient CNS uptake of LCPUFA is not

fully understood. To a certain extent, body stores may provide EFA for the brain(9), and

although astrocytes are capable of synthesizing AA and DHA from EFA, the plasma

compartment is the main source of brain LCPUFA. In plasma, LCPUFA are present

esterified in lipoproteins, unesterified bound to albumin, or as lyso-PL. The latter are

preferentially incorporated by the brain(4; 18; 19). Dietary PL are absorbed from the intes-

tine intact and as lyso-PL. In addition, oral PL could be a source of plasma (lyso-)PL

as components of HDL-particles, derived from excess CM surface material shed dur-

ing lipolysis. Possibly, PL on the CM surface are preferentially targeted to HDL before

exchanging with other lipoproteins and RBC(6; 15; 22). Either as albumin-bound lyso-PL

or as HDL-PL, oral PL provide a highly accessible source of LCPUFA for the brain,

supporting the concept of high post-absorptive bioavailability of enteral PL.

Obviously, human brain and liver are ethically not accessible for analysis of EFA

composition. Yet in light of the beneficial effects of PL supplementation on brain and

liver in mice, and the slightly better effects of supplementory PL compared to TG on

RBC EFA in cholestatic children, EFA supplementation in the form of PL may still be

preferred compared with TG for prevention or correction of EFA deficiency in children

with end-stage liver disease.

152

Chapter 8


CONCLUSIONS

Before the present studies were performed, it was apparent from existing literature

that EFA deficiency has important consequences for fat absorption and metabolism.

Evidence has been provided that EFA deficiency:

- decreases bile formation in rats;

- does not affect intraluminal lipolysis or enterocyte uptake of dietary fat;

- decreases LA and AA levels in intestinal mucosal membranes;

- leads to accumulation of fat in enterocytes after oral administration

- influences post-absorptive metabolism, i.e., alters plasma lipid profiles, lipoprotein

composition and lipolytic enzymes.

Still, the specific role of EFA deficiency in various steps of fat absorption was poorly

understood. Present results indicate that lipid malabsorption during EFA deficiency

is not due to decreased bile formation nor to altered EFA contents of biliary PL.

Considering the effects of EFA deficiency on lipoprotein formation, we hypothesize

that EFA deficiency-induced fat malabsorption may be due to EFA depletion of PL

membranes, which affects lipoprotein processing in intestine and in liver. In both

organs, the exit step of lipid transport appears to be changed during EFA deficiency;

in both, lipoprotein size is altered and fat accumulation occurs(3). PL synthesis and

membrane incorporation proceeds during deficiency of EFA, as EFA acyl chains are

substituted with available n-9, n-7 or saturated fatty acids. The mechanisms that

determine which fatty acid species are incorporated into PL, or which type of PL is

inserted into lipoprotein monolayers or cell membrane bilayers, are not fully known.

There are strong indications that the chemical structure of dietary fat, TG or PL, is

relatively maintained after intestinal digestion and absorption(1). This observation, in

line with our results on specific targeting of EFA from PL to liver and brain, suggests

that separate intracellular fatty acid pools exist. It will be challenging to elucidate the

mechanisms underlying the fatty acid or PL sorting towards different compartments

in enterocytes or hepatocytes, and the subsequent targeting to specific organs.

Endothelial lipase activity at the blood-brain-barrier has been suggested to play a

role in the specific EFA uptake by the central nervous system, as well as SR-B1-

mediated uptake of HDL-PL EFA by brain capillary endothelial cells(2;12;14).

In cholestatic mice, we clearly demonstrated the preferential capture by the brain of

dietary EFA in the form of PL compared to TG, supporting this form of supplemen-

tation for children with chronic liver disease. In future studies, it would be of great

physiological interest and probably of therapeutic value to elucidate the mechanisms

underlying the specific EFA/LCPUFA accretion by the brain, and the targeting of

dietary EFA-PL towards brain and liver.

