Università degli Studi di Trieste

PhD program in MOLECULAR MEDICINE

PhD Thesis

Effect of bilirubin on expression and localization of Mrp1 and Pgp in the central nervous system

Silvia Gazzin

Anno Accademico 2005-2007 (XX ciclo)

I

SUPERVISOR

Prof. Claudio Tiribelli; Università degli Studi di Trieste, Dipartimento di Biochimica,

Biofisica e Chimica delle Macromolecole; Centro Studi Fegato (CSF), Area Science Park, Ss

14, Km 163.5, 34012-Basovizza (Ts), Italy. E-mail:ctliver@unts.it

EXTERNAL SUPERVISOR

Dr. Jean-Francois Ghersi-Egea; INSERM U842, Neurobiologie experimentale et

physiopatologie, faculte Laennec, rue Gluillaume Paradin, Lyon-France. E-mayl: ghersi-

egea@lyon.inserm.fr

TUTOR

Prof. Stefano Gustincich; Scuola Internazionale Superiore di Studi Avanzati di Trieste

(SISSA), settore di Neurobiologia, Area Science Park, Ss 14, Km 163.5, 34012-Basovizza

(Ts), Italy. E-mail: gustinci@sissa.it

THESIS COMMETEE MEMBERS:

Prof. Francesco Tedesco; Università degli Studi di Trieste, Dipartimento di Fisiologie e

Patologia.

Prof. Stefano Gustincich; Laboratory of Molecular Neurobiology, International School for

Advanced Studies S.I.S.S.A., Area Science Park, Ss 14, Km 163.5, 34012-Basovizza

(Ts)

Prof. Massimo Levriero; Università degli Studi di Roma “lLa Sapienza”, Dip. di Medicina

Interna, Regina elena Cancer institute, Via delle Messi d’Oro 156, 00158-ROMA

Prof. Franco Vittur; Università degli Studi di Trieste, Dipartimento di Biochimica, Biofisica

e Chimica delle Macromolecole.

Prof. Silvia Giordano; UniversitÀ degli Studi di Torino, Dip. Scienze Oncologiche, Strada

Provinciale 142 di Piovesi, 10060-Candiolo (To).

Dott. Claudio Brancolini; Università degli Studi di Udine, Dip. di Scenze e Tecnologie

Biomediche, Piazzale Kolbe 4, 33100-Udine.

II

Prof. Giannino Del Sal; Università degli Studi di Trieste, Dipartimento di Biochimica,

Biofisica e Chimica delle Macromolecole, LNCIB-AREA science park, Padriciano 99,

34012-Trieste

Prof. Renato Gennaro; Università degli Studi di Trieste, Dipartimento di Biochimica,

Biofisica e Chimica delle Macromolecole

Table of Contents

III

Abstract V

List of papers VII

List of Abbreviations VIII

Introduction:

1.1 The bilirubin metabolism 1

1.1b Total, direct and indirect bilirubin meseaure in blood 5

1.2 The Bf theory 6

1.3 Adult and neonatal bilirubin metabolism 9

1.3a neonatal jaundice 9

1.3b BIND (Bilirubin Induced Neurological Dysfunction) 10

1.3c Kernicterus 11

1.4 Molecular basis of bilirubin toxicity 13

1.5 The Gunn rat 17

1.6 The Blood Brain Interfaces (BBI) 24

1.6a The Blood Brain Barrier (BBB) 26

1.6b The Blood-CerebroSpinal Fuid Barrier (BCSFB) 32

1.6c The parenchyma 36

1.7 The UCB transporters 38

1.7a Mrp1 39

1.7b Pgp 40

Aim of the study 41

Materials and Methods 42

1 Animals 42

2 Human samples 42

3 Rat tissues (CPs, Cx, CLL) dissection 42

4 Microvessels isolement procedure 43

5 Sample homogenisation and protein dosage 45

6 Spettrophptometric γGT activity measure 46

7 Enzymohistochemestry detection of γGT activity 46

8 SDS-Page and Western blot 47

9 Western blot quantification procedure 48

10 Immunohystochemestry 50

Table of Contents

IV

11 Total-RNA extraction 51

12 Total RNA retro-transcription 51

13 Real Time PCR 51

14 Total Bilirubin in Serum (TBS)and Albumin measure 52

15 Calculated free bilirubin (cBf) 53

16 Statistical analysis 53

Results 54

I) The Sprague-Dawley rat: the physiologiacal model. 54

Ia) Mrp1 and Pgp expression on adult brain barriers 54

Ib) Mrp1 and Pgp expression on rat brain barriers during the post-

natal development 62

II) The Gunn rat: the bilirubin implications on Mrp1 and Pgp expression on

rat BBI. 72

Discussion 92

I) The Sprague-Dawley rat: the physiologiacal model. 92

Ia) Mrp1 and Pgp expression on adult brain barriers 92

Ib) Mrp1 and Pgp expression on rat brain barriers during the post-

natal development 94

II) The Gunn rat: the bilirubin implications on Mrp1 and Pgp expression on

rat BBI. 96

Conclusions 101

Acknowledgments 103

References 104

Reprint of the papers 121

.

ABSTRACT

V

ABSTRACT

The free bilirubin (Bf), the part of UnCnjugated Bilirubin (UCB) exceeding the

binding capacity of serum albumin, causes encephalopathy in severely jaundiced neonates by

crossing Blood Brain Interfaces (BBI) and damaging specific areas of the brain. Around 70%

of children with kernicterus die within seven days, while the 30% survivors usually suffer

irreversible sequels, including hearing loss, paralysis of upward gaze, mental retardation, and

cerebral palsy with athetosis. Bilirubin encephalopathy is actually the leading cause of

hospital readmission of newborns within the first month after birth.

The endothelia of Micro Vessels (MV) at the blood brain barrier (BBB), and the

epithelia of the choroid plexuses (CP) at the blood cerebrospinal fluid barrier (BCSFB),

should therefore be viewed as dynamic sites of exchange controlling both influx into, and

efflux out of the brain.

Among efflux transporters, two "ATP Binding Cassette" (ABC) transport proteins,

Mrp1 and Pgp have been shown to play important neuroprotective functions and appear to be

actively involved in keeping extra cellular bilirubin concentration below toxic levels by

limiting their entry from blood to brain, or else in controlling intracellular bilirubin levels in

parenchyma cells.

Although the role of these transporters is central in neuroprotection, their pattern of

expression and cellular localization in the central nervous system (CNS), and in the two blood

brain interfaces (BBI), remains still unsettled and no studies to the effect of

hyperbilirubinemia on the transporters at the brain interfaces have been made before.

We used the Srague-Dowley rat as physiological model, and the Gunn rat, as the well-

established model for the hyperbilirubinemia, to study the relative expression and localisation

of the transporters at the BBI.

By quantitative Western blot, we have discovered a mirroring expression of Mrp1 and

Pgp in BBB and BCSFB; with the BBB characterized by Pgp and the BCSFB by Mrp1.

In the adult rat, the relative protein expression of Mrp1 reaches the highest level in 4th

Ventricle CP (4thV CP), and in LV CP (60%) and Pgp in isolated MV. In MV and in Cortex

(Cx), the Mrp1 amount is about 4.3% and 5.2% of the 4thV CP, respectively, suggesting that

Mrp1 may be present in others than endothelial cells, such as in astrocytes. Mrp1 localizes at

ABSTRACT

VI

the basolateral side of the choroids epithelium and their expression does not change

significantly from birth to adult life in both barriers.

In the homozygous jj Gunn rats the total bilirubin in serum (TBS) is several time

higher than in a normal rat, due to a genetic mutation in the Uridin di phospho Glucoronosyl

Transferase 1A1 (UGT1A1) enzyme, preventing the conjugation and subsequent biliary

clearance of bilirubin.

In the jj hyperbilirubinemic Gunn rat a strong and early down regulation of Mrp1 was

found in the CPs, respect the control (Jj rat). In both CPs the Mrp1 amount at P2 is similar

between genotypes but in jj Gunn rats, the decrease is rapid and consistent reaching a value

around 50% of the control genotype in the 4thV CP at P9 (P≤0.001 at P9, P17 and P60). The

decrease is less marked in LV CP and is apparent from P17 (both P17 and P60, P≤0.05). The

down regulation is not attended by a re-localisation of the transporter witch always stays

located at the basolateral side of the epithelial cell forming the BCSFB.

In the physiological situation, the Pgp relative expression in both CPs and in Cx is

around 0.5% and 7% of MVs Pgp expression and all the brain Pgp is located at the apical side

of endothelial cells forming MV, as confirmed by immunofluorescence. The transporters

expression is weak a P9 and increase 4.6 fold with maturation.

The hyperbilirubinemia enhance the Pgp expression in the BBB of hyperbilirubinemic

rats at every post-natal age, but at P17 is 5 times less present than in adult.

Taken together, these results suggest that, while the Mrp1 related protection seems to

be fully functional during post natal development, and strongly impaired in the

hyperbilirubinemic Gunn rat, the Pgp mediated neuroprotection may not be as efficient during

early stages as in adult, and no significant adaptatives changes are present in the jj Gunn rats.

We suggest that the down regulation of Mrp1 leads to the less efficient bilirubin out-

flow from CNS, and that the up-regulation of Pgp in BBB may be not sufficient to prevent the

passage of bilirubin across the barrier.

In addition, the hypolasia of cerebellum, marked at P9 (25% of weight loss) and

reaching the 47% at P60, indicates that the hazardous period in bilirubin encephalopathy

occur in the first days of life.

