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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 2
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 3
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 4
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 5
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 6
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 7
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)
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 8
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:23 Pagina 9
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 10
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
Unsaturated fatty acids
EFA non-EFA
n-9 familyn-3 family n-6 family n-7 family
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
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 11
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
General introduction
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 14
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
General introduction
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 15
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
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
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
Figure 5: Symptoms of EFA deficiency
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 16
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
General introduction
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 17
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 18
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
General introduction
proefschrift_def_v0870605.qxp 7-6-2005 22:43 Pagina 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
proefschrift_def_v0870605.qxp 7-6-2005 22:43 Pagina 20
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
General introduction
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 21
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
proefschrift_def_v0870605.qxp 7-6-2005 22:43 Pagina 22
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
General introduction
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 24
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
General introduction
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 25
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 26
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
General introduction
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 28
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
General introduction
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 30
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
General introduction
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 32
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
General introduction
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 34
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
General introduction
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 36
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
proefschrift_def_v010605def.qxp 2-6-2005 1:26 Pagina 37
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 38
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 39
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
proefschrift_def_v010605def.qxp 2-6-2005 1:19 Pagina 40
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 41
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 42
43
Fat malabsorption inessential fatty acid-deficient mice is not due toimpaired bile formation
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 43
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
proefschrift_def_v0870605.qxp 7-6-2005 22:44 Pagina 44
45
Fat malabsorption inessential fatty acid-deficient mice is not due toimpaired bile formation
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
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
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 45
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
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
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 46
47
Fat malabsorption inessential fatty acid-deficient mice is not due toimpaired bile formation
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 47
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.
proefschrift_def_v0870605.qxp 7-6-2005 22:44 Pagina 48
49
Fat malabsorption inessential fatty acid-deficient mice is not due toimpaired bile formation
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.
41.7 ± 7.6 *11.6 ± 1.5Cholesterol esters
81.4 ± 11.8 *45.8 ± 2.1Total cholesterol
225.8 ± 95.9 *101.0 ± 22.2Triglyceride
EFADEFAS
41.7 ± 7.6 *11.6 ± 1.5Cholesterol esters
81.4 ± 11.8 *45.8 ± 2.1Total cholesterol
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 49
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
0.7 ± 0.30.9 ± 0.4Chenodeoxycholate
0.8 ± 0.41.6 ± 0.9Deoxycholate
1.1 ± 0.31.0 ± 0.6Alpha-muricholate
5.2 ± 1.14.8 ± 1.7Omega-muricholate
35.9 ± 4.128.0 ± 8.0Beta-muricholate
56.2 ± 3.463.7 ± 7.3Cholate
EFADEFAS
0.7 ± 0.30.9 ± 0.4Chenodeoxycholate
0.8 ± 0.41.6 ± 0.9Deoxycholate
1.1 ± 0.31.0 ± 0.6Alpha-muricholate
5.2 ± 1.14.8 ± 1.7Omega-muricholate
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 50
51