153



REFERENCES1. Amate L, Gil A, Ramirez M. Dietary LCPUFA in the form of TG or PL influence lymph lipoprotein size and composition inpiglets. Lipids 37: 975-980, 2002.2. Anderson GJ, Tso PS, Connor WE. Incorporation of CM fatty acids into the developing rat brain. JCI 93: 2764-2767, 1994.3. Bennett Clark S, Ekkers TE, Singh A, Balint JA, Holt PR, Rodgers JB. Fat absorption in EFA deficiency: a model experimen-tal approach to studies of the mechanism of fat malabsorption of unknown etiology. J lipid res 14: 581-588, 1973.4. Bernoud N, Fenart L, Moliere P, Dehouck MP, Lagarde M, Cecchelli R, Lecerf J. Preferential transfer of 2-docosahexaenoyl-1-lysophosphatidylcholine through an in vitro blood-brain barrier over unesterified DHA. J Neurochem 72: 338-345, 1999.5. Bijvelds, M, Hulzebos, C., Bronsveld, H., Havinga, R., Stellaard, F., Sinaasappel, M., De Jonge, H., Verkade, H. J. Fat absorp-tion and enterohepatic circulation of bile salts in CF mice. Gastroenterology 124(4 Suppl 1), A434. 2003. 6. Bloom B, Kiyasu JY, Reinhardt WO and Chaikoff IL. Absorption of plasma PL. Am J Physiol 177: 84-86, 54 A.D.7. Cunnane SC. Problems with EFA: time for a new paradigm? Prog Lipid Res 42: 544-568, 2003.8. Fox JM. Polyene phosphatidylcholine: pharmacokinetics after oral administration - A review. In: Phospholipids and athero-sclerosis, edited by Avogaro P, Macini M, Ricci G and Paoletti R. New York: Raven, 1983.9. Lefkowitz W, Lim SY, Lin Y, Salem N, Jr. Where does the developing brain obtain its DHA? Relative contributions of dietaryALA, DHA, and body stores in the developing rat. Pediatr Res 57: 157-165, 2005.10. Lieber CS, DeCarli LM, Mak KM, Kim CI, Leo MA. Attenuation of alcohol-induced hepatic fibrosis by polyunsaturated lecithin.Hepatology 12: 1390-1398, 1990.11. Lieber CS, Robins SJ, Li J, DeCarli LM, Mak KM, Fasulo JM, Leo MA. Phosphatidylcholine protects against fibrosis and cir-rhosis in the baboon. Gastroenterology 106: 152-159, 1994.12. Presa-Owens S, Innis SM, Rioux FM. Addition of TG with AA or DHA to infant formula has tissue- and lipid class-specificeffects on fatty acids and hepatic desaturase activities in formula-fed piglets. J Nutr 128: 1376-1384, 1998.13. Robins SJ, Fasulo JM, Patton GM. Effect of different molecular species of PC on clearance of emulsion particle lipids. J LipidRes 29: 1195-1203, 1988.14. Scott BL, Bazan NG. Membrane DHA is supplied to the developing brain and retina by the liver. PNAS 86: 2903-2907, 1989.15. Scow RO, Stein Y, Stein O. Incorporation of dietary lecithin and lysolecithin into lymph CM in rats JBC 242: 4919-4924, 1967.16. Sealy MJ, Muskiet FAJ, Martini IA, Volmer M, Boersma ER, Bijleveld RJ, Vonk RJ, Verkade HJ. EFA deficientie bij pediatrischepatienten. Tijdschrift voor Kindergeneeskunde 65: 144-150, 1997.17. Setchell KDR, O'Connell NC. Inborn errors of bile acid biosynthesis. In: Bile acids in gastroenterology, 2000, p. 129-136.18. Thies F, Delachambre MC, Bentejac M, Lagarde M, Lecerf J. Unsaturated fatty acids esterified in 2-acyl-l-lysoPC bound toalbumin are more efficiently taken up by the young rat brain than the unesterified form. J Neurochem 59: 1110-1116, 1992.19. Thies F, Pillon C, Moliere P, Lagarde M, Lecerf J. Preferential incorporation of sn-2 lysoPC DHA over unesterified DHA in theyoung rat brain. Am J Physiol 267: R1273-R1279, 1994.20. Voshol PJ, Minich DM, Havinga R, OudeElferink RPJ, Verkade HJ, Groen AK and Kuipers F. Postprandial chylomicron for-mation and fat absorption in multidrug resistance gene 2 p-glycoprotein-deficient mice. Gastroenterology 118: 173-182, 2000.21. Wesson LG, Burr GO. The metabolic rate and respiratory quotients of rats on a fat-deficient diet. JBC 91: 525-539, 1931.22. Zierenberg O, Grundy SM. Intestinal absorption of polyenephosphatidylcholine in man. J Lipid Res 23: 1136-1142, 1982.

154

Chapter 8


Nederlandse samenvattingDankwoord

Curriculum VitaePublicaties

NLproefschrift_def_v010605def.qxp 2-6-2005 1:32 Pagina 155

SAMENVATTING

In dit proefschrift is de rol onderzocht van fosfolipiden (PL) bij absorptie en meta-

bolisme van essentiële vetzuren (EFA), onder fysiologische condities en bij lever-

aandoeningen. EFA kunnen niet door het lichaam zelf gesynthetiseerd worden, maar

EFA, en hun lange-keten meervoudig onverzadigde metabolieten (LCPUFA) van de

omega-3 of omega-6 serie, zijn wel onmisbaar voor normale functie en ontwikkeling

van het lichaam. EFA deficiëntie is geassocieerd met malabsorptie van voedingsvet,

groeiretardatie, steatose en stoornissen in de neurologische ontwikkeling, maar deze

symptomen zijn aspecifiek worden vaak pas duidelijk na een langdurige subklinische

fase. Kinderen hebben meestal een beperkte EFA voorraad in vetweefsel, en een

hoge EFA behoefte tijdens de groei, waardoor ze bijzonder afhankelijk zijn van ade-

quate inname en absorptie van EFA via de voeding. EFA zijn in voeding vooral aan-

wezig in de vorm van triglyceriden (TG), die slecht geabsorbeerd worden bij aan-

doeningen waarbij lipolyse of solubilizatie van voedingsvet gestoord is, zoals bij cys-

tic fibrosis (CF) of cholestase. EFA kunnen ook in de voeding aanwezig zijn in de

vorm van fosfolipiden (PL). PL zijn polairder, dus minder afhankelijk van de

aanwezigheid van gal voor solubilizatie in het waterige darmlumen, en ze worden

gedeeltelijk intact geabsorbeerd door de darm, zonder lipolyse(6; 22). PL kunnen

hierdoor makkelijker uit de darm worden opgenomen tijdens beperkte lipolyse of

galsecretie, en er zijn aanwijzingen dat PL na absorptie een bijzonder hoge

biologische beschikbaarheid hebben voor het lichaam(8; 10; 11).