List of papers

VII

“Differential expression of the multidrug-resistance associated proteins ABCb1 and

ABCc1 between blood-brain interfaces”

Silvia Gazzin1,2, Nathalie Strazielle3,1, Charlotte Schmitt1, Michelle Fèvre-Montange1, J.

Donald Ostrow4, Claudio Tiribelli2, Jean-François Ghersi-Egea1.

Journal of Comparative Neurobiology_in press.

List of abbreviations

VIII

BCSFB Blood CerebroSpinal Fluid Barrier

Bf free Bilirubin

BIND Bilirubin Induced Neuronal Disfunction CB Conjugated Bilirubin

CLL CerebeLLum

CN Crigler-Najjar

CNS central Nervous System

CP/CPs Choroid Plexuses CSF CerebroSpinal Fluid

Cx Cortex

HSA Human Serum Albumin

jj homozygous hyperbilirubinemic Gunn rats

Jj heterozygous normobilirubinemic Gunn rats KDa Kilo-Daltons

KR Krebbs-Ringher

LV CP Lateral ventricle Choroid Plexuses

Mdr Multi Drug Resistance

Mrp Multi drug Resistance Protein MV MicroVessels

MW Molecular weight

P2, 9, etc Post-natal age in days

PCR Polimerase Chain Reaction

Pgp P glyco-protein RBC red blood cells

T.test Student test

TBS Total Bilirubin in Serum

TJ Tight Junctions UCB UnConjugated Bilirubin

INTRODUCTION

1

INTRODUCTION

INTRODUCTION

2

1.1 Bilirubin metabolism

The bilirubin is the yellow pigment formed during the catabolism of heme containing

compounds, responsible for discoloration of skin, sclerae, and mucous membranes in jaundice

(Ostrow, 1987).

The major part of haemoglobin degradation products derive from the haemolysis of

senescent red blood cells (80%), but significant fractions are also derived from others heme

containing enzymes such as cytochromes, catalases, peroxidase, tryptophane pyrrolase and

muscle myoglobin (15-20%), and a minor part from the destruction of immature red blood

cells in the bone marrow, process termed “ineffective erythropoiesis” (less than 3%) (Ostrow,

1987; Gourley, 1997).

In Fig 1 are summarized the principal steps in heme degradation.

The first step in heme degradation is the oxidative excision of the alpha-carbon, with

the opening of the terapyrrole ring of protoporhyrin IX by microsomal heme oxygenase

enzyme, located primarily in the reticuloendothelial tissue and, to a lesser degree, in tissue

macrophages and intestinal epithelium; to yield equimolar amounts of Fe3+, CO, and

biliverdin (Ostrow, 1987; Gourley, 1997; Rubaltelli, 1993).

Then, the cytosolic enzyme biliverdin reductase reduces the central carbon (C10) of

the linear biliverdin IXa forming the bilirubin IXa (UnConjugared Bilirubin, UCB). The

resulting bilirubin is formed by two planar dipyrrolic halves connected by an -HCH-bridge

(Fig 2), folded at a 90° angle in a “ridge tile” or folded book structure. A trio of hydrogen

bonds ties each -COOH group on each half with the two NH groups and the lactam oxygen of

the opposite dipyrrolic half. In this rigid structure, all the polar groups (propionic acid side

chains) limit interaction with water and make the molecule very nonpolar and lipophilic. The

diacidic unconjugated molecule (H2B) is therefore virtually insoluble in water (minor 70 nM)

and has remarkably high pKa values (8.1 and 8.4). At the physiological plasma pH of 7.4,

approximately the 82% of UCB is in diacidic form (Ostrow et al., 1994). This lipophilicity

renders necessary the presence of a carrier molecule to transport the bilirubin in the aqueous

environment of the blood (Gourley, 1997; Ostrow, 1987).

INTRODUCTION

3

LIVER 1-2 UDP-glucoronic

acid

1-2 UDP

UDP-GlucoronosylTransferase

BLOAlbumin + UCB ↔ Albumin-UCB

(UCB)

CO

O O

CO

O

COO

OO

HO

OO

O O

CO

O

NAD(P)

NAD(P)H

Biliverdin Reductase

3 NADPH ½ NADPH 3 O2

3 NADP+ ½ NAPP+ 3 H2O

Heme-Oxygenase SPLEEN

(UCB)

INTESTINE

COHO

H O

COHO

H O

BLOOD

Fig 1: Bilirubin Metabolism In spleen, the haemoglobin is converted to biliverdin by the heme-oxigenase enzyme, then in bilirubin (UCB) by

the biliverdin reductase enzymatic activity. Bounded to albumin in blood, the UCB reach the liver where the

UGT1A1 bilirubin specific enzyme activity, conjugates it with one or two glucoronic acid. The glucoronidation

rends the Conjugated Bilirubin (CB) more soluble thus eliminable by the faeces in the intestine.

INTRODUCTION

4

This function is accomplished by albumin that posses two sites for the binding of

bilirubin. The affinity constant for the first site rang from the historical Ka of 6.8x107 M-1

(Brodersen, 1980) to the revised 2.3x106 M-1 (Weisiger et al., 2001; Roca et al., 2006). The

second binding site shows a lesser affinity (about 5x105 M-1) (Jacobsen, 1969).

The bilirubin tightly bounded to serum albumin, that in plasma is the 99.9% of the

total bilirubin, display a very limitate renal excretion and diffusion into tissues (Gourley,

1997; Ostrow, 1987; Ahlfors, 2001; Weisiger et al., 2001; Ostrow et al., 2004a).

The bilirubin-albumin complex reaches the liver where the bilirubin alone is taken up

into the aqueous environment of the hepatocyte across the basolateral membrane by facilitated

diffusion processes, whose mediating transporters have not be identified, and again bound to a

carrier protein, the glutathione S-transferase (GST) or ligandin or Y protein (Jagt et al., 1982),

preventing the refluxing back of bilirubin and its conjugates in the blood and promotes

transfer of UCB to the smooth endoplasmic reticulum for conjugation.

In the endoplasmic reticulum, the specific bilirubin glucuronosyltransferase

(UGT1A1) catalyses the conjugation of bilirubin and the glucoronic acid. The ionisation of

each –COOH group removes one hydrogen bond, allowing interaction of the -COO- group

with water. The conjugation of each –COOH group with polar glucuronic acid groups breaks

one H-bond and renders the conjugate water-soluble. This biotransformation is essential for

the efficient biliary secretion of bilirubin and only traces of UCB (2%) appears in the bile

(Ostrow, 1987). The conjugated bilirubin is excreted thought the hepatocyte canalicular

membrane multispecific organic anion transporter MRP2/Mrp2 in the bile canaliculi and then

they enter in the intestinal lumen (Ostrow, 1987; Gourley, 1997).

In intestine, various carbon double bonds in bilirubin are hydrogenate and

subsequently oxidized by intestinal flora, producing the urobilinoids that are mostly

eliminated in the faeces. Alternatively the bilirubin conjugates can also be hydrolyzated back

by bacterial or endogenous tissue β-glucoronidase. The UCB produced is rapidly adsorbed

from the intestine, and via the blood, reach the liver. The reabsorbtion in the intestine is very

limited, so that in the plasma, over 96% of bilirubin is unconjugated.

INTRODUCTION

5

1.1b Total, direct and indirect bilirubin measure in blood

The bilirubin species (UCB, CB) content in the blood could be measured by the diazo

method. Basically the internal hydrogen bonding of UCB rends the molecule inaccessible to

the diazo reagents (diazotized aromatic amines), able to split bilirubin into two dipyrroles,

each coupled with the reagent to form red-purple compounds (Ostrow, 1987).

In the CB, the hydrogen bonds are esterificated with highly hydrophilic glucuronic acid and

opened, allowing the methylene bridge to react rapidly (“directly”) with diazo reagents (direct

bilirubin). When molecules able to break the hydrogen bonding are added (e.g. methanol,

ethanol, dimethyl sulfoxide, caffeine, urea), also the UCB could react giving a measure of the

total serum bilirubin (TBS). The UCB concentration (indirect reacting) is derived by the

difference between the TBS and the direct reacting bilirubin (Doumas et al., 1985b; Doumas

et al., 1985a).

INTRODUCTION

6

1.2 The Bf theory

The total amount of bilirubin present in the blood is formed by the UnConjugated

Bilirubin bonded to albumin (UCB-A), the conjugated bilirubin also bound to albumin (CB-

A), although with much lower affinity than UCB, and the unbounded-UCB, the free bilirubin

(Bf), that is less than the 0,1% in physiological condition (Ostrow et al., 1994; Gourley,

1997).

The proportion of free UCB depends from the presence of albumin and its binding

affinity constants, actually considered of 3.1 and 2.3x106 M-1, in adults and full-term

newborns respectively, reflecting the change in binding affinity when the concentration of the

albumin moves from about the 450 μM in newborn to the 600 μM in adults (Brito, 2006). If

the amount of bilirubin does not exceed the albumin binding capacity (B/A ratio below the

unit) the maximal amount of UCB that could be transported bonded to albumin, is 435 μM

(25 mg/dL) (Brito et al., 2001) and the complex remains in the blood vessels (Diamond and

Schmid, 1966). At UCB to HSA (Human Serum Albumin) molar ratio exceeding the unity,

the amount of Bf rises (Brodersen, 1981), the TBS decreases and the bilirubin amount in

tissue increases (Diamond and Schmid, 1966).

