2013Digital Comprehensive Summaries of Uppsala Dissertations [603025]

ACTA
UNIVERSITATIS
UPSALIENSIS
UPPSALA
2013Digital Comprehensive Summaries of Uppsala Dissertations
from the Faculty of Medicine 898
Experimental and Clinical
Necrotizing Enterocolitis
NICLAS HÖGBERG
ISSN 1651-6206
ISBN 978-91-554-8655-6
urn:nbn:se:uu:diva-197754

Dissertation presented at Uppsala University to be publicly examined in Rosénsalen, Ing
95/96, NB, Akademiska Barnsjukhuset, Uppsala, Thursday, May 30, 2013 at 13:00 for the
degree of Doctor of Philosophy (Faculty of Medicine). The examination will be conducted
in Swedish.
Abstract
Högberg, N. 2013. Experimental and Clinical Necrotizing Enterocolitis. Acta Universitatis
Upsaliensis. Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of
Medicine 898. 47 pp. Uppsala. ISBN 978-91-554-8655-6.
Necrotizing enterocolitis (NEC), a severe inflammatory disorder of the gastrointestinal tract
with high morbidity and mortality, affects primarily preterm infants. The diagnosis represents
a challenging task, and no biomarker has been found to aid early diagnosis with high accuracy.
Microdialysis has been widely used to detect metabolites of anaerobic metabolism, enabling a
local and early detection of ischemia. This thesis aims to evaluate the possibility of detecting
intestinal ischemic stress in experimental and clinical NEC, by use of rectal intraluminal
microdialysis.
Intraluminal rectal microdialysis was performed on rats subjected to total intestinal ischemia.
Metabolites of ischemia were detectable in both ileum and rectum, with raised glycerol
concentrations and lactate/pyruvate ratios. Elevated concentrations of glycerol correlated to
increasing intestinal histopathological injury.
Experimental early NEC was induced in newborn rat pups, by hypoxia/re-oxygenation
treatment. Development of NEC was confirmed by histopathology. Elevated glycerol
concentrations were detected by rectal microdialysis.
The genetic alterations following experimental NEC in rat pups were studied with microarray.
Immunohistochemistry staining was performed for tight junction proteins claudin-1 and
claudin-8. Several genes were altered in experimental NEC, mainly genes regulating tight
junctions and cell adhesion. Immunohistochemistry revealed reduced expression of claudin-1.
A prospective study was conducted on preterm infants with a gestational age of less than 28
weeks. The infants were admitted to a neonatal intensive care unit, and observed during a 4-week
period. Rectal microdialysis was performed twice a week, and blood was drawn for analysis of
I-FABP. A total of 15 infants were included in the study, whereof four infants developed NEC,
and 11 served as controls. Rectal glycerol and I-FABP displayed high concentrations, which
varied considerably during the observation periods, both in NEC and controls. No differences
in either glycerol or I-FABP concentrations were seen in the NEC-group vs. the controls.
In conclusion, rectal microdialysis can detect metabolites of intestinal ischemia, both in
experimental and clinical NEC. Rectal microdialysis is safe and could provide a valuable non-
invasive aid to detect hypoxia-induced intestinal damage or ischemic stress in extremely preterm
infants. In this study however, it was not possible to predict the development of clinical NEC
using microdialysis or I-FABP.
Keywords: Necrotizing, Enterocolitis, Ischemia, Microdialysis, Intraluminal, I-F ABP
Niclas Högberg, Uppsala University, Department of Women's and Children's Health,
Paediatric Surgery, Akademiska barnsjukhuset, ing. 95 SV , SE-751 85 Uppsala, Sweden.
© Niclas Högberg 2013
ISSN 1651-6206
ISBN 978-91-554-8655-6
urn:nbn:se:uu:diva-197754 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-197754)

If you are unable to find the truth right where you are,
where else do you expect to find it?
Dōgen

Till Lina, Lukas och Klara

List of Papers
This thesis is based on the following papers, which are referred to in the text
by their Roman numerals.

I Högberg, N. Carlsson, PO. Hillered, L. Meurling, S. Stenbäck,
A. Intestinal ischemia measured by intraluminal microdialysis .
Scandinavian Journal of Clinical and Laboratory Investigation,
2012; 72(1):59 –66.
II Högberg, N. Carlsson, PO. Hillered, L. Stenbäck, A. Engstrand
Lilja, H. Intraluminal intestinal microdialysis detects markers
of hypoxia and cell damage in experimental necrotizing enter o-
colitis. Journal of Pediatric Surgery, 2012;47:1646 –1651 .
III Högberg, N. Stenbäck, A. Carlsson, PO. Wanders, A . Engstrand
Lilja, H. Genes regulating tight ju nction s and cell adhesion are
altered in a rat model of experimental necrotizing enterocolitis.
Submitted m anuscript .
IV Högberg, N. Carlsson, PO. Hillered, L. Stenbäck, A. Larsson,
A. Engstrand Lilja, H. Intestinal intraluminal glycerol and I –
FABP levels in preterm infants with necrotizing enterocolitis.
Manuscript.

Reprints were made with permission from the respective publishers.

Contents
Introduction ………………………….. ………………………….. ………………………….. ….. 9
 
Necrotizing enterocolitis ………………………….. ………………………….. …………. 9
 
Experimental necrotizing enterocolitis ………………………….. ………………… 13
 
Microdialysis, general principles ………………………….. ………………………… 14
 
Microdialysis in intestinal ischemia ………………………….. ……………………. 15
 
Aims of the investigation ………………………….. ………………………….. ………….. 17
 
Materials and methods ………………………….. ………………………….. ……………… 18
 
Animals (papers I -III) ………………………….. ………………………….. ………. 18
 
Study population and setting (paper IV) ………………………….. ………….. 18
 
Surgical procedure (paper I) ………………………….. ………………………….. 19
 
Induction of experimental NEC (papers II -III) ………………………….. …. 19
 
Microdialysis (papers I -II, IV) ………………………….. ……………………….. 19
 
Blood sampling (papers I, IV) ………………………….. ……………………….. 20
 
Enzyme -linked immunosorbent assay (paper IV) ………………………….. 21
 
Morphology (papers I -III) ………………………….. ………………………….. …. 21
 
Microarray analysis (paper III) ………………………….. ………………………. 21
 
Immunohistochemistry staining (paper III) ………………………….. ……… 22
 
Statistics (papers I -III) ………………………….. ………………………….. ……… 22
 
Results ………………………….. ………………………….. ………………………….. ……….. 24
 
Paper I ………………………….. ………………………….. ………………………….. .. 24
 
Paper II ………………………….. ………………………….. ………………………….. . 26
 
Paper III ………………………….. ………………………….. …………………………. 27
 
Paper IV ………………………….. ………………………….. …………………………. 30
 
Discussion ………………………….. ………………………….. ………………………….. ….. 34
 
Conclusions ………………………….. ………………………….. ………………………….. … 39
 
Acknowledgements ………………………….. ………………………….. ………………….. 40
 
References ………………………….. ………………………….. ………………………….. ….. 42

Abbreviations
NEC
VLBW
NO
iNOS
ET-1
I-FABP
pO 2
pCO 2
BE
Hb
mRNA
r2
Il
kDa
TJ
CAR Necrotizing enterocolitis
Very low birth weight (<1500g)
Nitric oxide
Inducible NO synthetase
Endothelin -1
Intestinal fatty acid binding protein
Oxygen partial tension
Carbon dioxide partial tension
Base excess
Hemoglobin
Messenger Ribonucleic acid
Correlation coefficient
Interleukin
Kilodaltons, molecular weight
Tight junction, cell adhesion
Coxsackie – and A denovirus receptor
ELISA Enzyme -linked immunosorbent assay

9 Introduction
Necrotizing enterocolitis
Necrotizing enterocolitis (NEC) is the most common gastrointestinal eme r-
gency occurring in neonates. It is largely a disease affecting preterm infants ,
where more than 90% of the infants who develop NEC are born premature.
The risk of developing NEC is directly related to decreasing gestati onal age
and birth weight (1). Among infants with ve ry low birth weight (VLBW
<1500 g), the incidence is between 10 to 15% (1-4). With mortality rates
approaching 30 % in VLBW infants , NEC represents a significant clinical
problem. The incidence of NEC has increased in parallel with the improved
survival of infants born before 24 weeks of gestation (1). An inverse rel a-
tionship between the gestational age and onset o f the disease has been
demonstrated, and l ate onset NEC may occur many we eks after birth.
NEC affects the gastrointestinal tract and, in severe cases, can have pr o-
found systemic impact. NEC results in variable degrees of ischemic damage
of the intestine s, ranging from mild ischemia of the intestinal mucosa to
trans mural necrosi s of the bowel wall. Both the small and large bowel can be
affected , but more commonly it is only the small bowel , and the most severe
lesions are often seen in the jejunum and ileum. The typical patient with
NEC is a preterm infant with abdominal distention and bloody stools deve l-
oping after enteral feedings are initiated. A p athological abdominal x -ray
reveals bowel dilation and pneumat osis intestinalis .
Prematurity is the dominant risk factor , besides perinatal asphyxia, con-
genital heart disease and pulmonary disease. Initial m edical treatment for
NEC involves bowel rest with total parenteral nutrition , broad -spectrum
antibiotics, gastric decompression through a nasogastric tube, and correction
of hematologic and metabolic abnormalities. Indications for surgical inte r-
vention are intestinal perforation, or abdominal erythema . Ideally, surgery is
performed when the intestine is gangrenous but not perforated. Objective
staging criteria developed by Bell (1A/B to 3A/B) have been used to deter-
mine the appropriate therapy according to disease severity.
Approximately 70% of NEC patients survive. However, 50% develop
long-term complications. The two most common complications are poor
neurodevelopmental outcome and short bowel syndrom e, a malabsorption
state resulting from the removal of excessive or critical portions of the small
bowel , necessary for absorption of essential nutrients. Also, in testinal stri c-

10 tures can develop with or without a prece ding perforation, and result in bo w-
el obstruction (1, 5).
Despite intensive studies over the past 4 0 years, the etiology and pathophy s-
iology of NEC remains elusive. The disease is characterized by inflamm a-
tion of the bowel that can progress to intest inal necrosis, sepsis and multiple
organ failure.
Although th e pathophysiology of NEC still remains unclear, current ev i-
dence suggests a multifactorial cause (3). Prema turity is the main risk factor,
due to immaturity of gastrointestinal motility and immune defense (6), along
with an impaired intestinal barrier function (7). Other contributing factors
are thought to be genetic predisposition (8-10), bacterial overgrowth and
translocation (6), and enteral feeding and intestinal ischemia (11).
Intesti nal ischemia is considered to be a pivotal factor in the pathogenesis of
NEC (1-3, 12-14). Hypoperfusion or hypoxia results in variable degrees of
necrosis of the small and large intestine, ranging from local damage of the
intestinal mucosa to tra nsmural necrosis , leading to perforation of the bowel
wall. It has been noted that infants exposed to intrauterine environments
marked by compromis ed placental blood flow (i.e. maternal hypertension,
pre-eclampsia ) have an increased incidence of NEC , although NEC never
occurs in utero (15). Similarly, infants with post -natally diminished systemic
blood f low, as occurs in patent ductus arteriosis or congenital heart disease s,
also have an in creased incidence of NEC (16-23), probably due to shunting
blood away from the visceral organs ( the Herring -Breur reflex). Furthe r-
more , newborns respond to ischemic events with a further vasoconstriction,
suggesting a diminished ability to increase oxygen uptake in re sponse to
ischemic stress. Recently, i t has been shown that clinical NEC is predisposed
by a mucosal compromise of blood flow (24), further aggravated when fee d-
ings are initiated leading to NEC stage 2 or 3 (definitive NEC) .
Intestinal ischemia or hypoxia initially results in mucosal damage, since it is
the part of the intestinal wall, which is the most vulnerable to hypoxia. This
is thought to be the result of the counter -current mechanism (25), where
shunting of oxygen between the afferent and efferent arterioles in the villi
takes place , leading to lowered pO 2 at the villus tip. In infants, this effect is
further aggravated due to the high affinity of oxygen to fetal hemoglobin ,
resulting in reduced oxygen delivery (26, 27). In newborns, the regulation o f
the intes tinal vascular resistance, and thus blood flow, is mainly determined
by the balance between vasoconstriction and vasodilation. Vasoconstriction
is mediated by Endothelin -1 (ET -1), and vasodilation is mediated by the
endothelial production of free radical n itric oxide (NO). In the case of end o-
thelial cell dysfunction, such as ischemia -reperfusion injury, or inflammat o-
ry activity, the balance between vasoconstri ction and dilation is altered in
favor of constriction.