Ons doel was de effecten van EFA deficiëntie op darm- en leverfuncties te

specificeren, en een effectief oraal supplement te ontwikkelen om EFA deficiëntie te

voorkomen of te genezen bij patiënten met risico op EFA deficiëntie, zoals bij cystic

fibrosis en cholestase.

Vetmalabsorptie in EFA-deficiënte muizen wordt niet veroorzaaktdoor verminderde galsecretieGalzouten en galfosfolipiden maken absorptie van voedingsvet mogelijk door intra-

luminale vet solubilizatie en door het leveren van oppervlaktemateriaal voor chylo-

micronen (CM). In ratmodellen voor EFA deficiëntie bleek lipolyse van voedingsvet

ongestoord te zijn, maar was galsecretie verminderd(17; 20). In hoofdstuk 2 hebben wij

echter aangetoond dat in EFA deficiënte muizen de galsecretie juist sterk is

toegenomen, waardoor een tekort aan gal als oorzaak van vetmalabsorptie bij EFA

deficiëntie kan worden uitgesloten. Om de rol van galfosfolipiden voor vetopname

nader te specificeren, hebben we de effecten van EFA deficiëntie vergeleken in

156



normale muizen en in genetisch gemodificeerde muizen die fosfolipid-vrije gal

uitscheiden (Mdr2 -/- muizen). Als veranderingen in vetzuursamenstelling van gal-

fosfolipiden betrokken zouden zijn bij EFA deficiëntie-geïnduceerde vetmalabsorptie,

dan zou vetabsorptie in Mdr2 -/- muizen niet veranderen tijdens EFA deficiëntie.

Echter, vetopname en galsecretie waren net zo aangedaan door EFA deficiëntie in

Mdr2 -/- als in controlemuizen. Dit impliceert dat vetmalabsorptie bij EFA deficiëntie

niet veroorzaakt wordt door verminderde galsecretie, maar waarschijnlijk gerelateerd

is aan stoornissen in de intracellulaire fase van vetabsorptie in enterocyten, mogelijk

door te lage EFA concentraties in plasma- en microsomale membranen.

EFA deficiëntie in muizen leidt tot steatose en secretie van grote VLDL deeltjesNaast vetmalabsorptie en hypersecretie van gal is EFA deficiëntie in muizen

geassocieerd met hypotriglyceridemie en steatose. In hoofdstuk 3 onderzochten we

of gestoorde VLDL secretie bijdraagt aan deze metabole consequenties van EFA

deficiëntie in muizen. Zowel in vivo als in vitro bleken lever triglyceride (TG)

concentraties verhoogd, maar was de hepatische VLDL-TG secretie kwantitatief

normaal, hoewel de uitgescheiden VLDL deeltjes opmerkelijk groot waren.

EFA onderdrukken lipogenese in de lever via down-regulatie van SREBP1c, dus kan

synthese van niet-essentiële vetzuren toenemen tijdens EFA deficiëntie. De stapeling

van lever-TG door deze toegenomen synthese, bij gelijkblijvende VLDL-TG secretie,

zou versterkt kunnen worden door toegenomen VLDL klaring. Expressie van genen

die VLDL katabolisme stimuleren (apoA-V, apoC-II) was inderdaad verhoogd, en

hoewel plasma- en leverlipase concentraties (HL en LPL) normaal waren in EFA-

deficiënte muizen, kunnen grotere VLDL deeltjes gevoeliger zijn voor snelle klaring.

Net zoals EFA het functioneren van celmembraan-PL beïnvloeden, kunnen ze de PL

laag die lipoproteïnen omhullen beïnvloeden. Zowel deeltjesgrootte (oppervlak-

tekromming) als lage EFA concentraties kunnen de VLDL structuur veranderen,

resulterend in betere toegankelijkheid voor lipasen of affiniteit voor apoCII of apoAV.

We concluderen dat steatose en hypotriglyceridemie in EFA-deficiënte muizen

waarschijnlijk een resultante is van toegenomen hepatische TG synthese, onveran-

derde VLDL-TG secretie en productie van grote VLDL deeltjes die sneller geklaard

worden. De effecten van EFA deficiëntie op VLDL grootte worden wellicht bepaald

door de beschikbaarheid van PL voor lipoproteïnevorming. De toegenomen secretie

van PL in gal door EFA-deficiënte muizen zou de beschikbaarheid van PL in de lever

voor VLDL vorming kunnen verlagen, waardoor grotere lipoproteïnen geproduceerd

worden.