Due to its remarkably high pKa values (8.1 and 8.4), the circulating Bf at the

physiological plasma pH of 7.4 is mainly formed (84%) by the non ionised, protonated diadic

molecule (Fig 2), able to cross cell membranes by passive diffusion and determine the cellular

content of UCB (Ostrow et al., 1994; Zucker et al., 1999). This is showed by the enhanced

extent of bilirubin toxic effects in red blood cells (RBC) and increased tissue content of

bilirubin (Brito et al., 2001), in acidosis, when the UCB diacidic molecules amount is

enhanced in the blood.

This idea has been postulate for the first time in 89 by Odell (Odell, 1959), and

sucefully popularized as the “free bilirubin theory” (Wennberg, 2000). In agreement with the

“free bilirubin theory”, displacing drugs such as sulphonamides, butolone or ceftriaxone

increase the passage of bilirubin from the blood to the tissues (Diamond and Schmid, 1966;

Hanko et al., 2003; Aono et al., 1989) and are historically correlated to an increase of the

Kernicterus incidence when administrates to newborn babies (Gourley, 1997; Diamond and

Schmid, 1966).

INTRODUCTION

7

By contrast the administration of albumin, also at high TBS, limit the entry of the

pigment in the CNS (Diamond and Schmid, 1966) and, as partially, reverses the bilirubin

triggered toxic effects on red blood cells and mitochondria, probably removing the UCB

diacidic molecules attached to the outer leaflet of the cell membrane (Brito et al., 2001;

Mustafa and King, 1970).

In vivo, in the AJR (un Albuminemic Jaundiced Rat) rat, obtained breading the

hyperbilirubinemic Gunn rat with an un-albuminemic rat, the total amount of bilirubin in

Albumin-UnConjugated Bilirubin

Albumin + UCB (Bf)

pH

Kaff

Fig 2: UnConjugated species of bilirubin in blood Proportion of chemical species of the unconjugated bilirubin at different pH values derived from partition syudiesas described in (Ostrow et al., 1994), and structure of diacidic, monoanion and dianion bilirubin species representingabout the 80%, 16%, and less than 2%, respectively, of the UCB in solution at the blood physiological pH of 7.4.

INTRODUCTION

8

brain is 1.2-2.7 time grater than in the control, and the animals die in 3 weeks, in respect to a

TBS content of 25% of a Gunn rat (Takahashi et al., 1984).

The TBS is actually considered a poor predictor of Kernicterus and the Bf, the part of

bilirubin that really enter the brain, is now supposed to be the critical determinant in cellular

uptake of bilirubin (Calligaris et al., 2007; Wennberg, 2000; Wennberg et al., 2006; Ostrow et

al., 2003b; Ostrow et al., 2003a).

Unfortunately, at date, no routine methods for measuring Bf in serum are available.

INTRODUCTION

9

1.3 Adult and neonatal bilirubin metabolism: neonatal jaundice, BIND and kernicterus

During the foetal development, the foetus is protected against bilirubin by the maternal

placenta. After the birth, the bilirubin clearance is totally accomplished by the infants.

In adult, the production of bilirubin is about of 3-4 mg/Kg die, and in full term newborns the

value reaches the 6-8 mg/Kg (Gourley, 1997). This difference is principally due to:

1) The minor red blood cell viability (Rubaltelli, 1993) (70-90 days vs. 120 days in adults)

(Gourley, 1997; Ostrow, 1987).

2) The immaturity of the hepatic bilirubin conjugation (Rubaltelli, 1993). The diglucuronide

normally accounts for 80-85% of the bilirubin conjugates in the bile of adult humans,

whereas the monoconjugate, easily re-converted in UCB, predominates in newborns.

3) The absence of anaerobic intestinal flora in infants up to 2 months of age, that causes a

decreased intestinal UCB degradation (Gourley, 1997; Ostrow, 1987).

4) The consequent enhancement to the UCB reabsorption and enterohepatic recycling,

causing higher serum levels of UCB (Vitek et al., 2000; Gourley, 1997; Ostrow, 1987).

5) The less efficacy of neonatal serum in binding unconjugated bilirubin due to the lesser

amount of albumin (Weisiger et al., 2001; Roca et al., 2006)

6) The highest level of expression of hepatic heme oxygenase early after birth, leading to an

increase of total serum bilirubin (Maroti et al., 2007; Drummond and Kappas, 1984).

1.3a Neonatal jaundice

The Immaturity of most steps of bilirubin metabolism causes a mild, temporary

retention of UCB in approximately 60% of healthy term neonates (Ostrow, 1987);

characterized by a serum UCB levels of less than 170 μM (10 mg/dL) (Ostrow et al., 1994)

within the first days of life. This situation usually resolves spontaneously in the first week of

life without sequelae (Ostrow et al., 2003a; Ostrow and Tiribelli, 2003).

The existence of hyperbilirubinemia is clinically evidenced by the deposition of

bilirubin and subsequent discoloration of tissues, phenomena named icterus (greek: ikteros) or

“jaundice” (French: “jaune”=yellow) (Gourley, 1997). Low, nanomolar concentrations of

UCB are not only not dangerous, but probably benefical by providing protection from

INTRODUCTION

10

oxidative injuries, such as ischemia (Dore and Snyder, 1999), the risk of coronary athery

diseases (Mayer, 2000) and cancer (Keshavan et al., 2004).

The antioxidant ability of UCB arises from the redox consuming cycling mechanism

that acts between the conversions of UCB in biliverdin. During this step the oxidant species

are consumed and the bilirubin regenerated via biliverdin reductase. The pigments may play a

role similar to the glutathione cycle in cytoplasm, acting against the lipophilic reactive oxygen

species produced from the cellular membranes, wile the GSH-GSSG cycle plays against the

cytosolic oxidative species (Tomaro and Batlle, 2002; Sedlak and Snyder, 2004).

Remarkably, bilirubin concentrations as little as 10 nM are able to protect cell cultures

from 10000 times higher concentrations of H2O2 (Baranano et al., 2002).

Similarly, in Gilbert's syndrome, a very common chronic, mild, fluctuating unconjugated

hyperbilirubinemia, due to a recessive insertional mutation in TATAA element upstream the

UGT1A1 gene resulting in an reduced level of expression and enzymatic activity (Kadakol et

al., 2000), the hyperbilirubinemia may protect against oxidant stress (Ostrow, 1987). Rigato

in a retrospective study reported that low-mild hyperbilirubinemia is negatively related to the

risk of different diseases, such as atherosclerotic disease, cancer, demyelinizatind

neuropathies (Rigato et al., 2005). Additionally the in vivo neointimal hyperplasia induced by

balloon-injury seems to be prevented by hyperbilirubinemia in Gunn rats (jj), due to the

antiproliferative properties of the bilirubin pigment that is able to arrest the cell cycle and

inhibits the p38 MAP Kinases (Ollinger et al., 2005).

1.3b BIND (Bilirubin Induced Neurological Dysfunction)

Moderate hyperbilirubinemia, (200-300 μM; 11.7 mg/dL) (Ostrow, 1987; Ostrow and

Tiribelli, 2003) occurs in at least 16% of infants. The increased entry of UCB into the CNS

may cause transitory effects such as hypotonia, lethargy, anorexia, poor suckling and

abnormal brainstem evoked potentials (BSAEP), symptoms referred to the deposition of

unbound UCB diacid in the central nervous system and development of bilirubin

encephalopathy or BIND (Bilirubin Induced Neuronal Dysfunction), which is usually

reversible (Ostrow, 1987).

In these cases, the preferred treatment of neonatal jaundice is phototherapy, which

converts UCB to photo isomers that can be excreted in bile and urine without conjugation or

INTRODUCTION

11

parenteral administration of tin mesoporphyrin IX, a potent competitive inhibitor of heme

oxygenase and thus of bilirubin synthesis. When severely jaundiced neonates respond

insufficiently to the above therapies, they are treated by exchange transfusion to physically

remove UCB from the circulation (Ostrow, 1987).

1.3c kernicterus

In about the 2% of infants (Gourley, 1997) at slightly higher or prolonged serum UCB

levels (13-26 mg/dL; 220-440 μM) a severe hyperbilirubinemia occurs (Soorani-Lunsing et

al., 2001) and results in permanent neurological sequelae ranging from delay in motor

development, impaired cognitive functions, auditory dysfunctions to more severe

extrapiramidal motor, auditory and cognitive disorders, termed Kernicterus (Hansen, 1994b),

or even death.

Similarly, in the Crigler-Najjar Type I syndrome (CN I), a severe, rare, hereditary,

recessive deficiencies, the non haemolytic hyper bilirubinemia is characterized by high levels

of TBS, that appear in the first days of life (mean 16±5 mg/dL; 270 μM) (CRIGLER, Jr. and

NAJJAR, 1952; Servedio et al., 2005) and increase with age by approximately 0.8 mg/dL per

year (Strauss et al., 2006). The UCB binds on red blood cells (RBC) surface and tissue

phospholipids and ultimately leads to accumulation in focal brain regions causing neural

dysfunctions followed by cell death and permanent disability if the patients are untreated. CN

patients remain vulnerable to brain injury throughout their life span (Chalasani et al., 1997).

The goals in treat CN patients are to maintain the B/A ratio at ≤0.5 in neonates and

≤0.7 in adults (Strauss et al., 2006) by prolonged and massive phototherapy, but the only way

to cure Crigler-Najjar I subjects is a liver transplantation (Ostrow, 1987). The genetic lesions

could be located in any of the five UGT1A1 gene exons, the only isoform involved in the

UCB conjugation in humans (Strauss et al., 2006; Rubaltelli, 1993) and cause premature stop

codon, that alter or delete amino acid residues. No detectable activity of UGT1A1 is presents.