Fat malabsorption inessential fatty acid-deficient mice is not due toimpaired bile formation
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 51
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 52
53
Fat malabsorption inessential fatty acid-deficient mice is not due toimpaired bile formation
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 53
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 54
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
proefschrift_def_v010605def.qxp 2-6-2005 1:27 Pagina 55
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 56
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 57
58
Chapter 3
MATERIALS AND METHODS
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 58
59
EFA deficiency in mice is associated with hepatic steatosis and secretion of large VLDL particles
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 59
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 60
61
EFA deficiency in mice is associated with hepatic steatosis and secretion of large VLDL particles
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 61
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 62
63
EFA deficiency in mice is associated with hepatic steatosis and secretion of large VLDL particles
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 63
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 64
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
EFA deficiency in mice is associated with hepatic steatosis and secretion of large VLDL particles
EFADEFAS
VLDL lipids 4 h after Triton
0
10
20
30
chol TG PL
mM
EFADEFASEFADEFADEFASEFAS
VLDL lipids 4 h after Triton
0
10
20
30
chol TG PL
mM
VLDL lipids 4 h after Triton
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
VLDL lipids 4h after Triton
0
20
40
60
80
100
%chol %TG %PL
%of
tota
l lip
id
VLDL lipids 4h after Triton
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
VLDL apoB 4h after Triton
band
inte
nsity
/ m
l VLD
L
0
20
40
60
apoB100 apoB48
*
VLDL apoB 4h after Triton
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 65
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
**
**
* *
fatty acid composition hepatocyte TG
fatty acid composition hepatocyte PL
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
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 66
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
EFA deficiency in mice is associated with hepatic steatosis and secretion of large VLDL particles
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 67
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 68
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
EFA deficiency in mice is associated with hepatic steatosis and secretion of large VLDL particles
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 70
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
EFA deficiency in mice is associated with hepatic steatosis and secretion of large VLDL particles
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 72
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
EFA deficiency in mice is associated with hepatic steatosis and secretion of large VLDL particles
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 73
74
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:27 Pagina 75
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 76
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 77
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(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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 78
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
Lymphatic chylomicron size is inversely related to biliary phospholipid secretion in mice
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 80
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
Lymphatic chylomicron size is inversely related to biliary phospholipid secretion in mice
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-/-
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
###
*
*##
##Mdr2+/+
Mdr2-/-
lipid composition lymph 0-60 min
0
2
4
6
8
10
12
0
2
4
6
8
10
12
PL TG chol
relative lipid composition lymph 0-60 min
0
20
40
60
80
100
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-/-
## #
#
lipid composition lymph >60 min
0
2
4
6
8
10
12
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
relative lipid composition lymph >60 min
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
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
** **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
% 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
relative lipid composition lymph VLDL 0-60min
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
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-/-
##
##
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 81
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-/-
core-surface ratio (0-60 min)
[TG
] / [P
L]
0
2
4
6
8
10
total lymph CM VLDL
**
#
core-surface ratio (0-60 min)
[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
**
#
core-surface ratio (>60 min)
0
2
4
6
8
10
total lymph CM VLDL
[TG
] / [P
L]
#
core-surface ratio (>60 min)
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 82
83
Lymphatic chylomicron size is inversely related to biliary phospholipid secretion in mice