157



Chylomicron grootte in lymfe is omgekeerd evenredig met gal-PLsecretie in muizenWe hebben de hypothese dat PL-beschikbaarheid de grootte van lipoproteïnen

bepaalt getest in hoofdstuk 4, door te onderzoeken of de grootte van intestinale

lipoproteïnen ook omgekeerd evenredig is met PL beschikbaarheid. Omdat gal-PL

secretie afwezig is in Mdr2 -/- muizen, maar juist sterk verhoogd in EFA-deficiënte

muizen, hebben we chylomicron (CM) grootte in lymfe gemeten in deze twee muis-

modellen, d.m.v. canulatie van de mesenterische lymfevaten. CM waren aanzienlijk

groter in muizen zonder gal-PL secretie (Mdr2 -/- muizen), terwijl hypersecretie van

gal-PL in de darm (EFAD muizen) secretie van kleine CM in lymfe induceerde. Deze

bevindingen bevestigen onze hypothese dat PL beschikbaarheid een cruciale deter-

minant is van intestinale lipoproteïne grootte. Aangezien EFA deficiënte muizen

grotere hepatische VLDL deeltjes produceren dan EFA sufficiënte muizen, is secretie

van kleine lipoproteïnedeeltjes echter geen intrinsiek maar een orgaanspecifiek

effect van EFA deficiëntie. Net als bij VLDL deeltjes kan veranderde grootte en vet-

zuursamenstelling van CM het intravasculaire metabolisme veranderen(13). Grotere

CM worden misschien trager in de lymfe uitgescheiden, maar worden vervolgens

wellicht zeer efficient gemetaboliseerd, aangezien grote CM een grotere affiniteit

hebben voor lipasen en sneller geklaard worden dan kleine CM. Ook worden EFA-

rijke CM sneller geklaard dan EFA-arme. Dit zou voor een deel het fenomeen kunnen

verklaren dat, hoewel post-prandiale verschijning in plasma van voedingsvet

vertraagd is in zowel EFA-deficiënte als in gal-PL-deficiënte (Mdr2 -/-) muizen, alleen

EFA-deficiënte muizen werkelijk vetmalabsorptie hebben. Dit ondersteunt het

concept dat PL in het darmlumen niet zozeer cruciaal zijn voor quantitatieve opname

van vet uit voeding, maar des te meer voor aborptiesnelheid en het daaropvolgende

metabolisme van voedingsvet.

Geen stoornissen in EFA metabolisme in 2 muismodellen voor CFBijvelds(5) beschreef dat in muismodellen voor cystic fibrosis (CF), cftr -/- of CFTR

knockout-muizen een aanzienlijke vetmalabsorptie hebben, wat kan resulteren in

EFA deficiëntie. Tekorten aan EFA of hun lange-keten metabolieten (LCPUFA)

worden inderdaad nog steeds beschreven in CF patiënten, ondanks de huidige

hypercalorische voeding en pancreasenzym substitutie. Er is gesuggereerd dat

verminderde functie van het CFTR eiwit direct de synthese van LCPUFA verstoort,

waarbij pro-inflammatoire n-6 concentraties verhoogd, en anti-inflammatoire n-3

LCPUFA concentraties verlaagd zouden zijn in de typsich in CF aangedane organen,

wat een primaire bijdrage zou leveren aan de symptomen van CF. Suppletie met n-3

LCPUFA, maar niet met de voorloper ALA, zou bepaalde symptomen van CF

158



verbeteren. Om vast te stellen of specifieke EFA- of LCPUFA-suppletie geïndiceerd is

in CF, hebben we vetzuurprofielen geanalyseerd in organen van 2 muismodellen

voor CF; met en zonder vetmalabsorptie. In beide CF modellen hebben we aange-

toond dat vetzuurprofielen in organen van CF muizen vrijwel identiek waren aan die

in niet-aangedane dieren van dezelfde sexe uit hetzelfde nest. In vivo conversie van13C-EFA in AA en DHA was ongestoord in CF muizen vergeleken met controles uit

hetzelfde nest, wat impliceert dat gestoorde LCPUFA synthese geen intrinsiek

kenmerk is van het CF fenotype, en dat vetmalabsorptie geen grote invloed heeft op

EFA status in CF muizen.

In de afgelopen jaren is een grote verscheidenheid aan CF muismodellen

ontwikkeld, die net als CF patiënten onderling fenotypisch enorm van elkaar

verschillen. Behalve aan de specifieke CFTR mutatie en aan omgevingsfactoren, is

de fenotypische variabiliteit in CF gerelateerd aan onafhankelijk overervende zoge-

naamde ‘ziekte-modulerende’ genen, die coderen voor eiwitten die (gedeeltelijk)

kunnen compenseren voor de effecten van CFTR dysfunctie. We hebben aange-

toond dat voeding en leeftijd, maar voornamelijk genetische achtergrond en

allesoverheersende determinant zijn van EFA status in CF muizen. In de studies

waarin gesuggereerd werd dat LCPUFA synthese door CFTR gereguleerd wordt, en

zo CF symptomen beïnvloedt, zijn de CF muizen vergeleken met controle muizen

van een geheel andere muizenstam. Voor elke zinnige vergelijking van EFA status

tussen CF muismodellen, en zeker voor extrapolatie van bevindingen naar de

humane situatie, moet de vergelijkbaarheid van de genetische achtergrond van de

muismodellen grondig gecontroleerd worden, en is het gebruik van controlemuizen

uit hetzelfde nest een absolute vereiste.