11 It is unusual for NEC to occur in the fast ing neonate , where 90% of cases
occur after feeding has been initiated (15). In a pig -model of NEC (28), a
rapid transition from parenteral to enteral nutrition induces histol ogical and
immunological changes, with altered expression of IL -1, 6, 18, TJ -proteins
(claudins), Toll -like receptors TLR -4 and TNF -alpha, suggesting a feeding –
induced inflammatory activation. Although this enteral feeding can activate
this inflammatory ca scade, other stress factors were needed (bacteria, inte s-
tinal ischemia, hypothermia) to induce animal experimental NEC (29-31).
Introduction of feedings provide carbohydrates, which in the presence of
bacterial overgrowth leads to fermentation of lactose into hydrogen gas. This
in turn leads to bowel distention and gas in the intestinal wall, pneumatosis
intestinalis , or gas in the vena portae . Bacterial overgrowth may primarily be
a result of immature bowel movements in infant s. Several common bacteria
have been found in NEC, but no specific pathogen ic microbe has been ide n-
tified. A broad range of microbes generally found in the di stal gastrointest i-
nal tract, have been recovered from the peritoneal cavity and blood of infants
with NEC. The predominant organisms include E nterobacteriaceae (i.e.,
Escherichia coli, Klebsiella pneumoniae), C lostridium spp., enteric path o-
gens (Salmonellae, Coxsackie B virus, Coronavirus, R otavirus), and pote n-
tial pathogens (Bacteroides fragilis, Cronobacter sakazakii ) (32-34). Regar d-
less of the microorganism present, infants, nevertheless , suffer from bacterial
overgrowth due to immature bowel movements. Outbreaks of NEC have also
been found to be caused by Cronobacter sakazakii, a conta minant of milk
powder formula (35).
The intestinal permeability is central to the defense against microbes. Ische-
mia leading to mucosal damage or necrosis results in breach of the intestinal
barrier, allowing for bacterial translocation and migration of bacterial end o-
toxin into the damaged tissu e, which further aggravates the inflammatory
cascade.
Intestinal permeability is increased in experimental NEC (36), in preterm
infants (37) as well as in infants suffering from NEC (38). In human NEC,
this increase in permeability could be secondary to ischemic damage to the
entire bowel wall or a primary structural immaturity. Increas ed intestinal
permeability as a primary cause leading to NEC may be considered, invol v-
ing structural proteins such as tight junctions (TJ) and cell adhesion proteins .
Increased permeability resulting from the "opening " of TJs may also be the
result of viru s-induced disruption of TJ formation (39). Other factors may
contribute to disrupt the function of TJs, such as intraluminal presence of
polyunsaturated fatty acids and other toxins (39, 40).
The i ntestinal mucosal cell layer is an important part of this barrier to i n-
traluminal microbes. The TJs are essential component s of this barrier, as
they bond the adjacent mucosal cells to each other. The junctional comple x-
es of the plasma membrane are not only epithelial barriers in paracellular

12 transport or barriers preventing diffusion in the plasma membrane, but also
contain proteins involved in signal transduction and the maintenance of the
physiological epithelial cell state. The TJ is composed of
both intracellular and membrane spanning proteins . Occludin, claudin, jun c-
tional adhesion molecules (JAMs), and the Coxsackie – and Adenovirus r e-
ceptor (CAR) are the major components of TJs (40). CAR is a trans –
membrane protein , which functions as a primary receptor for both the Cox-
sackie B virus and Adenovirus. CAR also functions as a marker for epithelial
TJs and regulation of permeability (43). The expression of these proteins,
functioning as receptors for Coxsackie B virus and Adenovirus, is related to
intestinal permeability and immune response. Binding of the virus to the
receptor leads to loss of T J-function and virus entry into the cell, activating
caspase -mediated apoptosis and T -cell activation.
The inflammatory reaction can be initiated by ischemia /reperfusion injury ,
and the introduction of feedings (11, 25, 28). The inflammatory reaction is
further augmented by an immature immune system, both specific (B – and T –
cell mediated) and other factors such as decreased intraluminal pH, reduced
cell adhesion , and immature intestinal motility. Inflammatory components
such as tumor necrosis factor α (TNF -α) and platelet activating factor (PAF)
have been shown to play an important part. Elevated levels of PAF lead to
further vasoconstriction and activation of TNF -α and interleukins such as
IL-6, mediators of an inflammatory response. Genetic studies in a pig model
of NEC has demonstrated an altered expression of innate immune defense
genes such as interleukins (IL -1 alpha, IL -6, IL -18), nitric oxide synthetase,
tight junction proteins (claudins), T oll-like receptors (TLR -4), and TNF –
alpha (28). In human enterocytes, the expression of TLR2, TLR4, NF kappa
B1, and IL -8 mRNA was increased in fetal vs. mature human enterocytes
and furthe r altered in NEC enterocytes (41). Another study suggests that
TLR genetic variants can alter susceptibility to NEC in VLBW infants (8).
Ischemia/reperfusion injury results in production of free oxygen radical s.
Previously, genetic analysis in experimental NEC demonstrated an up – regu-
lation in the antioxidant glutathione system (GSH) (42). The GSH antiox i-
dant system was shown to play a crucial role in intestinal barrier protection
by attenuating enterocyte death by caspase -mediated apoptosis.
From observing twins, there has been some speculation about a genetic pr e-
disposition for develop ing NEC, searching for a c andidate gene, but ev i-
dence has been scarce. Toll -like receptors have been identified as a genetic
variant in NEC (8). The difficulty with NEC is that it is a multi -factorial
disease and proc eeds through different stages, during which the different
inflammatory m ediators have different activities .

13 Today , NEC is diagnosed by a combination of clinical, laboratory, and r a-
diological findings ; however, these diagnostic methods lack high specificit y
and sensitivity for NEC . The accuracy of different biomarkers in diagnosing
NEC and in testinal ischemia has been studied extensively (44-48). D-lactate,
alpha glutathione S -transferase, intestinal fatty acid binding proteins, clau-
din-3, creatine kinase B, isoenzymes of lactate dehydrogenase (LD) , and
alkaline liver phosphatase (ALP) have been analyzed. D -dimer may be used
as an exclusi on test fo r intestinal ischemia; however, it lacks specificity (46)
and its role in NEC is uncertain.
Intestinal fatty acid binding protein (FABP 2, I-­‐FABP) has been reported
to be a useful plasma marker for early enterocyte cell death (47, 48). I-­‐FABP
is specifically present in mature enterocytes of the small and large intestine
and is released as soon as the integrity of the cell membrane is compromised.
I-­‐FABP is present in very small amounts in the plasma of healthy individ u-
als, probably representing the normal turnover of enterocytes, but levels rise
rapidly after episodes of acute intestinal ischemia and inflammation, inclu d-
ing NEC (47-51). Because of its low molecular weight, I -­‐FABP is present in
the systemic circulation and passes through the glomerular filter and can
readily be detected in the urine. Thus urinary values of I -­‐FABP as well as
plasma levels provide specific information about the number of dying inte s-
tinal epithelial cells and can be used in early diagnosis of NEC or intestinal
necrosis of other origin (49, 52-54).
Calprotectin is common ly used in the clinic to diagno se inflammatory
bowel disease. It has been proposed as a useful marker of NEC. Patients with
NEC have been shown to have a raised fecal c alprotectin level at the time of
diagnosis compared with controls (45, 55, 56). However, it does not seem to
be useful in the early stages of the disease (55), where no difference was
found between the controls and infants developing NEC at a later stage.
Claudins are part of the TJ, and have been evaluated as a marker of TJ –
degradation following ischemia or hypoxia -induced intestinal damage in
NEC. Decreased urinary levels of claudin -3 have been noticed in both expe r-
imentally induced intestinal damage, and in human inf lammatory bowel
disease (45, 57).
Experimental necrotizing enterocolitis
Experimental NEC has previously been studied by inducing NEC with h y-
poxia/re -oxygenation treatment in newborn rats (17, 18, 58), or a combin a-
tion of hypoxia/re -oxygenation treatment, cold stress and formula gavage
feeding (16-23). The induction of intestinal necrosis was verified by hist o-
pathology, revealing histopathological findings similar to what can be seen
in patients with NEC. Intestinal ischemia has also been studied extensively
in animal models of NEC.

14 Another method to induce experimental NEC in rat s is to infect them with
Cronobacter sakazakii (59). This method however, does not include any
hypoxia/re -oxygenation treatment.
Microdialy sis, general principles
Microdialysis is a technique to monitor the chemistry of the extracellular
space in living tissue. It enables monitoring of essentially any chemical
event taking place in the interstitial fluid (60).
A microdialysis probe is usually constructed as a concentric tube where
the perfusion fluid enters through an inner tub e, flows to its distal end, exits
the tube , and enters the space between the inner t ube and the outer dialysis
membrane. The semi -permeable membrane at the distal end of the microd i-
alysis catheter functions like a blood capillary and c hemical substances from
the extracellular fluid diffuse through the dialysis membrane into the cath e-
ter and the dialysate . Thereafter, it contain s a representative proportion of the
tissue fluid’s molecules. The dialysate consists of a physiological salt sol u-
tion, slowly pumped through the microdialysis probe where the solution is
equilibrated with the surro unding extracellular tissue fluid (Figure 1) .
The gradient of a particular compound depends not only on the difference
in concentration between the perfusate and the extracellular fluid but also on
the velocity of flow inside the microdialysis probe. The a bsolute recovery
(mol es/time unit) of a substance from the tissue depends on the “cut-off” of
the dialysis membrane , (usually defined as the molecular weight in Daltons
at which 80% of the molecules are prevented from passing through the
membrane), the len gth of the membrane , the flow rate of the perfusion fluid ,
and the diffusion coefficient of the compound through the extracellular fluid.
The advantage of microdialysis is the possibility to detect early signs of
tissue ischemia, before any clinical signs are evident. It provides semi –
continuous monitoring of tissue -specific metabolic changes, without the
need for repeated blood samples. Microdialysis samples substances involved
in the energy metabolism – glucose, lactate , and pyruvate – are markers for
ischemia, hypoxia , and hypoglycemia in peripheral and central tissues.

15
Figure 1. The principle of microdia lysis. A semipermeable membran e permits diff u-
sion of extracellular substances into the tube, allowing analysis of the metabolites of
interest. Reprinted with permission.
Glycerol is a marker for lipolysis or cell membrane damage (61, 62). Glut a-
mate is a marker of cytotoxicity in brain tissue (62). Urea is a marker for
urea clearance during hemodialysis.When the supply of glucose and oxygen
is diminished, there is an immediate increase of lactate/pyruvate ratio, and a
decrease of glucose. Obstruction of the blood fl ow will , thus, be detected
immediately by the change in metabolites.
Microdialysis in intestinal ischemia
Acute intestinal ischemia is a significant problem in clinical practice, both in
terms of treatment and diagnosis. It can manifest itself in general conditions
such as sepsis, multiple organ failure, necrotizing enterocolitis in neonates,
or following vascular surgery with clamping of the aorta, as well as in sp e-
cific intestinal conditions such as thrombosis or embolus in the mesenteric
arteries, volv ulus, and strangulation. The symptoms are often not apparent
until the intestinal ischemia is severe, leading to systemic responses, intest i-
nal bleeding , or perforation. Treatment at this stage often requires bowel
resections, and is associated with very high morbidity and mortality. Ische m-
ic or anoxic events are associated with a wide range of clinical presentations
both systemic and in various organ systems. Several indicators or markers
have been used to study ischemia. Systemic responses to severe ische mic
events and the following anaerobic metabolism such as elevated plasma lac-
tate and acidosis are late signs. Locally detectable markers of anoxia or i s-
chemia have the advantage of early detection of abnormalities. Local p hysi-
cal measurements of blood flo w through laser doppler flowmetry, pH, ATP,