In the Crigler-Najjar type 2 syndromes, the mutation frequently consists in a single

amino acid substitution with an autosomal dominant pattern of inheritance with variable

penetrance and responds to stimulatory effect played by phenobarbital. The UGT1A1 activity

is reduced but not abolished (Kadakol et al., 2000; Ostrow, 1987). Jaundice begins in late

childhood, is less severe, and seldom causes brain damage (Rubaltelli, 1993).

INTRODUCTION

12

The neurological signs of bilirubin encephalopathy and the following kermicterus are

selective, preferentially affecting only certain areas of the brain and only certain cells within

these regions. The classical pattern of discoloration/damage is symmetric, highly selective

involves the basal ganglia, the pallidum, the subtalamic nucleus, the Horn of Ammon and the

cerebellum (Hansen, 1994a; Rodriguez Garay and Scremin, 1971; Diamond and Schmid,

1966).

INTRODUCTION

13

1.4 Molecular basis of bilirubin toxicity

In spite of decades of scientific investigation our understanding of the biology of the

bilirubin encephalopathy dysfunction (BIND) is still incomplete (Ostrow et al., 2004b). This

is due to the poor comparability of data obtained in different laboratory, under different

condition in bilirubin/albumin ratio, B-A affinity constant applied in calculate the Bf, and

medium composition (Ostrow et al., 2003b). This is even more complicate by the lack of a

routine method to measure directly the Bf in blood. Additionally, several studies concerning

the UCB toxicology have been performed at higher UCB than those seen in jaundiced

neonates, rendering uncertain the relevance of the experimental finding, especially when dose

dependent reasoning of UCB effects have been made (Ostrow et al., 2002).

The first studies concerning the bilirubin toxicity indicated in the membrane the

principal target. Myelin shows the highest concentration of bilirubin (Brodersen, 1981;

Mustafa and King, 1970), possibly explaining the strong trophysm for brain (Nagaoka and

Cowger, 1978). When cultured red blood cells (RGB) have been used as model for cellular

membranes, severe B/A ratio dose dependent morphological changes, accelerated aging,

release of membrane phospholipids, appearance of haemoglobin depleted vesicles and cell

lysis were observed after UCB exposure.

The bilirubin interaction with the outer leaflet of the bilayer induces alterations in the

content of several classes of phospholipids, leads to lipid package and polarity, and renders

the membrane more fluid and more permeable to water (Brito et al., 2001) at B/A ratio of 0.4

and UCB concentration of 185 μM in vivo (Brito et al., 1996; Brito et al., 2002).

In 30 days old Sprague-Dawley rat, the disturbance of the brain micro vessels

membrane inhibits the glucose uptake, when the B/A ratio exceed the unit (Katoh-Semba and

Kashiwamata, 1980). The decrease in metabolic rates of glucose after a bilirubin infusion on

rats of 10 days stage, localizes with the auditory pathway, regions knows to preferentially

accumulate bilirubin in kernicteus (Roger et al., 1995). Choen, by contrast, reports an

increased glucose uptake and GLUT1 (Glucose receptor type 1) over expression on bovine

aortic cultured endothelial cells, and suggest a role for the oxidative stress (Cohen et al.,

2006).

The apoptosis seems to be the preferred way in bilirubin induced cell death.

INTRODUCTION

14

Observations in human kernicteric neonates indicate in the apoptotic vs. the necrosis with a

lack of inflammation the preferred way in bilirubin induce cell death (Ahdab-Barmada, 2000).

The activation of both mithocondrial and extrinsic pathway has been reported.

In mitochondria isolated from rat brain, the UCB induced permeabilysation is involved in

cytochrome C release (Rodrigues et al., 2000), increasing of the Bax levels and translocation

in the nucleus, degradation of PARP and activation of caspase 3, focusing on the

mitochondrial pathway in nerve cell apoptosis (Rodrigues et al., 2002a; Grojean et al.,

2000b). Additionally, bilirubin may activate the TNFR1 (tumour necrosis factor receptor) and

MAPKs (mitogen-activacted protein kinases), in the extrinsic apoptotic pathway (Fernandes

et al., 2006; Lin et al., 2003).

In vitro, cultured oligodendrocytes (Genc et al., 2003) and endothelial cells (Akin et

al., 2002) display both necrosis and apoptosis (Seubert et al., 2002). These finding, are in

contrast with the immunostimulant properties of UCB on cultured astrocytes and mycroglia,

where the IL1β production and TNFα secretion, together with glutamate influx and cell death

both by necrosis and apoptosis (Gordo et al., 2006), are induced by bilirubin. It has been

suggested that the clinical observation of a UCB encephalopathy aggravation during sepsis

may be corroborated by the in vitro results.

Inside the cell, bilirubin may damage the respiratory chain uncoupling the oxidative

phosphorylation, generates ROS witch leads to protein oxidation and lipid peroxidaion,

enhancing the oxidation of GSH to GSSG and reducing the restore of GSSG. Additionally

bilirubin leads to intracellular calcium overload and destruction of glutathione metabolism

(Hansen, 1994c), pointing out on oxidative stress as a key player in UCB neuronal induced

damage.

Bilirubin is able to inhibits different kinases (cAMP, cGMP, Ca-calmodulin and Ca

phospholipids dependent protein kinase A (Hansen et al., 1996) and probably binding to the

lysine residues on the ATP binding domain, as suggested by the anty-inhibitory effect of

polilysine but not poliglutamate or poliargynine (Hansen et al., 1997). In agreement with this

theory, the inhibition of the synapsin I responsible, in the phosphorylated form, of the

releasing of neurotransmitters, explains the inhibitory effect of bilirubin on auditory evoked

responses (Hansen, 1994c).

A role of glutamate in neuronal cell death has been postulated. In astrocytes culture

the uptake of glutamate is reduced and its release induced (Fernandes et al., 2004). In neurons

INTRODUCTION

15

the cell death seems to be related to the NMDA receptors, along the inhibition of PKC

activity (Grojean et al., 2000a). In vivo, administration of glutamate increases and the

glutamate receptor antagonist treatment reduce the entity of UCB induced injuries on Gunn

rat pups (McDonald et al., 1998). The release of glutamate is not modified by LPS, suggesting

the co-existence of different mechanisms in UCB nerve cell interaction, independents on the

membrane perturbation (Fernandes et al., 2004).

Additionally, bilirubin impairs DNA and protein synthesis (Yamada et al., 1977;

Schiff et al., 1985) and traslocates Mrp1 from the perinuclear region to the plasma membrane

and, simultaneously, up regulate both mRNA and protein expression on cultured primary

mice astrocytes (Gennuso et al., 2004).

In vivo, the selectivity for certain areas of the brain and, within these areas, for certain

cell types, is a landmark of kernicterus. The classical pattern of discoloration/damage in the

CNS is highly symmetric and highly selective, as revealed by brain slices of experimental

animals, autoptical samples, the use of BAESP as sensitive indicator of BIND (Shapiro, 2002;

Ahlfors and Shapiro, 2001) and finally, the in vivo magnetic resonance (Roger et al., 1996).

The brain regions most involved are the basal ganglia, the globus pallidus, the subtalamic

nucleus, the Horn of Ammon, the roof nuclei of the 4th ventricle, the cranial nerve nuclei of

the tegmentum, particularly the oculomotor and dentate nuclei, and the cerebellum (Hansen,

1994c; Rodriguez Garay and Scremin, 1971; Gourley, 1997; Diamond and Schmid, 1966;

Conlee and Shapiro, 1997). As consequence, the affected areas display a decrease in neurons

density and gliosys (Gourley, 1997).

The reasons for this specificity are unknown, a cellular specific sensibility and/or the

level of maturation of the cells, are the two most debated hypotheses for explain the bilirubin

trophysm in the brain.

A gradient in sensibility toward bilirubin between different cell types has been

reported in vitro. The neuronal cell lines (Ngai et al., 2000; Nagaoka and Cowger, 1978; Silva

et al., 2002; Rodrigues et al., 2002c) show the highest sensibility toward bilirubin, then the

hepatocytes and, at least, the fibroblast. The addition of albumin is able to reduce the entity of

damage but the patter of sensibility is not modified (Ngai et al., 2000; Nagaoka and Cowger,

1978). In brain mitochondria the respiration is inhibited by both, low and high bilirubin

concentrations. By contrast these concentration increase and inhibit, respectively the

INTRODUCTION

16

respiration in liver and hearth mitochondria (Mustafa and King, 1970). In vivo, the Purkinje

and the granular cells are more intensely stained than other cells (Hansen, 1994c).

Another consideration obtained from in vitro investigations, claims at the concept of

maturity of the cells in determining their sensibility toward bilirubin with the younger more

sensible than the adult ones. The rat glial or neuronal cells exhibits a “age in culture”

dependent sensitivity toward bilirubin (Falcao et al., 2006; Rodrigues et al., 2002c) as showed

by the decrease in mitochondrial functions (Amit and Brenner, 1993). And in neonatal RBC

the effects of UCB are exacerbated (Brito, 2006). Surprising, mitochondria swelling and

cytocrome C efflux, are most present in mitochondria isolated from adult rather than younger

rats (Rodrigues et al., 2002c).