We concluderen dat CFTR dysfunctie LCPUFA synthese niet verstoort, en dat sub-

optimale vetzuurprofielen in CF waarschijnlijk een secundair gevolg zijn van chro-

nische ontsteking of inadequate voeding. Voor CF patiënten zouden deze conclusies

impliceren dat een adequate EFA intake afdoende het ontstaan van EFA of LCPUFA

deficiëntie zou moeten kunnen voorkomen. Aangezien vetmalabsorptie geen

duidelijk effect had op EFA status in CF muizen, en specifieke EFA suppletie dus niet

geïndiceerd lijkt bij deze aandoening, hebben wij ons voor verdere studies naar

geschikte EFA suppletiestrategieen geconcentreerd op patiënten met cholestatische

leveraandoeningen.

Orale behandeling van EFA deficiëntie met TG of PL bij cholestaseEFA deficiëntie komt veel voor bij cholestatische aandoeningen, als gevolg van mal-

absorptie van EFA uit de voeding, waarin EFA vooral aanwezig zijn geësterificeerd tot

hydrofobe TG moleculen. De amfifiele PL moleculen zijn beter oplosbaar dan TG in

159



het waterige darmlumen, en worden gemakkelijker geabsorbeerd tijdens gal

deficiëntie. PL worden ofwel als intacte moleculen geabsorbeerd, of na vertering tot

lyso-PL. In hoofdstuk 5 hebben we aangetoond dat in EFA-deficiënte muizen met

acute cholestase, orale EFA suppletie als TG of als PL even effectief kon voorkomen

dat EFA concentraties in erythrocyten daalden. Echter, voor hersenen leverweefsel

waren orale PL duidelijk superieur aan TG voor het verhogen van EFA en LCPUFA

concentraties. Daarnaast verhinderde orale suppletie met PL gewichtsverlies tijdens

EFA deficiëntie en cholestase, en suppletie met TG niet, wat het concept onder-

steunt dat enterale PL een positief effect hebben op vetabsorptie tijdens tekort aan

gal in de darm.

In hoofdstuk 6 hebben we de effectiviteit van orale EFA suppletie als PL of TG

vergeleken voor de behandeling van EFA deficiëntie bij kinderen met chronische

cholestase, die frequent een suboptimale EFA status hebben(16). EFA-suppletie als TG

of PL gedurende drie maanden voorkwam verslechtering van linolzuur (C18:2n-6),

mead acid (C20:3n-9) en de som van n-6 vetzuren in rode bloedcellen (RBC) zoals

die te zien was bij niet-gesuppleerde kinderen. En hoewel we geen directe

verschillen konden aantonen van TG- of PL-suppletie bij groepsgewijze vergelijking

van EFA in RBC, resulteerde PL-suppletie duidelijk in een positieve maandelijkse toe-

name van linolzuur in RBC, en TG-suppletie niet.

Opmerkelijk genoeg was in cholestatische muizen het verschil in effect tussen TG en

PL suppletie niet meetbaar in RBC, maar wel heel duidelijk in de doelwitorganen van

EFA, lever en hersenen.

Aansluitend bij Korotkova en Rioux trekken wij het paradigma in twijfel dat RBC vet-

zuurprofielen een accurate weerspiegeling zijn van EFA status in het lichaam. Door

hun lange halfwaardetijd, en hun korte-termijn onafhankelijkheid van post-prandiale

voedingsvetconcentraties in plasma, zijn RBC inderdaad een stabiel en toegankelijk

compartiment voor analyse van EFA status. Maar vermoedelijk vindt er specifieke

kanalisatie plaats van EFA of LCPUFA naar doelwitorganen als brein, lever en darm,

ten koste van minder cruciale weefsels als RBC. Behalve internationale consensus

t.a.v. gevalideerde referentiewaarden die biochemische EFA-deficiëntie definieren,

zou het zeer nuttig zijn om sensitieve functionele tests te ontwikkelen ter beoordeling

van deficiëntie, sufficiëntie en inname-behoefte van EFA en LCPUFA in patiënten en

in diermodellen(7; 21).

Het mechanisme dat de opvallend efficiënte opname van EFA of LCPUFA door de

hersenen reguleert is nog onbekend. In beperkte mate kunnen vetvoorraden in het

lichaam de hersenen van EFA voorzien(9), en hoewel astrocyten AA en DHA kunnen

160



synthetiseren uit EFA, is plasma de belangrijkste bron van LCPUFA voor het brein. In

plasma zijn LCPUFA aanwezig als acyl-esters in lipoproteïnen, als vrije vetzuren

gebonden aan albumine en als lyso-PL. Die laatste vorm wordt preferentieel in de

hersenen geïncorporeerd(4; 18; 19). PL uit voeding worden uit de darm opgenomen als

intacte moleculen en als lyso-PL. Daarnaast kunnen orale PL een bron zijn van plas-

ma (lyso)-PL als componenten van HDL-deeltjes, die ontstaan uit overtollig opper-

vlaktemateriaal van chylomicronen (CM). Wellicht worden CM oppervlakte-PL prefe-

rentieel gedirigeerd naar HDL voor transport naar het brein, voordat uitwisseling met

andere lipoproteïnen of RBC plaatsvindt(6; 15; 22). Zowel in de vorm van albumine-

gebonden lyso-PL als in de vorm van HDL-PL, vormen orale PL een goed toeganke-

lijke bron van LCPUFA voor de hersenen, wat het concept ondersteunt dat enteraal

opgenomen PL een grote biologische beschikbaarheid hebben voor het lichaam.