16 pO 2, and pCO 2 have all been used with varying degrees of success. Blood
flow redistribution and heterogenous microcirculatory perfusion can explain
maintained regional aerobic metabolism during ischemic stress, despite local
signs of mesenteric hypoperfusion measured with laser doppler (superior
mese nteric arterial blood flow, mucosal mic rocirculation), and decreased
mucosal pCO 2 (63). Also, decreasing glucose concentrations suggest that
substrate supply may become crucial before oxygen consumption decreases.
Ischemic injury also leads to increased intestinal wall perm eability, but it
shows an uncertain correlation with the degree of necrosis (64).
Recent studies on intestinal ischemia with the microdialysis technique in
adult animals (13, 64-67) showed a typical metab olic response to anaerobic
metabolism; specifically, reduced glucose levels, production of lactate lea d-
ing to an elevated lactate/pyruvate ratio, accompanied by increased glycerol
levels as a result of cell -membrane phospholipid degradation caused by i s-
chemia-induced phospholipase activation. Markers of metabolism such as
glucose, lactate, pyruvate and markers of cell deterioration such as glycerol
and glutamate can be measured through microdialysis, and is a direct indic a-
tor of anaerobic metabolism and cel l injury.
Microdialysis has been used in various applicatio ns to monitor ischemic
events. Examples include i ntraperitoneal monitoring after intestinal surgery
and monitoring of free flaps after reconstructive surgery (67-69). Intrahepa t-
ic microdialysis has also been used as a method for early detection of pos t-
operative complications such as liver ischemia and graft rejection after liver
transplantation (70, 71). The microdialysis technique is minimally invasive
and suitable for different locations such as intraparenchymatous, intraperit o-
neal, intravasal (72), or intestinal intraluminal approach (64, 73-76).
Different locations have been used to measure these metabolic events. In
studies of intestinal ischemia, the intraperitoneal or serosal location close to
the affected site has been used extensively (69, 77-79), along with the i n-
traluminal approach. An intramural approach has also b een used, but the
inflammatory reactions around the probe were severe, affecting metabolism
and measurements thereof (65).
The intestinal mucosa seems to be most vulnerable to ischemic events,
leading to intraluminal release of the mentioned metabolites (64). The intr a-
peritoneal (serosal) compartment reflects the metabolism of the outer layers
of the intestinal wall (77). As the muscularis layer is less vulnerable to i s-
chemic stress than the mucosa, detection of anaerobic metabolites on the
serosal side reflects a later stage of the ischemic insult (64). This suggests
that the intraluminal compartment is the best location for detecting b i-
omarkers of anaerobic metabolism and cell decay at an early stage in intest i-
nal ischemia .

17 Aims of the investigation
The general aim of this investigation was to evaluate the microdialysis tec h-
nique on experimental and clinica l NEC. We were interested in its potential
use as a diagnostic tool, and to study the metabolic response following the
development of experimental and clinical NEC.
Specific aims were:

1. To evaluate the possibility of measuring ischemic damage to the
intestines, by placement of a rectal microdialysis catheter.
2. To set up an animal model of experimental early NEC.
3. To use the microdialysis technique on rat pups with experimental
NEC.
4. To study the genetic alterations in experimental NEC in rat pups
5. To study the potential diagnostic ability of intralu minal microdia l-
ysis compared to biomarkers and clinical routine diagnostic met h-
ods on preterm infants suffe ring from NEC.
In the first study , we evaluate d the metabolic response in the ileu m and the
colon by intraluminal rectal microdialysis during total intestinal ischemia.
The metabolic changes were also correlated to the histopathological findings
in the muc osa.
In the second study , we studied the possibility of detecting signs of hy-
poxic mucosal cell damage by using intraluminal rectal microdialysis in
newborn rat pups treated with hypoxia/re -oxygenation as an e xperimental
model of early NEC.
In the third study, we ex amined the genetic expression in the ileum from
rat pups induced with experim ental NEC , compared with controls. We also
used immunohistochemistry staining to study the specific expression of TJ
proteins.
In the fourth study, we evaluated rectal intraluminal microdialysis and
plasma I -FABP in a prospective, clinical setting. Extremely preterm infants
were foll owed during a four -week period, during which some of them deve l-
oped NEC.

18 Materials and methods
Animals (paper s I-III)
In paper I , adult male out -bred Sprague -Dawley rats were used. Their mean
weight was 360 g (range 280 -425 g), and the total number of rats used was
28. The rats were purchased from Scanbur AB (Sollentuna, Sweden). They
were housed in th e animal department at Uppsala Biomedical Cente r, and
kept at +22 șC, with a 12 -hour light/dark cycle, fed standard pellet food and
water ad libitum.
In paper s II and III , time -pregnant Sprague -Dawley rats were obtained on
day 15 of gestation. The rats were purchased from Charles River (Charles
River GmbH, Sulzfeld, Germany). They were housed in the animal depar t-
ment at Uppsala Biomedical Center, and ke pt at +22 șC, with a 12 -hour
light/dark cycle, fed standard pellet food and water ad libitum. On day 21,
the rat pups were delivered vaginally.
The studies were approved by the regional ethics committee on animal re-
search, and performed in accordance w ith the Guide for the Care and U se of
Laboratory Animals published by the National Research Council in 1996.

Study population and setting (paper IV)
In paper IV, preterm infants of gestational age <2 8 weeks and weighing
<1500g were included . The study was approved by the regional committee
on medical research ethics, and informed consent was obtained from the
parents. No abdominal symptoms or other clinical signs of illness were pr e-
sent in the infants on inclusion , and there was no evidence of any infectious
or otherwise complicating disease. The infants were admitted to a level III
neonatal intensive care unit, an d followed during a 4-week period. Any ro u-
tine blood testing and x -ray scans were performed o n a clinical basis. A total
of 15 infants wer e included during this period. Four infants developed clin i-
cal NEC stage 2 or 3, confirmed by radiology or histopathology . The r e-
maining 11 did not develop any complications or abdominal symptoms du r-
ing this period, and therefore served as controls.
The diagnosis of NEC was staged according to a simplification of the
Walsh and Kliegman classification (80), using three categories. Stage 1 was
defined as suspected NEC, with le thargy, abdominal distention, apnea , and

19 bloody stools . Stage 2 was considered present when either X -rays or ultr a-
sound revealed pneumatosis intestinalis, portal gas, intestinal perfo ration , or
paralysis with dilated bowel loops. Stage 3 was defined when the occurrence
of organ failure was present , in addition to the s tage 2 criteria.

Surgical procedure (paper I )
All animals were anaesthetized with thiobutabarbital sodium (Inactin®; Sig-
ma-Aldrich Swed en AB, Stockholm, Sweden ), at a dose of 120 mg/k g body
weight administered intraperitoneally. They were placed on a heated opera t-
ing table, maintained at body temperature (37°C) , and tracheostomized.
Heparinized polyethylene catheters were inserted into the right carotid artery
and jugular vein. The right carotid artery catheter was connected to a tran s-
ducer in order to continuously monitor the mean arterial blood pressure. The
arterial catheter was also used for blood gas sampling. The jugular vein cat h-
eter was used for continuous infusion (5 mL/h/kg body weight) with Ring-
er´s solution (Fresenius Kabi AS, Halden, Norway ) to compensate for loss of
body fluid . A laparotomy was performed and the aorta and cranial mesente r-
ic artery were identified. The aorta was permanently clamped proximal to
the cranial mesenteric artery, to create a state of total ischemia in the bow els.
In the sham -operated controls, no clamping of the aorta was performed.
Induction of e xperimental NEC (paper s II-III)
The rat pups were treated on the first day after birth. They were stressed with
hypoxia/re -oxygenation treatment , breathing 100% carbon dio xide for 10
minutes, followed by re-oxygenation using 100% oxygen for another 10
minutes. The rats wer e then returned to th eir mother s' cages and allowed ad
libitum nursing of maternal milk.

Microdialysis (paper s I-II, IV)
In paper I , all measurements were made using a CMA 20 Elite microdialysis
catheter (cut-off 20 kD a, 10 mm membrane length, CMA Microdialysis AB,
Stockholm, Sweden). The microdialysis catheters were connected to a m i-
croinjection pump (CMA 102 or CMA 402, CMA Microd ialysis AB) and
perfused with isotoni c Ringer´s solution with a flow -rate of 1 µL/minute. A
microdialysis catheter was placed in the sigmoid part of the colon through
the rectum. Another microdialysis catheter was placed in the ileum, through
a small ente rotomy that was five cm proximal to the caecum, and secured
with a suture. Another catheter was placed in the subcutaneous adipose ti s-

20 sue in the upper part of the b ody close to the scapula. A period of 30 minutes
was allowed for stabilization and in -situ p erfusion before the baseline mea s-
urement, which consisted of one 30 -minute period, before the aorta was
clamped. Microdialysate samples were then collected every 30 minutes for a
total of 210 minutes.
In paper II , all measurements were performed using a CM A 20 Elite m i-
crodialysis catheter (cut -off 20 kD a, 10 mm membrane length, CMA Micr o-
dialysis AB). The microdialysis c atheter was rectally inserted 10 mm, reac h-
ing the rectosigmoid part of the colon. The microdialysis catheters were
connected to the microinj ection pumps (CMA 102 and CMA 402, CMA
Microdialysis AB) and perfused with isotoni c Ringer´s solution with a flow –
rate of 0.7 µL/minute. Initially, in situ stabilization was allowed for 30
minutes. Microdialysate samples were then collected every 30 minutes for a
total of 90 minutes.
In paper IV, all measurements were performed using the clinically a p-
proved CMA 70 microdialysis catheter (cut -off 20 kD a, 10 mm membrane
length, Mdialysis AB, Solna, Sweden). The microdialysis catheters were
connected to the microinjection pumps (CMA 107, CMA Microdialysis AB )
and perfused with isotonic Ringer´s solution with a flow rate of 1.0
µL/minute. The infan ts were monitored during a 4 -week period, with micr o-
dialysis measurements twice a week. The micr odialysis catheter w as rectally
inserted 10 mm, and secured in position with tape. Initially, in situ stabiliz a-
tion was allowed for 5 minutes. Microdialysate samples were then collected
every 30 minutes for a total of 90 minutes.
Samples were immediately put in a freezer at -20°C. Analyses of glucose,
L-lactate, pyruvate , and glycerol were done using an enzymatic colorimetric
technique on a CMA 600 Microdialysis Analyz er (CMA Microdialysis AB).
The CMA 600 Analyz er was automatically calibrated at startup and re –
calibrated every sixth hour using standard calibration solutions from the
manufacturer (CMA Microdialysis AB ). Quality controls at two different
concentrations for each analyte were performed every weekday. Total i m-
precision coefficient of variation was <10% for all analytes.
Blood sampling (paper s I, IV)
In paper I, a rterial blood gas samples were collected at baseline and at the
end of the ischemic period. Me asurements were made for pH, pO 2, pCO 2,
base ex cess (BE) , and h emoglobin (H b), using an iSTAT -1 analyzer (iSTAT
Corporation, East Windsor, NJ, USA).
In paper IV, blood samples were drawn for analysis of I -FABP . Routine
testing of C-reactive protein was done using the standard clinical laboratory
method .