INTRODUCTION

17

1.5 The Gunn rat

The in vivo knowledge on kernicterus derives almost totally from the investigation on

Gunn rat (Fig 3) that is a natural model for bilirubin encephalopathy. In this animal model the

genetic lesion closely parallel those in Crigler-Najjar syndrome type I (Chowdhury et al.,

1993) and the neuropathological lesions are similar to those in humans, with cell loss and

gliosis most prominent in the auditory nuclei of the brainstem, the cerebellum, the

hippocampus and the basal ganglia (Chowdhury et al., 1993).

The Gunn rat (Gunn, 1938) is a mutant strain of Wistar rats that lack the uridin di

phospho glucoronosyl transferase (UDPGT) activity toward bilirubin. In the homozygous jj

animal, a life long hyperbilirubinemia in absence of heamolysis is present (Chowdhury et al.,

1991; Iyanagi et al., 1989). In the non jaundiced heterozygous Jj rats the activity of UDPGT is

reduced and did not result in retention of bilirubin in the plasma (Schmid et al., 1958).

Fig 3: The Gunn rat. In this picture of Gunn rats at 17-day post-natal age (P17), the yellowdiscoloration of the skin is clearly visible in the jj hyper-bilirubinemic animals,while the heterozygous (Jj) animals appear white.

INTRODUCTION

18

In rats, two family of UDPGT are present, the UGT1 and UGT2 families and The

capacity of a tissue to glucoronydate and excrete a substrate depends on the UGT isoforms

presence and level of expression (Shelby et al., 2003). The UGT2 family members are

encoded from individual genes, each of them containing 6 exons and are responsible for the

glucoronydation of steroids. Based on the sequence similarity, the UGT2 family is divided in

two sub family (Emi et al., 1995): the UGT2A, specific for the olfactory bulb (2A1), and the

UGT 2B, formed by 6 members (2B1-2B6). The presence of a second family of UGT

enzymes, explains why in Gunn rats the activity toward bilirubin is undetectable but activity

toward several other substrates is normal (Chowdhury et al., 1991).

The UGT1 locus spans plus than 120 Kb and forms a gene complex (Fig 4).

Nine unique first exons encoding NH terminus of each isoform were located at the 5’. This

region is followed by 4 common exons named II, III, IV and V, encoding the C-terminus of

the protein. Based on their sequence, the first unique exons are divided in two groups: the

Bilirubin cluster (B cluster) and the Phenol cluster (A cluster). The A cluster consists of four

(A1-A4) isoform-specific exons, while the B cluster of 5 (B1-B5); A4 and B4 are pseudo

genes. The major bilirubin glucoronydase activity is found in the B2-cluster (Emi et al., 1995;

elAwady et al., 1990). All genes are arranged in the same transcriptional orientation and

posses a distinct promoter the multiple first unique exons and a slicing site at the 3’. From the

UGT1 complex, by alternative utilisation of multiple first exons in combination with the

commonly used exons, different form of enzyme could be expressed in tissue-specific or drug

responsive manner (Emi et al., 1995).

In the Gunn rat, the deletion of a single base pairs in the UGT1, a GCT coding for the

leucine 413, produce a stop codon that remove the 115C-terminal amino acids of the protein

(approximately 13 KDa), responsible for the binding on the endoplasmic reticulum (ER), and

generate a truncated form of the enzyme. The protein is unstable and rapidly degraded

(elAwady et al., 1990; Chowdhury et al., 1991; Chowdhury et al., 1993; Iyanagi et al., 1989).

In Gunn rat the mutation in the 4th common exons causes the simultaneous deficiencies of all

UGT1 isoforms (Emi et al., 1995).

INTRODUCTION

19

The section above resumes the signs and symptoms of kernicterus obtained from

experiments on Gunn rats or others strains of rats traded with displacing agents to enhance the

bilirubinemia and mimic the kernicterus. The below section is organized in order to follow the

timing in witch the effects of hyperbilirubunemia appears or have been described during the

post-natal development of the animals from the birth indicated in Hs: hours, or in P: post-natal

age in days.

First unique exons

285 530

COOH NH2 T

NH2

1

CCT GGA AAT GAC TGC CG…… Normal UGT1A1 CCT GAA ATG ACT GCC G………Gunn Rat

STOP

Common exons II III IV V

Fig 4: Schematic representation of the UGT1 gene organization in rat, and of the mutation presents in Gunnrats. The UGT1 gene is composed by 9 first unique exons, coding for the N-ter of the enzyme. Each unique first exonhas a promoter, responsible of the different expression in different tissues. The first exon is spliced to the 4common exons giving rise to form an isoform specific RNA, thus translated in a 55 KDa full-length protein. InGunn rat a guanosine on the 4th common exon is lost causing a frame shift. The resulting stop codon leads to a C-terminal truncated - unstable enzyme synthesis.

INTRODUCTION

20

4-8Hs

36Hs

P3

P1 - P5

P6

P6 - P10

P7

P8

The yellow discoloration of brain is diffuse, then,

The basal ganglia, cerebellar cortex, cerebellar nuclei, the eight nerve nuclei and

the grey matter of the spinal cord were more stained (rat plus bilirubin displacer)

(Schutta and Johnson, 1969).

The earliest bilirubin abnormalities in the CNS of Gunn rats appear represented by

neuronal intra-cytoplasmc membranous inclusions and mitochondrial changes (rat

plus bilirubin displacer) (Schutta and Johnson, 1969). Some abnormalities in the

Purkinje cell of the posterior lobules of the cerebellum (IX) of treated animals are

present, but any weight difference between Jj and jj rats, treated or not, is

observed (Jj and jj rat plus displacing agents. Animals scarified 24 H after

treatment) (Conlee and Shapiro, 1997). If the animals are irradiated with a single

dose (for 24H) of blue light at this age, the hypoplasia of cerebellum still develops

(Keino et al., 1985).

The blue light treatment during this period is able to partially protect from the

hypoplasia the cerebellum development. But only the lobules I-V and VIII are

completely normal (new Gunn rat strain on Sprague-Dawley background) (Keino

et al., 1985).

The tymidin-kynases activity is reduced (50% respect the Jj) and the activity is

not restored by the photoinactivaction of UCB (Yamada et al., 1977).

The blue light treatment during this period is able to completely protect from the

bilirubin induced hypolasia the cerebellum development (new Gunn rat strain on

Sprague-Dawley background) (Keino et al., 1985).

The cerebellar hypoplasia in the AJR animals begins (new Gunn-un-albuminemic

strain (AJR), obtained breading the Gunn rats with the un-albuminemic animals.

The TBS corresponds to the 25% than in jj Gunn animals, with the total absence

of albumin) (Takahashi et al., 1984). Conversely any difference in the cerebellum

weight between Jj and jj rats is reported in Gunn rats.

Until now the l’85% of Purkinje cells are normal, but in the posterior lobules of

the cerebellum (IX) of treated animals the Purkinje cells show the presence of

vacuoles (Jj and jj rat plus displacing agents. Animals scarified 24 H after

treatment) (Conlee and Shapiro, 1997).

In animals younger than 8 days of life, the hypotonia is the first sign of

INTRODUCTION

21

P10

P4-P11

(Esp. P7)

P12

P11-13

P1-15

P15-17

hyperbilirubinemic brain damage (rat plus bilirubin displacer) (Schutta and

Johnson, 1969).

The specific inhibition of the tymidin kynase in the cerebellum (Takahashi et al.,

1984) and the DNA synthesis (H3-timidin incorporation) is decreased specifically

in the Purkinje cells of the cerebellum in jj Gunn rats, but no similar damage is

visible in liver or in different regions of the brain (Yamada et al., 1977). The

expression of the GSH, responsible for the binding of UCB within the cell, is

comparable between Jj and jj rats. No changes are detected in cerebral cortex,

thalamus, and hypothalamus and in the midbrain (Gunn rats) (Johnson et al.,

1993). Until now the cerebellum volume growth and stops at P20. The cerebellar

lobules more damaged are the IX and the X less (new Gunn rat strain on Sprague-

Dawley background) (Keino et al., 1985). Ay this post-natal age, local UBC

infusions induce a decrease in metabolic rates of glucose that localize with the

regions knows to preferentially accumulate bilirubin in kernicteus (Roger et al.,

1995).

A single dose of photo irradiation (24H) during the post-natal period ranging from

P4 to P11, and especially the P7 day, is the more effective in preventing

cerebellum hypoplasia (new Gunn rat strain on Sprague-Dawley background)

(Keino et al., 1985).

The irradiation with a single dose for 24H is not able to stop the hypoplasia of

cerebellum (new Gunn rat strain on Sprague-Dawley background) (Keino et al.,

1985). The Purkinje cells appears suffering and start to death (Takagishi and

Yamamura, 1989) .

The percentual of abnormal Purkinje cells in the cerebellum is 54-71%. The main

visible symptom is the ataxia that normally is transient and confined to the firs 3-4

weeks of life (rats plus bilirubin displacer) (Schutta and Johnson, 1969).

The blue light treatment is able to protect the cerebellum development from the

hypoplasia (new Gunn rat strain on Sprague-Dawley background) (Keino et al.,

1985).

In male sex all brain regions show a major content of bilirubin that in females

corrispective, after bilirubin infusion. The content is highest in the cerebellum (19

μg/g tissue), intermediate in the brainstem (11) and lowest in Cx (4.7) (Gunn rats

INTRODUCTION

22

P16-21

P17

P18 – 23

P21

P22

P23

P30

infused by H3-bilirybin and treated or not with displacing agents) (Cannon et al.,

2006).