Uiteraard is hersenen leverweefsel in humane studies ethisch gezien niet beschik-

baar voor analyse van vetzuurprofielen. Maar de gunstige effecten van PL suppletie

op de hersenen en de lever in cholestatische muizen, en de iets betere effecten van

PL suppletie op RBC EFA in cholestatische kinderen, vormen wellicht toch een

argument om EFA suppletie in de vorm van PL te prefereren boven TG voor

preventie of behandeling van EFA deficiëntie in kinderen met cholestase.

CONCLUSIES

Voorafgaand aan de in dit proefschrift beschreven studies, was uit de literatuur

bekend dat EFA deficiëntie belangrijke consequenties heeft voor vetabsorptie en

-metabolisme.

Aangetoond was dat EFA deficiëntie:

- galsecretie in ratten verlaagt;

- intraluminale lipolyse en enterocyt opname van voedingsvet niet beïnvloedt;

- linolzuur en arachidonzuur concentraties in intestinale membranen verlaagt;

- stapeling van vet in enterocyten veroorzaakt na orale toediening hiervan;

- vetmetabolisme na absorptie (bv. plasma lipid concentraties, lipoproteïne

samenstelling en activiteit van lipolytische enzymen) beïnvloedt.

Echter, de specifieke rol van EFA deficiëntie in de verschillende fasen van vetab-

sorptie was niet goed begrepen. In dit proefschrift is aangetoond dat vetmalabsorp-

tie tijdens EFA deficiëntie niet veroorzaakt wordt door verminderde galsecretie of

door verlaagde EFA concentraties in galfosfolipiden. Op basis van de geobserveerde

effecten van EFA deficiëntie op lipoproteïnevorming vermoeden wij dat EFA

161



deficiëntie-geïnduceerde vetmalabsorptie het gevolg is van EFA tekorten in

membraanfosfolipiden, waardoor lipoproteïnemetabolisme gestoord is in lever en

darm. In beide organen is tijdens EFA deficiëntie de secretiestap van het lipide-

transport aangedaan; in beide organen is lipoproteïne grootte veranderd en treedt

vetstapeling op. PL synthese en incorporatie in membranen gaat onverminderd door

bij een tekort aan EFA, waarbij essentiële vetzuurketens vervangen worden door

beschikbare n-9, n-7 of verzadigde vetzuren. De mechanismen die bepalen welk type

vetzuur wordt geïncorporeerd in PL, of welk type PL wordt ingebouwd in

lipoproteïne- of celmembranen zijn nog niet duidelijk.

Er zijn sterke aanwijzingen dat de chemische structuur van voedingsvet, TG of PL,

relatief goed bewaard blijft bij intestinale vertering en absorptie. Deze aanwijzingen,

samen met onze resultaten t.a.v. specifieke kanalisatie van EFA-PL naar lever en

brein, suggereren dat aparte intracellulaire vetzuur-pools bestaan. Het zou interes-

sant zijn om de mechanismen te achterhalen die het sorteren en dirigeren van

verschillende vetzuur- en PL-types naar verschillende compartimenten in enterocyten

of hepatocyten reguleren, en de daaropvolgende kanalisatie naar specifieke doel-

organen. Zowel endotheel-lipase activiteit van de bloed-hersen-barriere als SRB1-

gemedieerde opname van HDL-PL EFA door capillaire endotheelcellen zou een rol

kunnen spelen bij de specifieke opname van EFA door het brein. In cholestatische

muizen hebben de preferentiële opname van voedings-EFA in de vorm van PL door

het brein duidelijk kunnen aantonen, waardoor deze vorm van EFA suppletie wellicht

ook aan te bevelen is bij kinderen met chronische leverziekten. Voor toekomstige

studies zou het fysiologisch en therapeutisch interessant zijn om de mechanismen

die ten grondslag liggen aan de specifieke opname van EFA en LCPUFA door de

hersenen, en de kanalisatie van enterale fosfolipid-EFA naar lever en brein op te

helderen.

162



DANKWOORD

Dit proefschrift begint bij Ellen van der Gaag. Op een feestje vlak voor mijn arts-

examen vertelde ik haar in alle academische kinderklinieken te gaan solliciteren

behalve in Groningen en Maastricht, want dat vond ik te ver weg. Maar volgens Ellen

moest iemand die zonodig onderzoek wilde doen toch ook echt naar Groningen

schrijven, want daar gebeurden leuke dingen in het researchlab Kinder-

geneeskunde. Aldus geschiedde, en zo begon mijn reis door de wondere wereld van

wetenschap met een emigratie naar het hoge Noorden.

Ik bedank alle analisten die met veel (en soms wat minder) geduld hebben gezorgd

dat ik als onhandige dokter niet verdwaalde in de biochemische jungle van labland,

want verdwalen doe je eerst in je hoofd en dan pas in het bos. Ingrid, vet-

zuurkoningin die met wijze raad van al mijn orgaansapjes en -papjes chro-

matogrammen wist te brouwen; zonder jou wisten we nog steeds niet of de OLT-

kinderen hun vetzuurpudding opaten of in de plantenbak kieperden. Heel veel dank.