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-/-
##
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
0
5
10
15
PL TG chol
mM
Mdr2+/+
Mdr2-/-Mdr2+/+Mdr2+/+
Mdr2-/-Mdr2-/-
##
lipid composition lymph 1 hour after lipid bolus
0
5
10
15
0
5
10
15
PL TG chol
mM
Mdr2+/+
Mdr2-/-Mdr2+/+Mdr2+/+
Mdr2-/-Mdr2-/-
*
#
#
lipid composition lymph 2 hours after lipid bolus
0
5
10
15
0
5
10
15
PL TG chol
mM
Mdr2+/+
Mdr2-/-Mdr2+/+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-/-
core-surface ratio after lipid bolus
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
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
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-/-Mdr2+/+Mdr2+/+
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
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-/-Mdr2+/+Mdr2+/+
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
0
20
40
60
80
100
relative lipid composition lymph baseline
PL TG chol
% o
f tot
al li
pid
Mdr2+/+
Mdr2-/-Mdr2+/+Mdr2+/+
Mdr2-/-Mdr2-/-
#
#
*
4a
4b
4c
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 83
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 84
85
Lymphatic chylomicron size is inversely related to biliary phospholipid secretion in 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
EFASEFAD
#
#
0
20
40
60
80
100
relative lipid composition lymph 1 hour after lipid bolus
PL TG chol
% o
f tot
al li
pid
EFASEFAD
0
20
40
60
80
100
0
20
40
60
80
100
relative lipid composition lymph 1 hour after lipid bolus
PL TG chol
% o
f tot
al li
pid
EFASEFADEFASEFASEFADEFAD
#
#
*
0
20
40
60
80
100
relative lipid composition lymph 2 hours after lipid bolus
PL TG chol
% o
f tot
al li
pid
EFASEFAD
*
*
0
20
40
60
80
100
relative lipid composition lymph 2 hours after lipid bolus
PL TG chol
% o
f tot
al li
pid
EFASEFAD
0
20
40
60
80
100
0
20
40
60
80
100
relative lipid composition lymph 2 hours after lipid bolus
PL TG chol
% o
f tot
al li
pid
EFASEFADEFASEFASEFADEFAD
*
0
20
40
60
80
100
relative lipid composition lymph baseline
PL TG chol
% o
f tot
al li
pid
EFASEFAD
#
#
0
20
40
60
80
100
relative lipid composition lymph baseline
PL TG chol
% o
f tot
al li
pid
EFASEFAD
0
20
40
60
80
100
0
20
40
60
80
100
relative lipid composition lymph baseline
PL TG chol
% o
f tot
al li
pid
EFASEFADEFASEFASEFADEFAD
#
#
EFAS EFAD
core-surface ratio after lipid bolus
0
2
4
6
8
10
[TG
] / [P
L]
*
EFAS EFAD
core-surface ratio after lipid bolus
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
lipid composition lymph 1 hour after lipid bolus
PL TG chol
# mM
PL TG chol0
5
10
15
20
25 EFASEFAD
lipid composition lymph 2 hours after lipid bolus
mM
PL TG chol0
5
10
15
20
25 EFASEFAD
lipid composition lymph baseline
#
*
mM
0
5
10
15
20
25
0
5
10
15
20
25 EFASEFADEFASEFASEFADEFAD
lipid composition lymph 1 hour after lipid bolus
PL TG cholPL TG chol
# mM
PL TG chol0
5
10
15
20
25
0
5
10
15
20
25 EFASEFADEFASEFASEFADEFAD
lipid composition lymph 2 hours after lipid bolus
mM
PL TG cholTG chol0
5
10
15
20
25
0
5
10
15
20
25 EFASEFADEFASEFASEFADEFAD
lipid composition lymph baseline
#
*
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 85
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 86
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
Lymphatic chylomicron size is inversely related to biliary phospholipid secretion in mice
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 88
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
Lymphatic chylomicron size is inversely related to biliary phospholipid secretion in mice
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 90
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
proefschrift_def_v010605def.qxp 2-6-2005 1:28 Pagina 91
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 92
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 93
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 94
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
No indications for altered EFA metabolism in two murine models for cystic fibrosis
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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
50mx0.2mm Ultra 1 capillary column (Hewlett Packard, Palo Alto, CA) and a FID
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 96
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
No indications for altered EFA metabolism in two murine models for cystic fibrosis
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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 -/-
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 -/-
jejunum PL 13C -LA and 13C-AA
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
pancreas PL13C-LA and 13C-AA
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
pancreas PL13C-LA and 13C-AA
100
x A
TE
0
2
4
6
8
10
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
lung PL13C-LA and 13C-AA
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
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 -/-
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 -/-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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 98
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
No indications for altered EFA metabolism in two murine models for cystic fibrosis
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