21 Enzyme -linked immunosorbent assay (paper IV)
During a 4 -week period, b lood was drawn twice a week for analysis of I –
FABP. A volume of 100 µL in EDTA was centrifuged, and the plasma was
stored at -70°C until analysis. I-FABP was analyzed by a commercial sand-
wich ELISA (DY3078, R&D Systems, Minneapolis, MN, USA), in which a
monoclonal antibody specific for I -FABP was coated onto microtitre plates.
Standards and samples were pipetted into the wells and the peptide was
bound to the immobiliz ed antibodies. After washing, a biotinylated anti -I-
FABP antibody was added. Following incubation and washing , a streptav i-
dine-HRP con jugate was added to the wells. After incubation and washing, a
substrate solution was added. The development was stopped and the absor b-
ance was measured in a SpectraMax 250 (Molecula r Devices, Sunnyvale,
CA, USA). The concentrations in the samples were determined by compa r-
ing the optical density of the sample with the standard curve. The assays
were calibrated against highly purified recombi nant human I -FABP.
Morphology (p apers I-III)
In paper I , intestinal specimens were taken from the site of the sigmoid eum
microdialysis catheter, at the end of the ischemic period. Intestinal spec i-
mens were also taken from sham -operated controls at the en d of the exper i-
ments.
In paper s II and III , intestinal specimens were taken from the ileum . The
specimens were fixed in 4 % formaldehyde solution, further processed , and
embedded in paraffin. Sections of 5 µm were stained with hematoxylin and
eosin and exa mined by light microscopy . The speci mens were evaluated
blindly with respect to the degree of ischemic in jury and mucosal integrity
by a pathologist .
Microarray analysis (paper III )
Twenty -four hours after induction of NEC , all animals were sacrificed by
neck dislocation. A laparotomy was performed and intestinal specimens
were taken from the distal part of the small bowel. The intestinal specimens
were immediately placed in RNAlater solution (Qiagen GmbH, Hilden,
Germany) and stored at +6°C over night. Ex traction of mRNA was pe r-
formed using RNeasy Mini kit (Qiagen GmbH) according to the manufa c-
turer´s protocol.
The RNA concentration was measured with an ND-1000 spectrophotom e-
ter (NanoDrop Technologies, Wilmington, DE, USA) and the RNA quality
was evaluated using the Agilent 2100 Bioanalyzer system (Agilent Techno l-
ogies Inc., Palo Alto, CA, USA). Two hundred fifty nanograms of total RNA
from each sample was used to generate amplified and biotinylated sense –

22 strand cDNA from the entire expressed genome , accord ing to the Ambion
WT Expression Kit (P/N 4425209 Rev B 05/2009) and Affymetrix
GeneChip® WT Terminal Labeling and Hybridization User Manual (P/N
702808 Rev. 1, Affymetrix Inc., Santa Clara, CA, USA). GeneChip® ST
Arrays (GeneChip® Rat Gene 1.0 ST Array) we re hybridized for 16 hours in
a 45°C incubator, rotated at 60 rpm. According to the GeneChip® Expre s-
sion Wash, Stain and Scan Manual (PN 702731 Rev . 2, Affymetrix Inc.) , the
arrays were then washed and stained using the Fluidics Station 450 and f i-
nally sca nned using the GeneChip® Scanner 3000 7G.
Immunohistochemistry staining (paper III)
Tissue sections from the specimens were de -paraffinized in xylene, and then
rehydrated in graded alcohols , according to standard procedures. Sections
were heated in Tris E DTA -buffer, pH 9 (DAKO, S2367) in a microwave
oven at 750 W for 10 minutes, followed by 350W for 15 minutes for antigen
retrieval. The sections were allowed to cool for 20 minutes and then washed
in distilled water. The tissue sections were blocked and sta ined by immun o-
histochemistry techniques with antibodies directed against the tight junction
proteins claudin -1 (Invitrogen Corporation, Carlsbad, CA, USA, no. 51 –
9000, dilution 1:200) and claudin -8 (Invitrogen Corporation, no. 40 -0700Z,
dilution 1:2000). Immunohistochemistry was visualized by the use of Dako
REAL Envision Peroxidase /DAB detection system (Dako Denmark A/S,
Glostrup, Denmark) followed by hematoxylin counterstaining.
The epithelial cells of the intestinal villi in the immunohistochemistry
sections were assessed blindly by two investigators , in a semi -quantitative
manner. The claudin -1 stained cell membranes displayed only a negative or
a very faint staining. Consequently, only the cytoplasmic staining was
scored. Claudin -8 displayed a distin ct lateral membrane staining and a cyt o-
plasmic staining. In this case, both the staining of the membrane and the
cytoplasm were scored separately. Regarding the cytoplasmic staining, the
presence of small distinct granules in the cytoplasm was regarded as 1+, the
presence of medium sized granules or clumps as 2+, and the presence of
large sized granules or clumps was scored as 3+. Absence of granules was
judged as negative. Regarding the membranes, a weak or negative staining
intensity was defined as −/1+, moderately strong staining as 2+, and strong
staining 3+.
Statistics (p apers I-III)
For analytical statistics in paper I , the non -parametric Kruskal -Wallis test
was used , and for pairwise comparison between time points, the Tukey post –
hoc test was applied (Statview; Abacus Concepts, Berkeley, California,

23 USA). In part two of the paper, the data showed a normal distribution pa t-
tern, and one -way ANOVA was used with Holm -Sidak post -hoc test.
For analytical statistics in paper II , a Mann -Whitney U -test was pe r-
formed for all comparisons (Statview; Abacus Conc epts). A p-value < 0.05
was considered statistically significant for all comparisons in paper s I and II .
In paper III , subsequent analysis of the gene expression data was carried
out in the freely available statistical computing language R ( http://www.r –
project.org ) using packages available from the Bioconductor project
(www.bioconductor.org ). The raw data was normalized usin g the robust
multi -array average (RMA) method first suggested by Li and Wong in 2001
(81, 82). In order to search for the genes that were expressed differentially in
the NEC samples compared with the control samples, an empirical Bayes
moderated t -test was then applied (83), using the “limma” package (84). To
address the problem of multiple testing, the p -values were adjusted using the
method of Benjamini and Hochberg (85). A p-value < 0.01 was considered
statistically significant for these compari sons.
For analytical statistics of the immunohistochemistry staining scores,
Fischer´s exact test was performed (Statview, Ab acus Concepts ). A p-value
<0.05 was considered statistically significant.

24 Results
Paper I
In the first series of experiments, we studied the intraluminal rectal microd i-
alysis analyte levels of glucos e, lactate, pyruvate , and glycerol in adult rats
subjected to total intestinal ischemia.
On clamping the suprarenal aorta, immediate responses could be detected
by microdialysis. Intraluminal glycerol started to increase immediately after
clamping of the a orta, and reached a maximum at 240 minutes, with all va l-
ues higher than baseline levels, p< 0.001 (Figure 2).
Luminal lactate also started to increase directly after clamping of the ao r-
ta. Subcutaneous lactate increased gradually, approaching the luminal levels
at the end of the ischemic period.
Expressing the ratio between lactate and pyruvate levels, the luminal ratio
began to rise directly after clamping of the aorta, to reach the maximum at
the end of the ischemic period. The luminal lactate/pyruvate r atio was also
higher than at baseline, p< 0.01 (Figure 2).
Glycerol
Minutes0 50 100 150 200 250 300micromol/L
050100150200250300350
Sham control, lumen
Luminal
SubcutaneousLactate/Pyruvate Ratio
Minutes0 50 100 150 200 250 30005001000150020002500
Luminal ratio
Subcutaneous ratio
Figure 2. Intraluminal glycerol started to increase immediately after clamping the
aorta, and continued to rise during the ischemic period. Luminal lactate also i n-
crease d, although at a later stage. The lactate/pyruvate ratio also responded with
prompt elevation.

25 In the sham -operated control group, subcutaneous and intraluminal glycerol
and lactate values remained at baseline levels throughout the experiments.
In this stu dy, we also compared the intraluminal glycerol levels in the ileum
vs. the colon (Figure 3).
Glycerol
Minutes0 50 100 150 200 250micromol/L
0200400600800
Colon
Ileum
sham control
Figure 3. Intraluminal levels of glycerol in the colon and the ileum. The concentr a-
tions in the ileum were higher than the colon, and both differed from baseline levels
and sham controls .
Glycerol was measured in the colon and the ileum. B oth began to rise after
clamping of the aorta, reaching maximum levels at 210 minutes. Glycerol in
the ileum and the colon were both higher than baseli ne and sham operated
controls at 30 minutes and throughout the ischemic period ( p< 0.001). Gly c-
erol levels measured in the ileum were higher than glycerol levels in the
colon, at 90 minutes and at the end of the ischemic period ( p< 0.05).
Luminal lactate levels also increased directly after clamping of the aorta.
Colon lactate levels increased gradually, approaching the ileum lactate levels
at the end of the ischemic period. Lactate in the colon and the ileum were
different from baseline ( p< 0.001).
In the sham -operated control group, glycerol and lactate values in both the
colon and the ileum remained at baseline levels throughout the experiments.
Microscopic evaluation of the mucosal i ntegrity was made, using sham –
operated controls (N=3) , and ischemic s pecimens (N=6). The degree of
damage was classified into three degrees, and were evaluated by an ind e-
pendent observer . A linear regression was made on the grade of the mucosal
damage and the concentrations of luminal glycerol in each animal at 240
minutes. Correlation coefficient ( r²) was 0.705, p< 0.01. Increasing levels of
glycerol correlated to higher degrees of mucosal damage.

26 Visualized as a scatter plot, the luminal glycerol concentration is e x-
pressed on the y -axis, and the mucosa l integrity grade on the x -axis
(Figure 4).
Regression
Mucosal integrity0,5 1,0 1,5 2,0 2,5 3,0 3,5Glycerol
0100200300400500
Mucosal integrity vs Glycerol
Regression
Figure 4. Linear r egression on the grade of mucosal damage v ersus intraluminal
concentrations of glycerol in the colon. Increasing levels of glycerol correlate to
higher degrees of mucosal damage. N=9. Correlation coefficient was 0 .705.
Paper II
In the following experiments, we performed all microdialysis measurements
on rat pups with induced experimental early NEC. Intraluminal rectal micro-
dialysate levels of glucose, lactate , and glycerol were detectable. Levels of
pyruvate were too low to be measured in both the controls and NEC rats.
Therefore, no lactate/pyruvate quotients could be calculated.
The intraluminal levels of glucose, lactate , and glyce rol in controls were
compar ed with NEC rats. Glycerol and lactate levels were hig her in the NEC
group compared to the controls, p< 0.05 (Figure 5 ). Intraluminal levels of
glucose did not differ between the two groups.

27 Glycerolmicromol/L
020406080
Controls NEC *Lactatemmol/L
0,00,20,40,60,81,01,21,41,61,8
NEC Controls*

Figure 5. Intraluminal concentration of glycerol and lactate was higher in NEC
compared to control rats. Asterisk denotes significant difference compared with
controls .
Paper III
In this study, we analyzed the genetic transcriptome of rat pups with induced
experimental early NEC , and performed immunohistochemistry staining for
TJ proteins claudin 1 and 8 . The results were compared to healt hy rat pups.
We analyzed the transcriptional profile of gen es that were either up – or
down regulated at least two -fold, and with an adjusted p-value < 0.01 for each
sample. The total number of genes expressed at these levels of significance
was 509. The complete dataset comprised 26,000 genes. The genes were
further analyzed using the DAVID (Database for Annotation, Visualization
and Integrated Discovery ) bioinformatics software v ersion 6.7 provided by
the National Institute of Allergy and Infectious Diseases
(http://david.abcc.ncifcrf.gov/ ).
Among the genes that were up – or downregulated, several are involv ed in
TJ formation and cell adhesion, such as claudins (1, 8, 14, 15) and gap jun c-
tion protein beta 3. Occludin was not significantly upregulated (Table 1).
Fatty acid binding proteins (FABP) 5 and 6 (ileal) were upregulated, wher e-
as FABP2 (intestinal) was downregulated. Steroid 5 -alpha reductase, which
controls intestinal maturation and development during embryogenesis and
early development, was the most downregulated gene.