Marchien en Herman, carpe diem. Globetrotters Juul en Monsieur Renze, Anke

(jij die alles weet), Janneke, Nicolette, Janny, Mijnheer Marius, Henk W. (don't like no

short people) en Henk E. (doen we die 3042 monsters toch gewoon opnieuw?),

Frank (lymfe zuigt niet, lymfe plakt), Vincent en Fjodor; allen bedankt voor jullie

labologische adviezen. Trijnie, onze noeste LPL-assay-pogingen (zachtjes ultra-

sonificeren, straks moeten we de 14C van het plafond krabben) verdiende het

reviewerscommentaar "the authors have plenty of experience with the applied

methods" driedubbeldik. Herr Rick Deckelclick, grijnzende spin in het web der lab-

intriges, de eindeloze lymfecannulaties in ADL8 waren een beproeving maar wel erg

gezellig, al was mij nooit helemaal duidelijk wie nu eigenlijk aan wie de sappigste

roddels ontfutselde. Waar blijft die fles Martini?

De Baukje-hotline werd geopend toen na mijn eerste hepatocytisolatie alle celkweek-

expertise geëmigreerd bleek naar Ierland of op fietsvakantie in Vietnam. Onze kook-

boekmethode ("OK, links een bak cellen en rechts blauw spul, wat nu?") werkte

perfect en ontlokte opnieuw een reviewerscompliment dat we zo ervaren waren.

Dank voor je hulp en voor de gezellige XX-versterking in de club der vetklepjes. Ik

hoop dat je geniet van je tijd in Ierland en rustig aan doet met stout en bitter, cave de

klop met de hamer. Ook de andere ‘first generation’ AIO's, Tineke, Marieke,

Jacqueline, Tomoji, Jenny, Coen, Arjen, Aldo, Janine, Torsten, Lorraine, Lisethe,

Christian, Peter, Guido; bedankt voor jullie gezelligheid en de vaak zeer welkome tips

over Photoshop, SPSS en andere ongemakken. De afgelopen maanden heb ik met

163

Dankwoord


veel plezier de nieuwe ‘patatgeneratie’ AIO's leren kennen, piepjong maar des te

vrolijker: Esther, Maaike (“keileuk”), Juffrouw Jelske ("kunnen Dirkjan en professorVonk ons door de muur heen horen roddelen?"), Martijn, Mirjam, Titia en Jaap.

Hans, zoals beloofd hier geen ADHD-grappen; niets dan lof voor de koning der

opperkippen. Leonie (the woman most likely to go Los), als two white chickies inDa House Of Blues was het heerlijk dansen met onze voluptueuze brothers and

sisters, Chicago Rules! Anja (the woman most likely to föhn anything), ik vond het

erg gezellig met jou en de bikini-Martini’s in Chicago en met de zeehondjes in Seattleen op de Waddenzee. Er schijnen Chinese stoelen verkocht te worden bij Ikea,

wanneer gaan we die opblazen?

Ook wil ik heel hartelijk alle BKK arts-assistenten bedanken die met hun gezelligheiden geweldige collegialiteit hebben gezorgd dat een groen-als-gras-onderzoekertje

boven water bleef in de eerste onervaren kliniektijd, en maakten dat ik puf en af en

toe een avonduurtje overhield om aan dit proefschrift te werken. Andere Anniek("nee, jíj bent Andere Anniek! En er moeten snel meer koekjes komen!"), Hella, Baas

der Annieken en Heldin van M2 ("ga lunchen, nu!"), Marlon ("wie is de halve paar-

denkop?"), de Elizabethen ("wie is toch Bart Rottier?"), Gerbrich ("prikken moet jedóen!"), Tuesdays with Hannah ("het was wèl een zeehondje, en hij lachte naar

ons!"), Slapen met Stijntje ("snurk jij?"), Margreet ("Berenburgje?), Christian ("I am

friendly, and you are friendly too"), Ellis, Louise, Aline, Nynke, Ginny, Lisethe ("hoeoverleef je M4?") en de Zwollywood-diva's Meinke, Anouk en Marije ("ik kan altijd nog

fietsenmaker worden"): allemaal heel veel dank.

Renate en Robert, paranimf en parafaun, lief dat jullie mij bijstaan in dit uur U, ennatuurlijk dank voor alle grappen, grollen, biertjes, wijntjes, fotoklasjes, zeilklasjes,

mailmarathons en alles wat maakt dat collega's vriendjes worden.

Dat Folkerts goed volk waren wist ik al, maar dat er Folkert-achtige professoren

rondlopen, is een geruststellende gedachte. Ik wil jou graag hartelijk bedanken voor

je begeleiding, je snelle respons op mijn manuscripten waar steeds een vertrouwdesigaarlucht vanaf kwam in combinatie met geniale commentaren, en voor je

bereidheid onmiddellijk weer een plekje in het lab voor mij vrij te maken toen dat

nodig bleek. Ook "mijn" andere professor, professor Sauer, wil ik graag enormbedanken voor de steun en voor de hartelijke manier waarop ik werd opgevangen

toen mijn slechte nieuws boven tafel kwam. Henkjan, zeergeëerde promotor, als een

havik hield jij van grote hoogte haarscherp mijn wandel door de wereld van hard-corescience in de gaten. Jouw begaafdheid als wetenschapper behoeft geen toelichting,

maar die als begeleider vind ik nog bijzonderder. Iemand zei dat een goede baas

geen opdrachtgever is, maar een vragensteller, een motivator of een inspirator

164

Dankwoord


(Brent et al., the Office). Jij bent alledrie, en ik vond het een voorrecht om begeleid

te worden door iemand die zoveel enthousiasme en fascinatie voor wetenschap uit-straalt. Jouw inspanningen en medeleven de afgelopen maanden onderstreepten

wat ik al wist: 't kon veel slechter qua baas (voor niet-Noordelingen: ultiem compli-

ment).Ik bedank nogmaals heel hartelijk iedereen van de BKK (arts-assistenten, staf, ver-

pleegkundigen) en van het lab Kindergeneeskunde en MDL voor alle lieve reacties.