pancreas total lipid
pancreastotal lipid
pancreas total 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
LA AA ALA DHALA AA ALA DHA
0
10
20
30
LA AA ALA DHA0
10
20
30
LA AA ALA DHA0
10
20
30
0
10
20
30
LA AA ALA DHALA AA ALA DHA
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
LA AA ALA DHALA AA ALA DHA
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
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 DHALA AA ALA DHA
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 99
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
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
jejunum phospholipid
jejunum phospholipid
jejunum phospholipid
jejunum phospholipid
cftr +/+
cftr -/-N/N?F508/?F508
cftr +/+
cftr -/-cftr +/+
cftr -/-cftr +/+
cftr -/-N/N?F508/?F508N/N?F508/?F508
pancreas phospholipid
pancreasphospholipid
pancreas phospholipid
pancreasphospholipid
0
10
20
30
LA AA ALA DHA0
10
20
30
0
10
20
30
LA AA ALA DHALA AA ALA DHA
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
LA AA ALA DHALA AA ALA DHA
mo
l%
0
10
20
30
LA AA ALA DHA
mo
l%
0
10
20
30
0
10
20
30
LA AA ALA DHALA AA ALA DHA
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
LA AA ALA DHALA AA ALA DHA
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 100
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
No indications for altered EFA metabolism in two murine models for cystic fibrosis
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
0 10 20 30 40bodyweight (g)
mol
% fa
tty a
cid
DHA (22:6n-3) pancreas PL
012345
0 10 20 30 40bodyweight (g)
mol
% fa
tty a
cid
AA (20:4n-6) pancreas PL
0
5
10
15
0 10 20 30 40bodyweight (g)
+/+ -/-
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)
LA (18:2n-6) pancreas PL
0
10
20
30
0
10
20
0
10
20
30
0 10 20 30 400 10 20 30 40bodyweight (g)
ALA (18:3n-3) pancreas PL
0.00.20.40.60.81.0
0 10 20 30 40bodyweight (g)
mol
% fa
tty a
cid
ALA (18:3n-3) pancreas PL
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
DHA (22:6n-3) pancreas PL
012345
0 10 20 30 40bodyweight (g)
mol
% fa
tty a
cid
DHA (22:6n-3) pancreas PL
012345
012345
0 10 20 30 400 10 20 30 40bodyweight (g)
mol
% fa
tty a
cid
AA (20:4n-6) pancreas PL
0
5
10
15
0 10 20 30 40bodyweight (g)
AA (20:4n-6) pancreas PL
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
?F508N/N cftr +/+ cftr -/-
LA pancreas per mg protein
0
50
100
?F508N/N cftr +/+ cftr -/-
ALA pancreas per mg protein
0
1
2
?F508N/N cftr +/+ cftr -/-
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
AA pancreas per mg protein
0
10
20
30
?F508N/N cftr +/+ cftr -/-
AA pancreas per mg protein
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
?F508N/N cftr +/+cftr +/+ cftr -/-cftr -/-
LA pancreas per mg protein
0
50
100
?F508N/N cftr +/+ cftr -/-
LA pancreas per mg protein
0
50
100
0
50
100
?F508N/N cftr +/+cftr +/+ cftr -/-cftr -/-
ALA pancreas per mg protein
0
1
2
?F508N/N cftr +/+ cftr -/-
ALA pancreas per mg protein
0
1
2
?F508N/N cftr +/+ cftr -/-
4c
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 101
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
*
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
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
*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*
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
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 102
103
No indications for altered EFA metabolism in two murine models for cystic fibrosis
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
cftr +/+ Peptamencftr -/- Peptamencftr +/+chowcftr -/- chow
cftr +/+ Peptamencftr -/- Peptamencftr +/+chowcftr -/- chow
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).
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 103
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
pancreasphospholipid
*
***
mo
l%m
ol%
mo
l%m
ol%
LA AA ALA DHA
0
2
4
6
8
10
LA AA ALA DHA
lung phospholipid
# ##
LA AA ALA DHALA AA ALA DHA
0
2
4
6
8
10
0
2
4
6
8
10
LA AA ALA DHALA AA ALA DHA
lung phospholipid
# ##
0
5
10
15
20
25
cftr +/+
cftr -/-C57/Bl/6
pancreasphospholipid
*
***
0
5
10
15
20
25
0
5
10
15
20
25
cftr +/+
cftr -/-C57/Bl/6
cftr +/+
cftr -/-C57/Bl/6
pancreasphospholipid
*
***
6bm
ol%
0
10
20
30
LA AA ALA DHA
jejunum phospholipid
#
##
LA AA ALA DHA02468
1012
mo
l%
plasma phospholipid
*
**
*
mo
l%m
ol%
0
10
20
30
LA AA ALA DHA
jejunum phospholipid
#
##
0
10
20
30
LA AA ALA DHALA AA ALA DHA
jejunum phospholipid
#
##
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 104
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
No indications for altered EFA metabolism in two murine models for cystic fibrosis
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 106
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
No indications for altered EFA metabolism in two murine models for cystic fibrosis
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 108
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
No indications for altered EFA metabolism in two murine models for cystic fibrosis
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 109
110
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:29 Pagina 111
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 112
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 113
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.