The intensity of the yellow staining of the brain diminishes until the

disappearance (rat plus bilirubin displacer) (Schutta and Johnson, 1969).

In saline treated jj rats, the Purkinje cells of the II and VI lobules of the

cerebellum appear suffering. In jj rats, treated or not, the cerebellum growth is

quite totally abolished. Between the same genotype, treated rats, shown an

additional, but not significant, cerebellum weight loss (Jj and jj rat plus displacing

agents. Animals scarified 24 H after treatment) (Conlee and Shapiro, 1997). The

abnormalities in the uditive functions are evident and the Purkinje cells in the

cerebellum, the hippocampus and the basal ganglia are the most affected in the

brain, showing large mitochondria, inhibition of the phosphorilation (AJR rats)

(Takahashi et al., 1984). In a strain of Gunn rats, the BSEP abnormalities are

displayed only after administration of sulphonamides to animals (Shapiro, 2002).

The Purkinje cells presents in the cerebellum appears normal but the Purkinje

cells layer counts 3-5 cells (Takagishi and Yamamura, 1989).

The expression of the GSH, responsible for the binding of UCB within the cell, is

slightly enhanced in the Jj rats but high in the flocculus and in the lateral

hemispheres of the cerebellum of the jj rats. No changes are detected in cerebral

cortex, thalamus, and hypothalamus and in the midbrain (Johnson et al., 1993).

The bilirubin content in the cerebellum is 1,2-2,7 times higher and the animals die

until 3 weeks (AJR rats) (Takahashi et al., 1984).

The number of abnormal Purkinje cells amount at the 60-80% (rats plus bilirubin

displacer) (Schutta and Johnson, 1969).

The counts of abnormal Purkinjie cells vary considerably, probably depending in

read sorption of necrotic cells variability (rat plus bilirubin displacer) (Schutta and

Johnson, 1969).

The remnants Purkinjie are normal (Takagishi and Yamamura, 1989) and the

cerebellum weight of both treated or not jj rats is the half part of the Jj rats.

Between the same genotype, treated rats, shown an additional, but not significant,

cerebellum weight loss (Jj and jj rat plus displacing agents. Animals scarified 24

H after treatment) (Conlee and Shapiro, 1997).

INTRODUCTION

23

P60 The expression of the GSH in the flocculus and in the lateral hemispheres of the

cerebellum of the jj rats is increased respect P20. But any change is detected in

cerebral cortex, thalamus, and hypothalamus and in the midbrain of Jj animals

(Johnson et al., 1993).

INTRODUCTION

24

1.6 The Blood Brain Interfaces (BBI)

The blood brain interfaces (BBI) are the key determinant in the concentration of blood

derived potentially toxic compounds (Banks et al., 1999). They are the fundamental structures

in the transport, exchange and waste of molecules in a part of the body otherwise strongly

isolated (Van Bree et al., 1992). The BBI are in number of tree; the Blood Brain Barrier

(BBB) located at the level of endothelial cells forming the micro vessels (MV), the Blood

CerebroSpinal Fluid Barrier (BCSFB) at the epithelial cells of the choroid plexuses and the

second BCSFB, formed by the aracnoid membranes (Fig 5). The barrier phenotype derives

from the presence of Tight Junctions (TJ), sealing cells together and avoiding the paracellular

passage of molecules, the presence of metabolising phase I and II enzymes, and finally, the

presence of transporters able to modulate and control the bio availability of the compounds in

the CNS (phase III). The ependyme lining the ventricles and separating the CSF from the

parenchyma is not considered a barrier due to the absence of TJ avoiding he paracellular

passage of molecules.

Two barriers are more studied in literature and are the subject of this thesis: the BBB

and the BCSFB at the choroid plexuses.

Because the work of this thesis has been done on rat animals as model, the description

that follows is for the rat postnatal development.

INTRODUCTION

25

Choroid Plexuses (First BCSFB)

Ependyma Arachnoid membrane

(second BCSFB)

MicroVessels(BBB)

Fig 5: Cartoon representation of the Blood Brain Interfaces (BBI) in Human and Rat brain. The brain interfaces are in number of four: the Blood Brain Barrier (BBB), located at the level of theendothelial cells forming the MicroVessels (MV) inside the parenchyma, the Blood-CerebroSpinalFluid Barrier (First BCSFB), formed by the choroidal epithelial cells of the 4th, 3rd and LateralVentricles Choroid Plexuses (4thV CP; 3rdV Cp, LV CP), the arachnoids membrane located betweenthe brain and the skull (Second BCSFB) and the ependyma, lining the ventricles. The latter is notconsidered a barrier.

INTRODUCTION

26

1.6a The Blood Brain Barrier (BBB)

The BBB, located at the endothelia of micro vessels, is probably the best-studied

barrier in brain. Paul Ehrlich in 1885 observed that water soluble dyes (aniline) injected into

the peripheral circulation were rapidly taken up by all the organs except the brain and

interpreted these findings as a lack of dye affinity of the nervous system. Lewandowski

interpreted differently this finding, coining the term blood-brain barrier to describe the

phenomenon. The generalized staining of the brain tissue when the trypan blue was

administered in the CSF, and a lack of brain staining if administered intravenously finally

demonstrated this concept. By the use of the electron microscopy, the blood brain barrier was

localized at the endothelial cells and tight junctions were founded to be the anatomical

responsible structure for barrier function (Brightman and Reese, 1969; Reese and Karnovsky,

1967). In 1969, Saikotos et al. reported the isolation of brain capillaries (Siakotos and Rouser,

1969). Betz et al. isolated brain capillary endothelial cells and successfully established an in

vitro blood-brain barrier model system in 1980 (Betz and Goldstein, 1980b; Betz and

Goldstein, 1980a). Since then, various in vivo, ex vivo and in vitro blood-brain barrier models

have been used in basic research and industrial drug screening, mainly with the goal of

improved drug delivery to the brain (Abbott and Bundgaard, 1992).

The BBB is a dynamic conduit for the transport between blood and brain of nutrients,

peptides, proteins, or immune cells that have access to specific transport systems localized

within the capillary membranes. This barrier protects the brain from many exogenous toxins

and sudden fluctuations in the levels of systemic substances (Jette et al., 1995a), such as

neurotransmitters.

1.6aI Development

The micro vessels forming the blood brain barrier develop late in respect to the

BCSFB.

In the foetal period, cell division produces 94-97% of the cells of the adult brain

(Caley and Maxwell, 1970) and at E11-13, the fenestrae in rat vasculature disappears (Stewart

and Hayakawa, 1994; Bauer et al., 1993).

INTRODUCTION

27

1) In the neonatal period ranging from P1 to P9, defined as “capillary sprouting period”,

the capillaries are formed by single endothelial cell, thickened from 0.3 to 1.0 μm, and

joined to it self. The pericytes are partially separated from the endothelial cells by a

discontinuous layer of basal lamina that conversely, surrounds continuously the

pericyte-endothelial complex. The arterioles appears fully developed and the post

capillary venules follow the developmental profile already described for capillaries

(Simionescu et al., 1988). The vascular profiles cont 80-90 units/mm2. During this

period the major growth in size of the brain is accomplished, mainly due to the

important increase of the cell volume, not in cell numbers of axons and dendrites. The

density of the cells (no cells/mm2) falls from 3000 at birth to 800 at P9 (Caley and

Maxwell, 1970; Stewart and Hayakawa, 1994; Stewart and Hayakawa, 1994).

2) In the myelinization period (P10-20), the endothelium became progressively

attenuated, the number of coated pits and vesicles increase remarkably. The TJ are

well developed and the basal lamina, now mature, surround continuously the pericyte–

endothelial cell complex, as well as it is present between the pericyte and the

endothelial cell (Simionescu et al., 1988). The rate of brain growth decrease markedly

and at the end of this period, the dimensions of the brain approach those of the adult

(Caley and Maxwell, 1970; Stewart and Hayakawa, 1987).

3) From P20 to P90 (young-adult period) the structure, in terms of basal-lamina,

pericytes and TJ is similar with those described or the previous stage, but the basal

lamina that surround the pericyte-endothelial cell is now surrounded by a complete

glial sheet, absent before myelinization (Simionescu et al., 1988). The astrocytic end-

feet processes develop rapidly, and synchronously the large extra cellular spaces

disappear. The density of the parenchyma cells reach the 500 units /mm2 and the

density of vessels increase strongly, reaching the 250-300 units/mm2 between P15 and

21 (Caley and Maxwell, 1970; Stewart and Hayakawa, 1987). The combined effect of

reduced cell density and increased vessel presence is summarized by the brain

cells/blood vessels ratio that from the 40 fold at P1, reaches the 1.7 score at P21

(Caley and Maxwell, 1970).

INTRODUCTION

28

1.6aII Description

With a diameter as small as 3-7 μm, the brain capillaries are the smallest vessels of the

vascular system (Rodriguez-Baeza et al., 2003). In humans, the BBB is formed by about 100

billion capillaries organized in a highly branched micro vascular network (Zlokovic and

Apuzzo, 1998) that in an adult human brain reach the 600-650 km with a surface area of about

20 m2, which makes the blood-brain barrier the third largest drug exchange surface area after

intestine and lung (Pardridge, 2003). This high capillary density, where single capillaries are

only about 40 μm apart from each other, ensures a distance short enough for small molecules

to diffuse within 1 second (Rodriguez-Baeza et al., 2003) and allow every neuron is perfused

by its own capillary to guarantee efficient nutrient supply.