Het heeft me veel goed gedaan om te merken dat er in Groningen een warm nestbleek te zijn voor mij.

En over nesten gesproken; pappie, mammie, als ik jullie begin te bedanken houd ikniet meer op. Jullie wisten ons op te voeden zonder dat we het in de gaten hadden.

Als het vermogen van een individu tot empathie afhangt van de hoeveelheid liefde

die het als kind heeft gehad, en van hoe het als kind is gewaardeerd en ge-respecteerd, dan zijn Frank en ik eindeloos empathisch. Frank, grote Werner Bro,

duizend keer dank dat je altijd alles beter weet dan ik (?), en dat je met zoveel andere

dingen en boeven aan je hoofd toch dit boekje hebt vormgegeven. Wel cool dat iknu eindelijk iets heb gedaan dat jij niet allang kon. Suzanne, Quinten, dank voor het

mogen lenen van Frank, en natuurlijk voor de kwak-kwakgrappen en nog veel meer.

Folkert, grote liefde, dit boekje begint met jou, en na veel citaten vanzeergeleerde types sluit ik graag af met jouw motto: neus in de wind en

moedigh voorwaertsch!

There is more to explore.

165

Dankwoord


CURRICULUM VITAE

The author of this thesis was born on February 10 th 1973 in Nijmegen, theNetherlands. After obtaining her Athenaeum-ß diploma in Nijmegen, she studied

medicine at the University of Leuven in Belgium for one year, and continued her

medical studies at the University of Leiden. To get a closer look at scientific research,she worked as a student-assistant at the Leiden University laboratory for experimen-

tal cardiophysiology (head: prof. dr. J. Baan) from 1993 to 1996. Her interest in

research was confirmed during her graduation project at the neonatal intensive careunit in Edinburgh, where she investigated the possibilities of artificial neural networks

as an aid for early diagnosis of neonatal sepsis, supervised by prof. dr. N. McIntosh

and prof. dr. A.F. Murray.After obtaining her M.D. degree in 1999, she started this PhD project on essential

fatty acid absorption and metabolism described in this thesis, supervised by prof. dr.

H.J. Verkade, prof. dr. F. Kuipers and prof. dr. P.J.J. Sauer. In the summer of 2002 sheworked as a resident in the children's nutritional rehabilitation ward of the Mporokoso

District Hospital in Zambia. In October 2003, she started her residency in Pediatrics

in the Beatrix Children's Hospital (head: prof. dr. P.J.J. Sauer), University MedicalCenter Groningen, the Netherlands.

166

Curriculum Vitae


PUBLICATIES

Verkade H.J., Bijleveld C.M.A., Werner A.Intestinale vetmalabsorptie: pathofysiologie en diagnostiek.

Tijdschrift Kindergeneeskunde 2000; 68: 175-182

Werner A., Minich D.M., Havinga H., Van Goor H., Kuipers F., Verkade H.J.

Fat malabsorption in EFA deficient mice is not due to impaired bile formation.

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

Werner A., Kuipers F., Verkade H.J.

Fat Absorption and Lipid Metabolism in Cholestasis.Book chapter in: Molecular Pathogenesis of Cholestasis. 2003; p321-335.

Trauner M., Jansen P.L.M. (Eds.), Landes Bioscience

Werner A., Havinga H., Bos T., Bloks V.W., Kuipers F., Verkade H.J.

Essential fatty acid deficiency in mice is associated with hepatic steatosis and

secretion of large VLDL particles.Am J Physiol Gastrointest Liver Physiol 2005; 288(6): G1150-1158

Werner A., Havinga H., Perton F., Kuipers F., Verkade H.J. Lymphatic chylomicron size is inversely related to biliary phospholipid secretion in

mice.

Am J Physiol 2005 (In press)

Werner A., Bongers M., Bijvelds M.J., de Jonge H.R., Verkade H.J.

No indications for altered EFA metabolism in two murine models for cystic fibrosis.J Lipid Res. 2004; 45(12): 2277-2286

Werner A., Havinga H., Kuipers F., Verkade H.J.Treatment of essential fatty acid deficiency with dietary triglycerides or phospholipids

in a murine model of extrahepatic cholestasis.

Am J Physiol Gastrointest Liver Physiol 2004; 286(5): G822-832

Werner A., Bijleveld C.M.A., Martini I.A., van Rijn M., van der Heiden J., Sauer P.J.J.,

Verkade H.J.Oral treatment of essential fatty acid deficiency with triglycerides or phospholipids

in children with end-stage liver disease.

Submitted

167

Publications


Essential fatty acid absorption and metabolism in …ESSENTIAL FATTY ACIDS Linoleic acid (LA,...

Documents

Transcript of Essential fatty acid absorption and metabolism in …ESSENTIAL FATTY ACIDS Linoleic acid (LA,...