28
Table 1. Among the genes that were up – or down regulated, s everal are involved in
tight junction formation and cell adhesion , as well as inflammatory response and
apoptosis.
Gene name Symbol Fold
change Log2 Adj. p –
value Function
Claudin 1 Cldn1 -1.43 0.000297
414 Tight junction
Claudin 8 Cldn8 3.14 7.63E -7 Tight junction
Claudin 14 Cldn14 -1.13 0.004930
587 Tight junction
Claudin 15 Cldn15 -1.22 0.000578
282 Tight junction
Occludin Ocln 0.55 0.001948
6 Tight junction

Gap junction protein, beta 3 Gjb3 1.54 0.000556
758 Gap junction
Fatty acid binding protein 2 Fabp2 -0.52 0.002 Cytosolic protein, inte s-
tinal
Fatty acid binding protein 5 Fabp5 1.36 7.85E -5 Cytosolic protein, ep i-
dermal
Fatty acid binding protein 6 Fabp6 2.14 7.4E -7 Cytosolic protein, ileum
Interleukin 1 beta Il1b -1.95 1.04E -5 Cytokine activity

Interleukin 18 Il18 -1.13 0.001638
442 Cytokine activity
TNF receptor -associated factor
6 Traf6 1.01 0.000132
556 NFKB1
Caspase 3 Casp3 -1.06 0.000566
715 Apoptosis
5-alpha reductase type 2 Srd5a2 -6.48 3.96E -5 Steroid biosynthetic
process
Glutathione S-transferase A2 Gsta2 3.74 0.005376
712 Antioxidation
RT1 class II, locus Db1 RT1Db1 -1.41 0,000182
78 MHCII
RT1 class II, locus DM beta RT1DMb -1.35 3.04E -5 MHCII
RT1 class II, locus Ba RT1Ba -1.21 4.37E -5 MHCII
RT1 class II, locus Da RT1Da -1.09 0.000255
672 MHCII

29
Genes regulating the inflammatory response and apoptosis were also found
altered, including IL -1, IL -18, TNF alpha, caspase 3, and NF-kappa -beta 1
(TNF receptor associated factor 6).
Several downregulated genes included RT1 class II loci Ba, Da , and Db1
(major histocompatibility complex (MHC) class II family), MHC class II
loci DM beta and antigen E alpha. These are involved in the formation of the
MHC class II, functioning as receptors for viruses in antigen presenting
cells. Glutath ione S -transferase A2 was upregulated, as were several genes
involved in glutathione transferase activity: glutathione S -transferase Yc2
subunit , glutathione S -trans ferase mu 2, 3, 4 , and 7 and glutathione S –
transferase theta 3.
The immunohistochemical staining for claudin -1 was scored only with
regard to the intra cytoplasmic components (Figure 6 A, B). Regarding the
staining of claudin -1, the NEC -group received a lo wer score than the control
group ( p= 0.005). In the NEC group, 2 rats received a moderate or strong
staining score (2+ or 3+), and 10 rats received a low (1+ or negative) stai n-
ing score. In the control group, 7 rats received a moderate or strong (2+ or
3+) staining score, and only one rat was scored low (1+). The immunohist o-
chemical staining for claudin -8 was also scored, but no difference in score s
could be seen between the two groups. In the NEC group, 8 rats received a
moderate or strong score (2+ or 3+) , and 4 rats received a low score ( 1+ or
negative). In the control group , 2 rats were scored as moderate or strong (2+
or 3+), whereas 6 rats were scored as low (1+ or negative).
A B
Figure 6. A. Moderate to strong lateral membrane staining of claudin -1 as seen
mostly in the control group (original magnification x400). B. Moderate to strong
staining of claudin -1 indicated by the presence of large intracytoplasmic aggregates.
This was seen in ma ny animals in the NEC group and also in a few animals in the
control group (original magnification x400). A very similar picture was observed in
the claudin -8 stained intestinal epithelium.

30 Paper IV
This paper presents a prospective , clinical pilot study on preterm infant s. We
included infants with a gestational age of less than 28 weeks and weighing
<1500g after obtaining parental informed consent . A total of 15 infants wer e
included during this period. Four infants devel oped signs of NEC, whereof
three died during the observation period. T he remaining 11 did not de velop
NEC , and therefore served as controls.

A B
C D
Figure 7. A-D. Glycerol, I -FABP , and CRP levels in four infants developing NEC.
Glycerol values are medians with bars for maximum and minimum values. Birthdate = 2011−11−275 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
NECAd mortem
Antibiotics
Respirator!
!!
!!0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!!!!0.1 0.2 0.5 1 2 5 10 20 50
I−FABP (ng/ml)
Glycerol
CRPI−FABPBirthdate = 2012−03−045 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
NECPneu
Antibiotics
Respirator!!
!!
!
!0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!
!!!0.1 0.2 0.5 1 2 5 10 20 50!I−FABP (ng/ml)
Glycerol
CRPI−FABP
Birthdate = 2013−01−245 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
NECAd mortem
Antibiotics
Respirator!
!!!
!0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!!!
!!!
!0.1 0.2 0.5 1 2 5 10 20 50
I−FABP (ng/ml)
Glycerol
CRPI−FABPBirthdate = 2011−11−275 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
NECSepsis
Operation
Antibiotics
Respirator!
!
!
!!!
!
!!!!0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!!
!!
!!!
!!0.1 0.2 0.5 1 2 5 10 20 50
I−FABP (ng/ml)
Glycerol
CRPI−FABP

31 Microdialysis catheters were placed in the rectum for intraluminal measur e-
ments of glucose, glycerol, lactate , and pyruvate. B lood sample s were taken
for analysis of I -FABP. Routine testing of C -reactive protein (CRP) was
performed on a clinical basis.
Intraluminal microdialysate lev els of glycerol were detectable; however , the
concentrations of lactate, glucose , and pyruvate were too low for analysis for
all infants. During the observation periods, the concentrations of glycerol, I –
FABP , and CRP varied considerably in infants with NEC (Figure 7A -D) and
in the controls (Figure 8 A -K). The mean levels of glycerol o r I-FABP at
NEC diagnosis were not higher than before clinical diagnosis, nor compared
to the controls.
In control infants, both glycerol and I -FABP levels also revealed a high
degree of variation, with rising and falling concentrations during the obse r-
vation periods (Figure 8A -K). One infant developed symptoms of intestinal
ischemia , but no signs of NEC on x -ray (Figure 8K). Laparotomy revealed
expansive intestinal necrosis, as a result of mid -gut volvulus. In this patient,
I-FABP clearly displayed an earl y elevation, accompanied by a later rise of
glycerol and CRP.

A B Birthdate = 2012−09−255 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
Pneu
Antibiotics
Respirator!! !0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!
!!! !!!!0.1 0.2 0.5 1 2 5 10 20 50
I−FABP (ng/ml)
Glycerol
CRPI−FABPBirthdate = 2012−09−245 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
Antibiotics
Respirator!!
!0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!!!
!!!!!0.1 0.2 0.5 1 2 5 10 20 50
I−FABP (ng/ml)
Glycerol
CRPI−FABP

32 C D

E F
G H Birthdate = 2012−04−095 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
Sepsis
Antibiotics
Respirator!!!
!
!!!0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!
!!!!!!!0.1 0.2 0.5 1 2 5 10 20 50
I−FABP (ng/ml)
Glycerol
CRPI−FABPBirthdate = 2012−04−025 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
Ad mortem
Antibiotics
Respirator!!
!!0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!
!!
!
!0.1 0.2 0.5 1 2 5 10 20 50
I−FABP (ng/ml)
Glycerol
CRPI−FABP
Birthdate = 2012−02−025 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
Operation
Antibiotics
Respirator!!!
!
!
! !0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!!
!!!
!!!0.1 0.2 0.5 1 2 5 10 20 50
I−FABP (ng/ml)
Glycerol
CRPI−FABPBirthdate = 2012−02−025 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
Operation
Antibiotics
Respirator!
!!
!0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!!!
!!
!!
!!0.1 0.2 0.5 1 2 5 10 20 50
I−FABP (ng/ml)
Glycerol
CRPI−FABP
Birthdate = 2011−12−125 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
PneuAd mortem
Antibiotics
Respirator!!
!!!
!0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!!!!!0.1 0.2 0.5 1 2 5 10 20 50
I−FABP (ng/ml)
Glycerol
CRPI−FABPBirthdate = 2013−02−025 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
Antibiotics
Respirator!
!!
!0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!
!!!
!
!0.1 0.2 0.5 1 2 5 10 20 50!I−FABP (ng/ml)
Glycerol
CRPI−FABP

33 I J
K
Figure 8. A-J. Glycerol, I -FABP , and CRP levels in ten infants without clinical or
radiological signs of NEC. Values for glycerol are medians with bars for maximum
and minimum values. Glycerol and I -FABP levels varied considerably during the
observation periods, with both rising and fall ing concentrations at different time
points. Two infants ( E, F) were operated on for patent ductus arteriosus. K. One
infant suffered from expansive intestinal necrosis due to mid -gut volvulus.
All infants were intubated and ventilation was maintained on r espirator du r-
ing the observation periods as indicated (Fi gure 7A -D, 8A -K). Three of them
suffered from pneumothorax due to ventilator trauma (Figure 7B, 8A, G).
Antibiotic treatment was also initiated in all children, and maintained for
different periods a s indicated (Figure 7A -D, 8A -K). Two of the infants di s-
played a significant persistent ductus arteriosus, and were operated on there-
after (Figure 8E, F). Severe infection and sepsis was present in one patient
(Figure 8C). Birthdate = 2013−01−215 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
Ad mortem
Antibiotics
Respirator!!
!!
!!0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!
!!
!!!!!0.1 0.2 0.5 1 2 5 10 20 50
I−FABP (ng/ml)
Glycerol
CRPI−FABPBirthdate = 2013−02−025 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
Antibiotics
Respirator!
!
!0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!!!!0.1 0.2 0.5 1 2 5 10 20 50
I−FABP (ng/ml)
Glycerol
CRPI−FABP
Birthdate = 2012−12−305 10 20 50 100 200 500 2000 5000
0 10 20 30 40 50
Days since birthGlycerol (µmol/L)
Ad mortem
Antibiotics
Respirator!!
! !!
!!
!!!0.1 0.5 1 2 5 10 20 50 200 500
CRP (mg/l)
!!! !!
!
!!0.1 0.2 0.5 1 2 5 10 20 50
I−FABP (ng/ml)
Glycerol
CRPI−FABP

34 Discussion
Today , NEC is diagnosed by a combination of clinical, laboratory, and r a-
diological findings. However, t hese diagnostic methods lack specificity and
sensitivity for NEC, especially in the early phase. Any indications of NEC
are evident at a late r stage during the course of the di sease, when intestinal
damage is manifest. Early diagnosis and treatment is important to reduce the
morbidity and mortality associated with NEC.
Microdialysis has previously been used to study intestinal isch emia, both in
humans and animals . The advantage of the microdialysis approach is that it
measures metabolites of anaerobic metabolism , locally , in the organ of inte r-
est. In early stage experimental NEC, the mucosa is pr imarily affected as it is
the part of the intestinal wall that is most sensitive to hypoxia (86).
In humans, patients with recent abdominal surgery have been monitored
with microdialysis (77, 87), and these studies have demonstrated that micr o-
dialysis is a valuable tool for detecting visceral ischemia using the intraper i-
toneal approach. However, no clinical studies have previously been pe r-
formed using the intraluminal approach, partly because of th e difficulty of
placing the microdialysis catheters in the lumen. The intraluminal approach ,
however, seems promising, and raises the opportunity to ident ify anaerobic
stress at an earlier stage, before systemic levels of the metabolites are
reached, and b efore the organ of interest is severely damaged.
To our knowledge, the studies in this thesis, are the first to apply intral u-
minal rectal microdialysis to detect hypoxic stress in extremely preterm i n-
fants, as well as in an animal model of NEC. Although it was noted that an
elevation of glycerol was seen in infants with NEC, it was not possible to
detect any significant increase in glycerol concentrations prior to clinical
diagnosis. In infants with co mplications other than NEC, intraluminal gly c-
erol values also varied considerably, both rising and falling during the o b-
servation period. Therefore it was not possible to differentiate between NEC
and the controls by observing any increase in glycerol conc entrations.
We have also for th e first time been able to analyz e plasma levels of I –
FABP in extremely preterm infants born before 28 weeks of gestation. Pla s-
ma levels of I -FABP displayed a similar pattern as glycerol, with high co n-
centrations before the de velopment of NEC, as well as in the control infants.
In this study, elevated intraluminal glycerol levels, as well as plasma I –

35 FABP, were detected in infants developing NEC as well as in those who had
no abdominal symptoms. These controls , however, were se verely ill, with
complications following extreme prematurity. Primarily, respiratory – and
ventilation -associated problems dominated, resulting in long periods with
low blood oxygen saturation levels. This relative hypoxic state may result in
a compromise d oxygenation of the intestines; particularly , the sensitive m u-
cosal cell -layer, which could result in a hypoxia -induced mucosal cell me m-
brane decay and release of glycerol and I -FABP into the intestinal lumen.
Other complications and diseases such as sepsis , infections, persi stent ductus
arteriosus, and an emia were also present in the control group, which aggr a-
vates the intestinal distress.
In a previous stud y of I -FABP as a diagnostic marker of intestinal isch e-
mia, the suggested cut -off point for non -rever sible intestinal ischemia was
1.3 ng/ml (88). In a study on healthy preterm infants with gestational age
between 28 and 33 weeks, the plasma concentrations ranged between 0.46 –
4.5ng/ml (50). An interesting finding in our present study is that the conce n-
trations of I-FABP, in controls as well as infants with NEC, exceeded these
levels even at an early stage. These findings may suggest that the previous
suggested cut -off point or normality range of I -FABP is not relevant in this
patient category of extremely prete rm infants. Infants with high enterocyte
turnover should theoretically display higher levels of I -FABP. Another e x-
planation could be that the high concentrations of I -FABP in the present
study reflect intestinal enterocyte damage. This fact is supported by the high
intraluminal concentrations of glycerol at an early stage in controls as well
as in infants later developing NEC. It is highly valuable to establish
knowledge regarding the normality range of I -FABP levels in this patient
category.
Our study on e xperimentally induced early NEC in rat pups, demonstrated
that intraluminal microdialysis can detect signs of hypoxic intestinal cell
damage (86). In the study, elevated levels of glycerol and lactate were mea s-
ured by placement of a rectally inserted microdialysis catheter.
Experimental NEC has previously been studied by inducing NEC through
a combination of asphyxia and cold stress on newborn rats or hypoxia/re –
oxygenation treatment only (17, 18, 58, 86, 89). The mode l that we used is
based on the induction of intestinal damage using hypoxia/re -oxygenation
treatment on rat pups, whereas NEC in human infants is more complex in its
pathogenesis. However, the histopathological findings and the genetic tra n-
scriptome in the present model indicate similarities to clinical NEC.
Using this model with newborn rats , subjected to hypoxia/re -oxygenation
treatment only, histological changes of moderate ischemic damage were
observed in the intestinal specimens from the ileum, and is chemic metabolic
changes were detected using intraluminal microdialysis in the rectosigmoid