Despite the huge number of brain capillaries, they occupy only about 0.1% of the brain

volume, or 1 ml in an adult human brain of 1200-1400 g (Pardridge, 2003).

Blood

Astrocytic end-feet

Endothelial cell

Basal lamina

Perycite

Fig 6: The Blood Brain Barrier Unit. The BBB unit is composed by the endothelial cells forming themicrovessels, where the TJ are located, the surrounding basal-lamina, the pericytes and the astrocytic end-feet processes. A thinlayer of basal-lamina surrounds the MV, passing between theendothelial celll and the pericyte, and sourronding the latter.

INTRODUCTION

29

The “blood barrier unit” is formed by the endothelial cells, the basement membrane,

the perycites and the astrocytic processes contacting the vessels (Begley, 2004; Abbott et al.,

2006) (Fig 6). Brain capillary endothelial cells are long and spindle-shaped polarized cells,

with an apical membrane facing the blood (luminal membrane) and a basolateral membrane

facing the brain tissue (abluminal membrane) (Betz et al., 1980), separated by only about 300

nm of cytoplasmic space (Pardridge, 2007; Pardridge, 2003). They are surrounded by the

basement membrane, composed of collagens and proteins, providing external support for the

endothelial cells (Goldstein and Betz, 1983). The cerebral endothelial cells show low

pinocytotic activity and have no intercellular clefts. However, they contain a large number of

mitochondria to meet the energy demand for active processes like metabolism and ATP-

driven efflux transport (Goldstein and Betz, 1983).

1.6aIII Tight junctions (TJ)

The structural basis of the blood-brain barrier is the presence of tight junctions sealing

adjacent endothelial cells together, in contrast to the peripheral tissue capillaries that have

interendothelial cleft passages (Deeken and Loscher, 2007; Sage, 1982).

The TJ prevent the paracellular diffusion of solutes (Nag, 2003) which are not lipid

soluble enough to dissolve into the cell membranes. Functional investigations support the idea

that the tightness of the BBB is due to the composition of the TJ (Furuse et al., 2001), which

include occludin, claudin 5 (Liebner et al., 2000; Lippoldt et al., 2000a) and 3 (Wolburg et al.,

2003) but not claudin 2, which appears to be associated with tight junction strands of lesser

resistance (Wolburg et al., 2001; Lippoldt et al., 2000b). The TJ are also responsible for the

polarization of the barrier membranes, affecting the localization of both membrane-bound

enzymes and transporters.

1.6aIV Metabolising enzymes

Several drug-metabolising enzymes have been recorded in isolated brain capillaries, at

higher levels than in the cortex (Betz et al., 1980; Ghersi-Egea et al., 1988), suggesting their

implication as an enzymatic barrier between the blood and the brain (Ghersi-Egea et al.,

1995). Alkaline phosphatase, which hydrolyses phosphorylated metabolites, is present on the

INTRODUCTION

30

endothelial cells (both luminal and abluminal) (Lawrenson et al., 1999). Cytochrome P450 is

expressed in most cerebral micro vessels, in the astrocytic foot processes and pericytes (Volk

et al., 1991). The conjugating enzyme UDP-glucuronsyltransferase is localized to rat brain

capillaries (Ghersi-Egea et al., 1994) and the glutathione S-transferase has been detected in

both cerebral capillaries and astrocytic foot processes (Bendayan, 2002). The γGT activity is

strongly expressed in micro vessels, associate to the luminal membrane, and in CPs, but is

absent in parenchyma and large vessels (Sanchez del Pino et al., 1995).

1.6aV ATP Binding cassette transporters

About the transporters, by different techniques based on the mRNA, protein and

activity detection, Pgp has been reported in human and rat brain homogenates (both protein

and Mdr1b mRNA), freshly isolated micro vessels (both protein and Mdr1a mRNA)

(Thiebaut et al., 1989; Jette et al., 1995b; Regina et al., 1998; Sugiyama et al., 2003) and in

cultured rat brain endothelial cells (both Mdr1a and 1b). A weak expression has been reported

also in neonatal rat cultured astrocytes (both protein and Mdr1b mRNA) (Decleves et al.,

2000). Functional evidence of the role of Pgp in avoiding the passage of substrates has been

obtained by the use of Pgp inhibitors (Kemper et al., 2004) (Miller et al., 2000) or by KO

mice (Schinkel et al., 1996; Karssen et al., 2001; Uhr et al., 2002).

The Pgp role in the BBB may be the control of the bio availability of several

endogenous opioid peptides such as adrenorphin and endomorphin 1 and 2, as well as the

neurokinin, Substance P (Oude Elferink and Zadina, 2001), additionally a role of Pgp in

controlling the efflux of glutamate from the BBB was suggested by in vitro and in vivo

inhibitory experiments (Liu and Liu, 2001). The Pgp seems avoid the brain influx of the

antiepileptic drug phenytoin (Potschka and Loscher, 2001), confirming the role of this

transporter up-regulation in BBB and astrocytes in resistant epilepsy (Rizzi et al., 2002;

Tishler et al., 1995; Sisodiya et al., 1999; Sisodiya et al., 2001; Sisodiya et al., 2002).

Despite its importance, the localisation of Pgp in the BBB is still controversial. Pgp is

reported to be luminal, together with the GLUT-1 (glucose transporter) but not GFAP (glial

fibrillary acidic protein), in membrane enriched fraction of MV (Beaulieu et al., 1997). The

staining appears as a continuous endothelial like signal (Mercier et al., 2004) co localized

with caveolin1 on the human MVs ex vivo (Virgintino et al., 2002). At the contrary, others

INTRODUCTION

31

authors report the co-localization of Pgp with the GFAP, but not the GLUT1, by confocal

microscopy (Pardridge et al., 1997), and suggest that the Pgp trasporter in human MVs could

be expressed from the astrocytes end feet contacting the endothelial cell (Golden and

Pardridge, 1999). Additionally, in the RBE4 cell culture, that maintain the morphological

characteristics of a endothelial cell forming MV, Pgp is on the plasma membrane,

plasmalemma vesicles, and nuclear envelope (Bendayan, 2002).

Mrp1 is reported on isolated MV (both protein, mRNA and functional assay)

(Sugiyama et al., 2003; Miller et al., 2000; Kusuhara et al., 1998). The discontinuous

labelling, co-localized with glial fibrillary acidic protein (GFAP), suggests a preferential

expression of Mrp1 by the astroglial component of the blood–brain barrier (Mercier et al.,

2004). The expression of Mrp1 is most present in cultured endothelia cells (Decleves et al.,

2000) than in isolated MV. Generally, Mrp1 is up-regulated in cell culture (Decleves et al.,

2000) in a time-dependent manner. In primary cultured endothelial cells co-cultured with

astrocytes, the expression of Mrp1 mRNA is less important and the Mrp1 activity is low

(Regina et al., 1998; Torok et al., 2003) indicating that the culture conditions may modulate

drastically the ABCs transporters expression (Berezowski et al., 2004). Also the barrier

properties, such as TJ and transporters expression, is clearly modulate (Abbott, 2002),

rendering all data derived from cultured cells doubtful when applied to the in vivo situation.

The mRNA of MRP4, MRP5 and MRP6 has been demonstrate in bovine primary

endothelial cell cultures and in capillary-enriched fraction from bovine brain homogenates

(Leggas et al., 2004; Nies et al., 2004; van Aubel et al., 2002). In mouse brain Mdr1 and

Mrp5 have been localized on the luminal side, Mrp1 on the abluminal (basal) side, and Mrp2

in between. Mrp3 and Mdr3 are undetectable (Soontornmalai et al., 2006) and MRP4 is at the

apical pole of human brain endothelial cells (van Aubel et al., 2002; Leggas et al., 2004; Nies

et al., 2004). The Mrp2 mRNA has not been detected nor isolated MV nor on cultured

endothelial cells (Sugiyama et al., 2003; Zhang et al., 2000), but their protein expression and

function in BBB have been revealed in rat (Potschka et al., 2003). Oatp2 is expressed at the

rat blood-brain barrier (Gao et al., 2000), and is both apical and luminal on rat (Sugiyama et

al., 2001).

INTRODUCTION

32

1.6b The Blood Cerebro-Spinal Fluid Barrier (BCSFB)

If the BBB is the better know brain barrier and our actual knowledge reassume more than 120

years of investigations, the BCSF barrier has been less investigated.

In 1922 the cerebral blood vessel compartment in which choroid plexus was semi

permeable, facilitating the flow of substances from the blood into the CSF, was defined as the

"Barriére hémato-encephalique" (hemato encephalic barrier), and in the 1963, was described

the "sink effect" gradient favouring the passage of substances in extra cellular fluids from

brain to CSF with the CSF constantly circulating and carrying substances away (DAVSON,

1963).

The development of choroid plexuses is precocious in foetal life, suggesting a major

role in the nutrition of embryonic early brain. The CPs are involved in the control of the CNS

extra cellular environment, participate to the brain development secreting morphogens,

mitogens and tropic factors driving both the cerebral and cerebellar growth, and participate in

the neuroprotection (Borlongan et al., 2004; Strazielle et al., 2005; Dziegielewska et al.,

2001).

1.6bI Development

The CPs develops early in all mammalian species, at the time when the brain is poorly

vascularized. In the rat the 4thV CP appears at the E12, the lateral at the E13 and the 3rd at

E16. The blood flow in both the lC and 4thV CP increase considerably between 3 and 4 weeks

postnatal, reaching values 5 times higher than in the cerebral cortex at every post-natal age.