MATERIALS AND METHODS
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 114
115
Treatment of EFA deficiency with dietary TG or PL in a murine model of extrahepatic cholestasis
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 115
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
50mx0.2mm Ultra 1 capillary column (Hewlett Packard, Palo Alto, CA) and a FID
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 + EFA-rich PL
EFAD diet
EFAD diet + EFA rich TG
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 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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 116
117
Treatment of EFA deficiency with dietary TG or PL in a murine model of extrahepatic cholestasis
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 117
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 118
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
Treatment of EFA deficiency with dietary TG or PL in a murine model of extrahepatic cholestasis
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)
non-cholestatic EFAS control
weeks
0
10
20
30
40
0 3 4 5 6 71 2 8
TG
PL
EFAD
*
#
#
Bodyweight
(gra
m)
non-cholestatic EFAS control
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 119
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 120
121
Treatment of EFA deficiency with dietary TG or PL in a murine model of extrahepatic cholestasis
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 EFADPL TG EFAD
**
PL TG EFAD
mol
%
EFAS
AA (20: 4n-6)
0
6
8
10
12
4
2
**
PL TG EFADPL TG EFAD
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 121
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 122
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
Treatment of EFA deficiency with dietary TG or PL in a murine model of extrahepatic cholestasis
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
PL TG EFADPL TG EFAD
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 123
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 124
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
Treatment of EFA deficiency with dietary TG or PL in a murine model of extrahepatic cholestasis
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
proefschrift_def_v0870605.qxp 7-6-2005 22:45 Pagina 125
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 126
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
Treatment of EFA deficiency with dietary TG or PL in a murine model of extrahepatic cholestasis
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 128
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
Treatment of EFA deficiency with dietary TG or PL in a murine model of extrahepatic cholestasis
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 130
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
proefschrift_def_v010605def.qxp 2-6-2005 1:30 Pagina 131
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 132
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 133
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 134
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
Oral treatment of EFA deficiency with TG or PL in children with end-stage liver diease
pre-OLT cholestatic patients
RBC-EFA
plasma liver fuctions
plasma vitamin A, E
dietary EFA intake questionnaire
RBC-EFA
plasma liver fuctions
plasma vitamin A, E
RBC-EFA
plasma liver fuctions
plasma vitamin A, E
RBC-EFA
plasma liver fuctions
plasma vitamin A, E
dietary EFA intake questionnaire
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 liver fuctions
plasma vitamin A, E
dietary EFA intake questionnaire
RBC-EFA
plasma liver fuctions
plasma vitamin A, E
RBC-EFA
plasma liver fuctions
plasma vitamin A, E
RBC-EFA
plasma liver fuctions
plasma vitamin A, E
dietary EFA intake questionnaire
EFA-rich PL
EFA-rich TG
no EFA supplement
month 1 month 2 month 3
t=0 t=1 t=2 t=3
RBC-EFA
plasma liver fuctions
plasma vitamin A, E
dietary EFA intake questionnaire
RBC-EFA
plasma liver fuctions
plasma vitamin A, E
RBC-EFA
plasma liver fuctions
plasma vitamin A, E
RBC-EFA
plasma liver fuctions
plasma vitamin A, E
dietary EFA intake questionnaire
EFA-rich PL
EFA-rich TG
no EFA supplement
month 1 month 2 month 3
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 135
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 136
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
Oral treatment of EFA deficiency with TG or PL in children with end-stage liver diease
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 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
primary sclerosing cholangitis
cystic intra+extrahepatic bile ducts
Alagille syndrome
biliary atresia
auto-immune hepatits
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
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
primary sclerosing cholangitis
cystic intra+extrahepatic bile ducts
Alagille syndrome
biliary atresia
auto-immune hepatits
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 138
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
Oral treatment of EFA deficiency with TG or PL in children with end-stage liver diease
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 4 8 12time (weeks)
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
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 4 8 12time (weeks)
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
0 4 8 12time (weeks)
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)
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 4 8 12time (weeks)
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 4 8 12time (weeks)
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
proefschrift_def_v010605def.qxp 2-6-2005 1:20 Pagina 139
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.
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 140
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
Oral treatment of EFA deficiency with TG or PL in children with end-stage liver diease
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
no supplementno supplement
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
no supplementno supplement
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
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 141
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
no supplementno supplement
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
no supplementno supplement
proefschrift_def_v010605def.qxp 2-6-2005 1:42 Pagina 142
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
Oral treatment of EFA deficiency with TG or PL in children with end-stage liver diease
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 144
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
Oral treatment of EFA deficiency with TG or PL in children with end-stage liver diease
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 146
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
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 148
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
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 149
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
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 150
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
General discussion and summary
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 152
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
General discussion and summary
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 154
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
Nederlandse samenvatting
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 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
Nederlandse samenvatting
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 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
Nederlandse samenvatting
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 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
Nederlandse samenvatting
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 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 hersen- en 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
Nederlandse samenvatting
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 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 hersen- en 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
Nederlandse samenvatting
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 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
Nederlandse samenvatting
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 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
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 163
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
proefschrift_def_v0870605.qxp 7-6-2005 23:05 Pagina 164
(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
proefschrift_def_v0870605.qxp 7-6-2005 23:05 Pagina 165
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
proefschrift_def_v160605.qxp 16-6-2005 21:50 Pagina 166
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
proefschrift_def_v0870605.qxp 7-6-2005 22:54 Pagina 167
proefschrift_def_v010605def.qxp 2-6-2005 1:21 Pagina 168