36 part of the colon. The ability to detect these metabolites by rectal microdia l-
ysis is highly valuable when clinical NEC is suspected.
Ischemic or hypoxic stress to the intestinal mucosa initially leads to release
of the metabolites of anaerobic metabolism into the bowel lumen (12). We
have demonstrated that these metabolites are easily detectable using micr o-
dialysis in the bowel lumen by a rectal microdialysis cat heter . Increased
intraluminal levels of glycerol, lactate , and raised lactate/pyruvate ratio are
all markers of ischemic stress, mucosal disintegration , and cell membrane
decay. Increased luminal lactate levels have a positive correlation with pr o-
longed oc clusion of the superior mesenteric artery, but whether this is due to
spill-over from elevated systemic hyperlactatemia is uncertain (13, 74, 75).
An increased level of intraluminal lactate also correlates with ischemic stress
at an early stage, but also increases in subcutaneous adipose tissue, and is
therefore unable to differentiate between local and systemic ischemia.
Intraluminal rectal lactate ha s also been found to be ele vated during cor o-
nary artery by pass operations (90). However, intraluminal changes in lactate
levels, intesti nal permeability , or adenosine triphos phate levels do not corr e-
late with the extent of histopathological intestinal damage (64), whereas
elevated glycerol levels have a positive correlation to prolonged ischemic
insult and severity of intestina l damage (12, 13, 64).
In our previous study on experimental intestinal ischemia (12), signs of
intestinal damage in the lumen were measured by microdialysis, before sy s-
temic levels of the anaerobic metabolites were reached. Prolonged local in-
testinal ischemia eventually l eads to systemic levels of these anaerobic m e-
tabolites. The intraluminal levels of glycerol also had a positive correlation
to aggravated histological mucosal damage , in accordance with our previous
work (64). This suggests that glycerol is the best marker for mucosal damage
thus far.
Increased levels of glycerol are considered to origi nate from phospholipid
degradation of the cell membrane (61, 62). Systemic glycerol can be gene r-
ated through stress -induced lipolysis in response to anaerobic metabolism
(61, 62). As b owel permeability increases during ischemia (13), any in-
creased intraluminal glycerol levels could originate from spill over from the
plasma. However, this has not been supported in previous studies of intest i-
nal ischemia, where ele vated intraluminal glycerol was found along with low
plasma levels (64). In our firs t study of intestinal ischemia, clamping of the
suprarenal aorta creates a state of total ischemia, so no spill -over to the inte s-
tinal lumen is therefore possible (12). Glycerol can also be generated from
glucose in the bowel lumen (62); however, no difference could be found in
the intraluminal levels of glucose in this study.
A potential methodological problem of microdialysis is that it only measures
a relative concentration of the metabolites in the compartment of interest.

37 This fact makes it difficult to compare the absolute values of two different
measurements performed at different time intervals. To overcome the pro b-
lem of relative concentrations, ratios such as the lactate/pyruvate ratio are
often used. The lactate/pyruvate ratio is considered to be independent of
changes in relative recovery , making it a useful quantitat ive measure [18]. In
our studies of both experimental and clinical NEC , intraluminal levels of
lactate and pyruvate were too low to be measu red. The lactate/pyruvate ratio,
therefore , could not serve as an indicator of hypoxic damage in the intestines
in these setup s. The concentrations of glycerol , on the ot her hand, were
much higher and easily detectable well above the detection limit of the CMA
600 analyzer . A higher relative recovery could be achieved by using micro-
dialysis catheters with longer membran es or by using a lower perfusion flow
rate. This theoretically results in higher concentrations of lactate and p y-
ruvate, enabling calculation of the lactate/pyruvate ratio. Initially, we tried to
use a 30 mm membrane, but it was not possible due to the ana tomical limit a-
tions of the extremely preterm infants.
In our analysis of the transcriptome using microarray in early experimental
NEC, genes regulating TJs and cell adhesion were clearly affected secondary
to hypoxia. Increased intestinal permeability may be considered a major
factor in NEC pathogenesis, involving these structural proteins such as TJ
and other cell adhesions (20, 91-93). In human NEC, this increase in perm e-
ability could however be secondary t o ischemic damage to the entire intest i-
nal wall or a primary structural immaturity due to prematurity.
In this model of experimental NEC, we have found several T J genes to be
both up – and down regulated. The most important ones are the claudins 1, 8,
14, and 15. This was also supported by the immunohistochemical staining,
where TJ protein claudin -1 was less abundant in the NEC intestine compared
with the controls. Interestingly, claudin -1 was ma inly located in the intr a-
cytoplasmic compartment. It could be speculated that the intracellular local i-
zation of the TJ proteins may reflect a downregulation of the TJ proteins
from the outer membrane after cellular stress resulting in an increased ep i-
thelial permeability. Other junctional adhesion genes such as gap junction
protein and nucleoporin were also affected.
An increased bacterial translocation is also present in NEC, which could lead
to a secondary degradation of the TJs following inflammation. I n addition,
increased permeability resulting from the "un-lockning" of the TJs may be
the result of virus -induced disruption of TJ formation (39). Occludin, cla u-
din, junctional adhesion molecules (JAMs), and the Coxsackie – and Aden o-
virus receptor (CAR) are major components of TJs (40). These receptors are
used by the Coxsackie – and Adenoviruses to gain access to the cells, and TJ
viral receptors binding to viruses can also disrupt the paracellular barrier (39,
40). This could represent an alternative mechanism by which the intestinal

38 barrier function is destroyed, leading to increased permeability to both bact e-
ria and toxins. In this context , it is of special interest that virus receptor mo l-
ecules constituting part of the MCH class II were also downregulated. Alt-
hough previous research has focused on intestinal bacteria as a cause of
NEC, there is a possibility that viruses play a critical role in NEC pathogen e-
sis. However, the role of viruses in NEC pathogenesis is very unclear at pr e-
sent.
Moreover, alterations in previously identified genes involved in the i n-
flammatory response was confirmed, along with several genes regulating
proteins used as biomarkers for NEC. Better understanding of the genes i n-
volved in the pathogenesis of NEC may lead to novel strategies for the pr e-
vention and treatment of NEC.

39 Conclusion s
Intestinal intraluminal microdialysis can detect metabolites of ischemia or
hypoxia -induced damage , and cell membrane decay in rat models of total
intestinal ischemia and experimental early NEC in rat pups. A rectal micro-
dialysis catheter can be used for intraluminal measurements.
Microdialysate intraluminal glycerol is the best marker for early intesti nal
cell-membrane deterioration thus far, and correlates positively to increasing
mucosal damage. Glycerol is also a marker for hypoxic damage mucosal
cell-membrane decay in rat experimental early NEC.
In extremely preterm infants monitored during the neon atal period, elevated
intraluminal glycerol and plasma I -FABP indicates intestinal hypoxia in-
duced mucosal cell -membrane decay. This elevation is seen in infants deve l-
oping NEC, as well as in preterm infants with other complications resulting
in hypoxia, s uch as respiratory problems, sepsis, anemia, or persistent ductus
arteriosus .
In experimental early NEC, genes regulating the TJs and cell adhesion are
altered , accompanied by a decreased protein expression of claudin -1.
In conclusion, this thesis has demo nstrated that rectal intraluminal microd i-
alysis is safe and could provide a valuable non -invasive aid to detect hypo x-
ia-induced intestinal damage , or ischemic stress in extremely preterm i n-
fants. However , it was not possible to predict or differentiate NEC from ot h-
er diagnoses with hypoxia in preterm infants , by detecting elevated levels of
glycerol or I -FABP.

40 Acknowledg ements
Denna avhandling är utfört som ett samarbete mellan institutionen för kvi n-
nors och barns hälsa, samt institutionen för medicinsk c ellbiologi vid Upps a-
la Universitet. Finansieringen har kommit från institutionen för kvinnors och
barns hälsa vid Uppsala Universitet, samt från Kronprinsessan Lovisas
förening för barnasjukvård.

Jag skulle vilja tacka följande personer som gjort detta ar bete möjligt:

Huvudhandledare, docent Heléne Engstrand Lilja . Utan handledning är det
omöjligt att bedriva doktorandutbildning. Tack för allt ditt vänliga stöd, dina
idéer och din suveräna förmåga att arbeta snabbt och effektivt!

Bihandledare , med.dr Anders Stenbäck . Du har varit med mig ända sedan vi
startade detta projekt . Utan dig hade det aldrig blivit några råttförsök!

Bihandledare , professor Per-Ola Carlsson . Av någon outgrundlig anledning
lät du oss låna ditt lab på fysiologen för alla råttför sök. Tack för din a idéer,
hjälp och vänskap!

Professor Staffan Meurling . Du, jag och Anders startade detta projekt våren
2007. Utan dina idéer och ditt stöd i början hade det aldrig blivit något!

Docent Rolf Christoffersson . Tack för din insats som "hand ledare " i en tid av
osäkerhet. Dina synpunkter får mig alltid att fundera en gång extra!

Professor Lars Hillered . Tack för att du hjälpt mig ända sedan starten, med
excellenta synpunkter och tillgång till labbet på NIVA!

Docent Alkwin Wanders. Utan dina idéer och hjälp med histologin samt
immunfärgningarna hade detta arbete inte blivit detsamma!

Professor Anders Larsson. Tack för din vänlighet och s karpa synpunkter
angående analysen av I -FABP!

41 Verksamhetschef, docent Arne Stenberg . Utan dig hade jag varken jobbat
eller forskat på barnkirurgiska kliniken i Uppsala, tack för allt stöd och v ä-
nlighet under åren !

Ett stort tack till f .d. verksamhetschef professor Uwe Ewald och forsknings s-
juksköterska Cecilia Ewald på avdelning 95F. Utan hjälp är klinisk forskning
omöjlig att bedriva! Tack även till alla neonatologer för hjälp med att
rekrytera patienter!

Jag skulle även vilja tacka Inger Ståhl Myllyaho på NIVA och Inga Hansson
på patologen för ovärderlig hjälp i labbet!

Övriga kollegor på barnkirurgiska klinisken: Johan Danielsson (Mr Science),
Erik Sköldenberg , Elisabet Gustafson , Göran Läckgren , Gillian Barker , Ger-
trud Angsten , Peter Flacker , Ammar Al -Mashhadi, Thóra Ólafsdó ttir.

Tack även till professor Ulf Eriksson och pro fessor Leif Jansson på instit u-
tionen för medicinsk cellbiologi .

42 References
1. Lin, P. W., and Stoll, B. J. Necrotising enterocolitis. Lancet 368: 1271 -1283,
2006.
2. Henry, M. C., and Moss, R. L. Necrotizing enterocolitis. Annu Rev Me d 60:
111-124, 2009.
3. Neu, J., and Walker, W. A. Necrotizing enterocolitis. N Engl J Med 364: 255 –
264, 2011.
4. Petrosyan, M., Guner, Y. S., Williams, M., Grishin, A., and Ford, H. R. Current
concepts regarding the pathogenesis of necrotizing enterocolit is. Pediatr Surg
Int 25: 309 -318, 2009.
5. Hintz, S. R., Kendrick, D. E., Stoll, B. J., Vohr, B. R., Fanaroff, A. A., D o-
novan, E. F., Poole, W. K., Blakely, M. L., Wright, L., and Higgins, R. Neur o-
developmental and growth outcomes of extremely low birth we ight infants after
necrotizing enterocolitis. Pediatrics 115: 696 -703, 2005.
6. Neu, J., Chen, M., and Beierle, E. Intestinal innate immunity: how does it relate
to the pathogenesis of necrotizing enterocolitis. Semin Pediatr Surg 14: 137 –
144, 2005.
7. Upperman, J. S., Potoka, D., Grishin, A., Hackam, D., Zamora, R., and Ford, H.
R. Mechanisms of nitric oxide -mediated intestinal barrier failure in necrotizing
enterocolitis. Semin Pediatr Surg 14: 159 -166, 2005.
8. Sampath, V., Le, M., Lane, L., Patel, A. L. , Cohen, J. D., Simpson, P. M., Ga r-
land, J. S., and Hines, R. N. The NFKB1 (g. -24519delATTG) variant is assoc i-
ated with necrotizing enterocolitis (NEC) in premature infants. J Surg Res 169:
e51-57, 2011 .

9. Treszl, A., Heninger, E., Kalman, A., Schuler, A ., Tulassay, T., and Vasarhelyi,
B. Lower prevalence of IL -4 receptor alpha -chain gene G variant in very -low-
birth -weight infants with necrotizing enterocolitis. J Pediatr Surg 38: 1374 –
1378, 2003.
10. Treszl, A., Kocsis, I., Szathmari, M., Schuler, A., Heninger, E., Tulassay, T.,
and Vasarhelyi, B. Genetic variants of TNF -[FC12]a, IL -1beta, IL -4 receptor
[FC12]a -chain, IL -6 and IL -10 genes are not risk factors for sepsis in low -birth –
weight infants. Biol Neonate 83: 241 -245, 2003.
11. Nowicki, P. T. Ischemia and necrotizing enterocolitis: where, when, and how.
Semin Pediatr Surg 14: 152 -158, 2005.
12. Hogberg, N., Carlsson, P. O., Hillered, L., Meurling, S., and Stenback, A. Inte s-
tinal ischemia measured by intraluminal microdialysis. Scand J Clin Lab Invest
72: 59-66, 2012.
13. Solligard, E., Juel, I. S., Bakkelund, K., Johnsen, H., Saether, O. D., Gronbech,
J. E., and Aadahl, P. Gut barrier dysfunction as detected by intestinal lu minal
microdialysis. Intensive Care Med 30: 1188 -1194, 2004.
14. Fox, T. P., and Godavitarne, C. What really causes necrotising enterocolitis?
ISRN gastroenterology 2012 : 628317, 2012.

43 15. Berseth, C. L. Feeding strategies and necrotizing enterocolitis. Current opinion
in pediatrics 17: 170 -173, 2005.
16. Ozkan, K. U., Ozokutan, B. H., Inanc, F., Boran, C., and Kilinc, M. Does m a-
ternal nicotine exposure during gestation increase the injury severity of small i n-
testine in the newborn rats subjected to experim ental necrotizing enterocolitis. J
Pediatr Surg 40: 484 -488, 2005.
17. Meyer, K. F., Martins, J. L., Freitas Filho, L. G., Oliva, M. L., Patricio, F. R.,
Macedo, M., and Wang, L. [Evaluation of an experimental model of necrotizing
enterocolitis in rats]. Acta Cir Bras 21: 113 -118, 2006.
18. Meyer, K. F., Martins, J. L., de Freitas Filho, L. G., Oliva, M. L., Patricio, F. R.,
Macedo, M., and Wang, L. Glycine reduces tissue lipid peroxidation in hypoxia –
reoxygenation -induced necrotizing enterocolitis in rats. Acta Cir Bras 21: 161 –
167, 2006.
19. Travadi, J., Patole, S., Charles, A., Dvorak, B., Doherty, D., and Simmer, K.
Pentoxifylline reduces the incidence and severity of necrotizing enterocolitis in
a neonatal rat model. Pediatr Res 60: 185 -189, 2006.
20. Clark, J. A., Doelle, S. M., Halpern, M. D., Saunders, T. A., Holubec, H., Dv o-
rak, K., Boitano, S. A., and Dvorak, B. Intestinal barrier failure during exper i-
mental necrotizing enterocolitis: protective effect of EGF treatment. Am J Phy s-
iol Gastrointest Liv er Physiol 291: G938 -949, 2006.
21. Halpern, M. D., Holubec, H., Clark, J. A., Saunders, T. A., Williams, C. S.,
Dvorak, K., and Dvorak, B. Epidermal growth factor reduces hepatic sequelae
in experimental necrotizing enterocolitis. Biol Neonate 89: 227 -235, 2006.
22. Halpern, M. D., Holubec, H., Saunders, T. A., Dvorak, K., Clark, J. A., Doelle,
S. M., Ballatori, N., and Dvorak, B. Bile acids induce ileal damage during e x-
perimental necrotizing enterocolitis. Gastroenterology 130: 359 -372, 2006.
23. Jilling, T., Simon, D., Lu, J., Meng, F. J., Li, D., Schy, R., Thomson, R. B.,
Soliman, A., Arditi, M., and Caplan, M. S. The roles of bacteria and TLR4 in rat
and murine models of necrotizing enterocolitis. J Immunol 177: 3273 -3282,
2006.
24. Mannoia, K., Boskovi c, D. S., Slater, L., Plank, M. S., Angeles, D. M., and
Gollin, G. Necrotizing enterocolitis is associated with neonatal intestinal injury.
J Pediatr Surg 46: 81-85, 2011.
25. Shepherd, A. P., and Kiel, J. W. A model of countercurrent shunting of oxygen
in the intestinal villus. Am J Physiol 262: H1136 -1142, 1992.
26. Bustamante, S. A., Jodal, M., Nilsson, N. J., and Lundgren, O. Evidence for a
countercurrent exchanger in the intestinal villi of suckling swine. Acta Physiol
Scand 137: 207 -213, 1989.
27. Bustamante, S. A., Srkalovic, G., and Lundgren, O. Hemodynamics during
increased intraluminal pressure in the ileum of suckling swine. J Pediatr Ga s-
troenterol Nutr 9: 94-97, 1989.
28. Siggers, J., Sangild, P. T., Jensen, T. K., Siggers, R. H., Skovgaard, K., Stoy, A.
C., Jensen, B. B., Thymann, T., Bering, S. B., and Boye, M. Transition from
parenteral to enteral nutrition induces immediate diet -dependent gut histological
and immunological responses in preterm neonates. Am J Physiol Gastrointest
Liver Physiol 301: G435 -445, 2011.
29. Claud, E. C., and Walker, W. A. Hypothesis: inappropriate colonization of the
premature intestine can cause neonatal necrotizing enterocolitis. FASEB J 15:
1398 -1403, 2001.
30. Halpern, M. D., Holubec, H., Dominguez, J. A., Meza, Y . G., Williams, C. S.,
Ruth, M. C., McCuskey, R. S., and Dvorak, B. Hepatic inflammatory mediators

44 contribute to intestinal damage in necrotizing enterocolitis. Am J Physiol Ga s-
trointest Liver Physiol 284: G695 -702, 2003.
31. Waligora -Dupriet, A. J., Dugay , A., Auzeil, N., Huerre, M., and Butel, M. J.
Evidence for clostridial implication in necrotizing enterocolitis through bacterial
fermentation in a gnotobiotic quail model. Pediatr Res 58: 629 -635, 2005.
32. Lake, A. M., Lauer, B. A., Clark, J. C., Wesenb erg, R. L., and McIntosh, K.
Enterovirus infections in neonates. J Pediatr 89: 787 -791, 1976.
33. Johnson, F. E., Crnic, D. M., Simmons, M. A., and Lilly, J. R. Association of
fatal Coxsackie B2 viral infection and necrotizing enterocolitis. Arch Dis Child
52: 802 -804, 1977.
34. Brook, I. Microbiology and management of neonatal necrotizing enterocolitis.
Am J Perinatol 25: 111 -118, 2008.
35. Hunter, C. J., Petrosyan, M., Ford, H. R., and Prasadarao, N. V. Enterobacter
sakazakii: an emerging pathogen in infa nts and neonates. Surgical infections 9:
533-539, 2008.
36. Zani, A., Ghionzoli, M., Lauriti, G., Cananzi, M., Smith, V. V., Pierro, A., De
Coppi, P., and Eaton, S. Does intestinal permeability lead to organ failure in e x-
perimental necrotizing enterocoliti s? Pediatr Surg Int 26: 85-89, 2010.
37. Stratiki, Z., Costalos, C., Sevastiadou, S., Kastanidou, O., Skouroliakou, M.,
Giakoumatou, A., and Petrohilou, V. The effect of a bifidobacter supplemented
bovine milk on intestinal permeability of preterm infants. Early Hum Dev 83:
575-579, 2007.
38. Piena -Spoel, M., Albers, M. J., ten Kate, J., and Tibboel, D. Intestinal perme a-
bility in newborns with necrotizing enterocolitis and controls: Does the sugar
absorption test provide guidelines for the time to (re -)introduce enteral nutr i-
tion? J Pediatr Surg 36: 587 -592, 2001.
39. Coyne, C. B., Voelker, T., Pichla, S. L., and Bergelson, J. M. The co x-
sackievirus and adenovirus receptor interacts with the multi -PDZ domain pr o-
tein-1 (MUPP -1) within the tight junction. J Bio l Chem 279: 48079 -48084,
2004.
40. Hossain, Z., and Hirata, T. Molecular mechanism of intestinal permeability:
interaction at tight junctions. Mol Biosyst 4: 1181 -1185, 2008.
41. Nanthakumar, N., Meng, D., Goldstein, A. M., Zhu, W., Lu, L., Uauy, R., Ll a-
nos, A., Claud, E. C., and Walker, W. A. The mechanism of excessive intestinal
inflammation in necrotizing enterocolitis: an immature innate immune response.
PLoS One 6: e17776, 2011.
42. Kelly, N., Friend, K., Boyle, P., Zhang, X. R., Wong, C., Hackam, D. J., Zam o-
ra, R., Ford, H. R., and Upperman, J. S. The role of the glutathione antioxidant
system in gut barrier failure in a rodent model of experimental necrotizing e n-
terocolitis. Surgery 136: 557 -566, 2004.
43. Raschperger, E., Thyberg, J., Pettersson, S ., Philipson, L., Fuxe, J., and Petter s-
son, R. F. The coxsackie – and adenovirus receptor (CAR) is an in vivo marker
for epithelial tight junctions, with a potential role in regulating permeability and
tissue homeostasis. Exp Cell Res 312: 1566 -1580, 2006.
44. Thuijls, G., van Wijck, K., Grootjans, J., Derikx, J. P., van Bijnen, A. A.,
Heineman, E., Dejong, C. H., Buurman, W. A., and Poeze, M. Early diagnosis
of intestinal ischemia using urinary and plasma fatty acid binding proteins. Ann
Surg 253: 303 -308, 2011.
45. Thuijls, G., Derikx, J. P., van Wijck, K., Zimmermann, L. J., Degraeuwe, P. L.,
Mulder, T. L., Van der Zee, D. C., Brouwers, H. A., Verhoeven, B. H., van
Heurn, L. W., Kramer, B. W., Buurman, W. A., and Heineman, E. Non -invasive

45 markers for early diagnosis and determination of the severity of necrotizing e n-
terocolitis. Ann Surg 251: 1174 -1180, 2010.
46. Block, T., Nilsson, T. K., Bjorck, M., and Acosta, S. Diagnostic accuracy of
plasma biomarkers for intestinal ischaemia. Scand J Clin Lab Invest 68: 242 –
248, 2008.
47. Kanda, T., Fujii, H., Tani, T., Murakami, H., Suda, T., Sakai, Y., Ono, T., and
Hatakeyama, K. Intestinal fatty acid -binding protein is a useful diagnostic
marker for mesenteric infarction in humans. Gastroenterology 110: 339 -343,
1996.
48. Lieberman, J. M., Sacchettini, J., Marks, C., and Marks, W. H. Human intestinal
fatty acid binding protein: report of an assay with studies in normal volunteers
and intestinal ischemia. Surgery 121: 335 -342, 1997.
49. Edelson, M. B., Sonnino, R. E., Bagwell, C. E., Lieberman, J. M., Marks, W.
H., and Rozycki, H. J. Plasma intestinal fatty acid binding protein in neonates
with necrotizing enterocolitis: a pilot study. J Pediatr Surg 34: 1453 -1457,
1999.
50. Guthmann, F., Borchers, T., Wolfrum, C., Wus track, T., Bartholomaus, S., and
Spener, F. Plasma concentration of intestinal – and liver -FABP in neonates su f-
fering from necrotizing enterocolitis and in healthy preterm neonates. Mol Cell
Biochem 239: 227 -234, 2002.
51. Evennett, N., Alexander, N., Petro v, M., Pierro, A., and Eaton, S. A systematic
review of serologic tests in the diagnosis of necrotizing enterocolitis. J Pediatr
Surg 44: 2192 -2201, 2009.
52. Derikx, J. P., Evennett, N. J., Degraeuwe, P. L., Mulder, T. L., van Bijnen, A.
A., van Heurn, L. W., Buurman, W. A., and Heineman, E. Urine based detection
of intestinal mucosal cell damage in neonates with suspected necrotising enter o-
colitis. Gut 56: 1473 -1475, 2007.
53. Evennett, N. J., Hall, N. J., Pierro, A., and Eaton, S. Urinary intestinal fatt y
acid-binding protein concentration predicts extent of disease in necrotizing e n-
terocolitis. J Pediatr Surg 45: 735 -740, 2010.
54. Thuijls, G., Wijck, K., Grootjans, J., Derikx, J. P., van Bijnen, A. A., Heineman,
E., Dejong, C. H., Buurman, W. A., and Po eze, M. Early diagnosis of intestinal
ischemia using urinary and plasma Fatty Acid binding proteins. Ann Surg 253:
303-308, 2011.
55. Ayse Selimoglu, M., Temel, I., Yildirim, C., Ozyalin, F., Aktas, M., and
Karabiber, H. The role of fecal calprotectin and lactoferrin in the diagnosis of
necrotizing enterocolitis. Pediatr Crit Care Med , 2012.
56. Carroll, D., Corfield, A., Spicer, R., and Cairns, P. Faecal calprotectin conce n-
trations and diagnosis of necrotising enterocolitis. Lancet 361: 310 -311, 2003.
57. Thuijls, G., Derikx, J. P., de Haan, J. J., Grootjans, J., de Bruine, A., Masclee,
A. A., Heineman, E., and Buurman, W. A. Urine -based detection of intestinal
tight junction loss. J Clin Gastroenterol 44: e14 -19, 2010.
58. Canpolat, F. E., Yurdakok, M., Oz soy, S., Haziroglu, R., and Korkmaz, A. Pr o-
tective effects of recombinant human granulocyte colony stimulating factor in a
rat model of necrotizing enterocolitis. Pediatr Surg Int 22: 719 -723, 2006.
59. Hunter, C. J., Singamsetty, V. K., Chokshi, N. K., Bo yle, P., Camerini, V.,
Grishin, A. V., Upperman, J. S., Ford, H. R., and Prasadarao, N. V. Enterobacter
sakazakii enhances epithelial cell injury by inducing apoptosis in a rat model of
necrotizing enterocolitis. The Journal of infectious diseases 198: 586 -593, 2008.
60. Ungerstedt, U. Microdialysis –principles and applications for studies in animals
and man. J Intern Med 230: 365 -373, 1991.

46 61. Marklund, N., Salci, K., Lewen, A., and Hillered, L. Glycerol as a marker for
post-traumatic membrane phospholipi d degradation in rat brain. Neuroreport 8:
1457 -1461, 1997.
62. Hillered, L., Vespa, P. M., and Hovda, D. A. Translational neurochemical r e-
search in acute human brain injury: the current status and potential future for
cerebral microdialysis. J Neurotrauma 22: 3-41, 2005.
63. Krejci, V., Hiltebrand, L., Buchi, C., Ali, S. Z., Contaldo, C., Takala, J., Si g-
urdsson, G. H., and Jakob, S. M. Decreasing gut wall glucose as an early marker
of impaired intestinal perfusion. Crit Care Med 34: 2406 -2414, 2006.
64. Solligard, E., Juel, I. S., Bakkelund, K., Jynge, P., Tvedt, K. E., Johnsen, H.,
Aadahl, P., and Gronbech, J. E. Gut luminal microdialysis of glycerol as a
marker of intestinal ischemic injury and recovery. Crit Care Med 33: 2278 –
2285, 2005.
65. Emmertsen, K . J., Wara, P., Soerensen, F. B., and Stolle, L. B. Intestinal micr o-
dialysis –applicability, reproducibility and local tissue response in a pig model.
Scand J Surg 94: 246 -251, 2005.
66. Sommer, T. Microdialysis of the bowel: the possibility of monitoring intestinal
ischemia. Expert Rev Med Devices 2: 277 -286, 2005.
67. Sommer, T., and Larsen, J. F. Detection of intestinal ischemia using a microd i-
alysis technique in an animal model. World J Surg 27: 416 -420, 2003.
68. Norgren, L., and Jansson, K. Intraperit oneal and intraluminal microdialysis in
the detection of experimental regional intestinal ischaemia (Br J Surg 2004; 91:
855-861). Br J Surg 91: 1653, 2004.
69. Ungerstedt, J., Nowak, G., Ericzon, B. G., and Ungerstedt, U. Intraperitoneal
microdialysis (IPM): a new technique for monitoring intestinal ischemia studied
in a porcine model. Shock 20: 91-96, 2003.
70. Nowak, G., Ungerstedt, J., Wernerman, J., Ungersted t, U., and Ericzon, B. G.
Metabolic changes in the liver graft monitored continuously with microdialysis
during liver transplantation in a pig model. Liver Transpl 8: 424 -432, 2002.
71. Silva, M. A., Richards, D. A., Bramhall, S. R., Adams, D. H., Mirza, D . F., and
Murphy, N. A study of the metabolites of ischemia -reperfusion injury and s e-
lected amino acids in the liver using microdialysis during transplantation.
Transplantation 79: 828 -835, 2005.
72. Backstrom, T., Liska, J., Oldner, A., Lockowandt, U., an d Franco -Cereceda, A.
Splanchnic metabolism during gut ischemia and short -term endotoxin and he m-
orrhagic shock as evaluated by intravasal microdialysis. Shock 21: 572 -578,
2004.
73. Sommer, T., and Larsen, J. F. Intraperitoneal and intraluminal microdialys is in
the detection of experimental regional intestinal ischaemia. Br J Surg 91: 855 –
861, 2004.
74. Tenhunen, J. J., Jakob, S. M., and Takala, J. A. Gut luminal lactate release du r-
ing gradual intestinal ischemia. Intensive Care Med 27: 1916 -1922, 2001.
75. Tenhunen, J. J., Kosunen, H., Alhava, E., Tuomisto, L., and Takala, J. A. Inte s-
tinal luminal microdialysis: a new approach to assess gut mucosal ischemia. An-
esthesiology 91: 1807 -1815, 1999.
76. Juel, I. S., Solligard, E., Skogvoll, E., Aadahl, P., and Gr onbech, J. E. Lactate
and glycerol released to the intestinal lumen reflect mucosal injury and perme a-
bility changes caused by strangulation obstruction. Eur Surg Res 39: 340 -349,
2007.
77. Jansson, K., Ungerstedt, J., Jonsson, T., Redler, B., Andersson, M. , Ungerstedt,
U., and Norgren, L. Human intraperitoneal microdialysis: increased la c-

47 tate/pyruvate ratio suggests early visceral ischaemia. A pilot study. Scand J
Gastroenterol 38: 1007 -1011, 2003.
78. Jansson, K., Strand, I., Redler, B., Magnuson, A., Unge rstedt, U., and Norgren,
L. Results of intraperitoneal microdialysis depend on the location of the cath e-
ter. Scand J Clin Lab Invest 64: 63-70, 2004.
79. Sommer, T., and Larsen, J. F. Validation of intramural intestinal microdialysis
as a detector of intes tinal ischaemia. Scand J Gastroenterol 39: 493 -499, 2004.
80. Walsh, M. C., and Kliegman, R. M. Necrotizing enterocolitis: treatment based
on staging criteria. Pediatr Clin North Am 33: 179 -201, 1986.
81. Irizarry, R. A., Hobbs, B., Collin, F., Beazer -Barclay, Y. D., Antonellis, K. J.,
Scherf, U., and Speed, T. P. Exploration, normalization, and summaries of high
density oligonucleotide array probe level data. Biostatistics 4: 249 -264, 2003.
82. Li, C., and Wong, W. H. Model -based analysis of oligonucleotide arrays: e x-
pression index computation and outlier detection. Proc Natl Acad Sci U S A 98:
31-36, 2001.
83. Smyth, G. K. Linear models and empirical bayes methods for assessing diffe r-
ential expression in m icroarray experiments. Stat Appl Genet Mol Biol 3: Art i-
cle3, 2004.
84. Smyth, G. K. Bioinformatics and Computational Biology Solutions using R and
Bioconductor . New York: Springer, 2005.
85. Benjamini Y, H. Y. Controlling the false discovery rate: a practi cal and powe r-
ful approach to multiple testing. Journal of the Royal Statistical Society Series
B: 289 -300, 1995.
86. Hogberg, N., Carlsson, P. O., Hillered, L., Stenback, A., and Lilja, H. E. I n-
traluminal intestinal microdialysis detects markers of hypoxia and cell damage
in experimental necrotizing enterocolitis. J Pediatr Surg 47: 1646 -1651, 2012.
87. Jansson, K., Redler, B., Truedsson, L., Magnuson, A., Ungerstedt, U., and No r-
gren, L. Postoperative on -line monitoring with intraperitoneal microdialysis is a
sensitive clinical method for measuring increased anaerobic metabolism that
correlates to the cytokine response. Scand J Gastroenterol 39: 434 -439, 2004.
88. Vermeulen Windsant, I. C., Hellenthal, F. A., Derikx, J. P., Prins, M. H., Buu r-
man, W. A., Jaco bs, M. J., and Schurink, G. W. Circulating intestinal fatty acid –
binding protein as an early marker of intestinal necrosis after aortic surgery: a
prospective observational cohort study. Ann Surg 255: 796 -803, 2012.
89. Ozdemir, R., Yurttutan, S., Sari, F. N., Uysal, B., Unverdi, H. G., Canpolat, F.
E., Erdeve, O., and Dilmen, U. Antioxidant effects of N -acetylcysteine in a ne o-
natal rat model of necrotizing enterocolitis. J Pediatr Surg 47: 1652 -1657, 2012.
90. Solligard, E., Wahba, A., Skogvoll, E., Stense th, R., Gronbech, J. E., and A a-
dahl, P. Rectal lactate levels in endoluminal microdialysate during routine cor o-
nary surgery. Anaesthesia 62: 250 -258, 2007.
91. Shiou, S. R., Yu, Y., Chen, S., Ciancio, M. J., Petrof, E. O., Sun, J., and Claud,
E. C. Erythro poietin protects intestinal epithelial barrier function and lowers the
incidence of experimental neonatal necrotizing enterocolitis. J Biol Chem 286:
12123 -12132, 2011.
92. Rentea, R. M., Liedel, J. L., Welak, S. R., Cassidy, L. D., Mayer, A. N.,
Pritchard , K. A., Jr., Oldham, K. T., and Gourlay, D. M. Intestinal alkaline
phosphatase administration in newborns is protective of gut barrier function in a
neonatal necrotizing enterocolitis rat model. J Pediatr Surg 47: 1135 -1142,
2012.
93. Han, X., Fink, M. P. , and Delude, R. L. Proinflammatory cytokines cause NO* –
dependent and -independent changes in expression and localization of tight
junction proteins in intestinal epithelial cells. Shock 19: 229 -237, 2003.

Acta Universitatis Upsaliensis
Digital Comprehensive Summaries of Uppsala Dissertations
from the Faculty of Medicine 898
Editor: The Dean of the Faculty of Medicine
A doctoral dissertation from the Faculty of Medicine, Uppsala
University, is usually a summary of a number of papers. A few
copies of the complete dissertation are kept at major Swedish
research libraries, while the summary alone is distributed
internationally through the series Digital Comprehensive
Summaries of Uppsala Dissertations from the Faculty of
Medicine.
Distribution: publications.uu.se
urn:nbn:se:uu:diva-197754
ACTA
UNIVERSITATIS
UPSALIENSIS
UPPSALA
2013