For developmental stages have been identified for the choroid plexuses cells.

1) In the first stage the epithelial cells are tall, pseudo stratified; they lack glycogen and

have central nuclei.

2) In the second stage they lower and columnar, with an apical nuclei and abundant

glycogen.

3) In the 3Th are cubical, the nuclei are in the apical side and the cells posses only

moderate glycogen.

4) Finally, at the stage 4, the epithelial cells are cubical, squamous without glycogen and

the nuclei are central or basal.

INTRODUCTION

33

The TJ in CPs probably develop as early as the epithelial cells differentiates, and are

impermeable to HRP as early as E14 (Dziegielewska et al., 2001; Dziegielewska et al., 1980;

Knott et al., 1997). So, the paradox of higher CSF protein concentration in early postnatal

development respect than in adult life, has been resolved by the identification of regulated

pathway for the transfer of molecules, rather than with a decrease in permeability due to the

weakness of the barrier phenotype in the early age of life (Dziegielewska et al., 2001)

(Johansson et al., 2006).

1.6bII Description

The blood cerebrospinal fluid barrier is located at the “cauliflower like” choriod

plexuses (Fig7) within the for ventricles of the brain (one in each LV CP, one in the 3rdV CP

and one in the 4thV CP) where they form the main interface between the blood and the CSF

(blood-CSF barrier or BCSFB).

Fig 7: Structure of the Choroid Plexuses. In CPs the barrier is shifted on the specialized epithelial cells monolayerforming the choroid plexuses. Here the TJ are locate and seals cellstogether, surrounding a central stroma in witch the fenestrated vessels arepresents. The epithelial cells forming the adjacent ependyma, lining theventricle cavities does not posses barrier properties.

INTRODUCTION

34

The CPs belong to the para-ventricular organs and are characterized by a conjunctive

stroma containing large fenestrated capillaries that let pass the molecules from the blood to

the single layer of choroidal epithelial cells. These cells display as increased surface area, due

to the basolateral infoldings and apical microvilli (Keep and Jones, 1990). In rodents, this

apical surface area is close to these developed by the BBB [26]. The macroscopical

morphology of CPs differs (see Fig 9), the LV CP appear as a thin “velae” containing large

fenestrated vessels and, in adult rat weight about 120 μg, wile the 4thV CP posses a more

convolute structure and weight 80 μg in adult rat. Also if many authors point out to the CPs as

a single entity, and most of the studies probably involve only the lateral CPs, recently

biochemical differences between CPs from the different anatomical districts, have been

described (Strazielle and Ghersi-Egea, 1999; PAPPENHEIMER et al., 1961). The CPs are in

the perfect location for distributing molecules both locally and globally on the brain (Emerich

et al., 2005). In fact they are located between two circulating fluid, posses a high surface area

available for exchange, the highest blood flow among the brain structures (Al-Sarraf and

Philip, 2003; Chodobski et al., 1995) and a more ‘leaky’ phenotype, that raise the exchange

properties of the BCSFB and favour blood-to-CSF rate of influx (Strazielle and Ghersi-Egea,

2005; Strazielle et al., 2005; Strazielle et al., 2004).

1.6bIII Tight junctions

Within CPs the barrier function is shifted from the vasculature to the apical side of the

epithelial cells, where the tight junctions are located. Differently from the BBB, the tight

junctions at the choroid plexuses include occludins, claudin 2 and 1, but not 5 (Lippoldt et al.,

2000b; Wolburg et al., 2001). This difference maybe account for the more ‘leaky’ phenotype

of this barrier, and for the resulting higher paracellular permeability (DAVSON, 1963) as

demonstrated by the slowly diffusion into the CSF through the CPs of polar drugs, whereas

their penetration across the BBB is almost completely prevented (Thomas and Segal, 1998).

INTRODUCTION

35

1.6bIV Metabolising enzymes

The CPs posses the higher expression of phase I and II drug metabolising enzymes ,

that renders this organ able to influence the central availability of toxic compounds (Strazielle

and Ghersi-Egea, 2005) since the early development (Strazielle et al., 2004). The cytochrome

P-450s (Pardridge et al., 1983), the NADPH–cytochrome P-450, UDP-glucuronosyl

transferase (UGT), glutathione S-transferase (GST) (Strazielle et al., 2003) or

sulfotransferase, together with antioxidant enzymes, in particular glutathione peroxidase,

catalase, superoxide dismutase and glutathione reductase renders this organ the “liver of the

brain” (Strazielle and Ghersi-Egea, 1999; Strazielle et al., 2004; Strazielle and Ghersi-Egea,

2005). Additionally, CPs contains high concentration of glutathione and cysteine that may

sequester toxic agents circulating in CSF, and together with the enzymes, provide a barrier

protecting against free-radical oxidative species (Emerich et al., 2005).

1.6bv ATP Binding cassette transporters.

In CPs both Pgp and Mrp1 proteins have been identified in rat, mouse, and human. In

this study, Pgp has been reported to be subapical, conferring an apical-to-basal transepithelial

permeation barrier to drugs (Rao et al., 1999). But the presence of this transporter in the CP

has been denied, at least at the level of the endothelium of the sinusoidal vessels (Cordon-

Cardo et al., 1989; Drion et al., 1996; Drion et al., 1996).

In the choroidal cells Mrp1 is well expressed (Mercier et al., 2004) and localize

basolaterally, conferring an opposing basal-to-apical drug-permeation barrier (Rao et al.,

1999). The functional importance of Mrp1 in the choroidal tissue has been confirmed by the

10-fold increase of etoposide levels in the Mrp1/Mdr1a/Mdr1b triple-knockout (TKO) mice,

respect to the Mdr1a/Mdr1b DKO (Wijnholds et al., 2000)

The mRNA of several ABC transporters has been reported. Mrp4, and 5 seems to be

well expressed (Loscher and Potschka, 2005), Oatp3, Oatp2 are present, whyle the remaining

transporters (Mrp2, Mrp3, Oatp1, Oatp4, Oatp5, Oatp12, Mdr1a, Mdr1b, Mdr2) are expressed

at very low levels (Choudhuri et al., 2003). About their localization, Mrp2 localize on

endothelial cells, and Mrp3 was co-localized with zonula occludens-1 at tight junctions

(Soontornmalai et al., 2006), Mrp4 is expressed baso-laterally (Leggas et al., 2004), indicating

a role for this transporter in limiting organic anion influx from blood and in driving organic

INTRODUCTION

36

anion efflux from the brain to blood (Dallas et al., 2003). The Oatp1 and Oatp2 (Breen et al.,

2004) are localized at the apical and basolateral domains, respectively, of the rat choroid

plexus epithelium (Gao et al., 2000) were they transport opiod peptides.

The best-known function of the CP is the secretion of most of the CSF, which

comprises half part of the extra cellular fluid in the mammalian brain. This fluid, specific to

the CNS, is mainly secreted by the choroid plexuses located in the lateral ventricles and flow

into the ventricular cavities of the brain by a tightly regulated process. It flow s trough the

ventricular system connections formed by the Foramen of Monroe (from lateral to third

ventricle), the Sylvious aqueduct (from the caudal part of the cortex to the forth ventricle),

from Foramen of Luschka (lateral to the 4thV) and the Magendie foramen (roof of the 4thV) to

the Cisterna Magna. Then the CSF reaches the different subarachnoid spaces surrounding the

whole brain, and finally goes out from the brain through the arachnoids villi into the venous

system, and to some extent along the nerve roots into the lymphatic (Johansson et al., 2006).

1.6c The parenchyma

The ventricular CSF and the interstitial fluid of the parenchyma are separated by the

ependyma that boards the cavities of the brain. The epithelial cells forming the ependyma are

joined only by adherents junctions, which leave open the paracellular pathway, and are not

able to confer barrier property. Thus the molecules could diffuse in both directions. The

permanent circulation of the CSF drains the molecules by the parenchyma and, by the other

side, the molecule able to cross the CPs barrier are quickly distributed by the permanent flow

of this fluid in brain parenchyma across the leaky ependyma.

If the metabolising enzymes are barely detectable in brain parenchyma, several

transporters have been recognized and they may act as a second barrier, inside the tissue. Pgp

has been demonstrated in microglia (Lee et al., 2001; Dallas et al., 2003), in tanycytes lining

the 3rd ventricle (protein) (Mercier et al., 2004) and has been reported in parenchyma (mRNA,

Mdr1b ). Mrp1 is present in human perilesional samples (mRNA) (Nies et al., 2004;

Hirrlinger et al., 2002), cultured astrocytes, microglial, neurones, oligodendrocytes (mRNA)

(Hirrlinger et al., 2002)and in tanycytes, astrocytes, in glia limitans and in the ependyma

(protein) (Mercier et al., 2004). Mrp4 and Mrp5 (mRNA) have been detected in human

perilesional samples (Nies et al., 2004; Hirrlinger et al., 2002), in astrocytes of the sub cortical

INTRODUCTION

37

white matter (Nies et al., 2004), and in cultured astrocytes and microglia, whereas the Mrp4

mRNA was only marginal present in cultures of neurones and oligodendrocytes (Hirrlinger

et al., 2002).

The Mrp3 mRNA has been detected in human perilesional samples (Hirrlinger et al.,

2002; Liebner et al., 2000; Nies et al., 2004) and as weakly expressed on cultures of neurones

and oligodendrocytes (Hirrlinger et al., 2002).

_

INTRODUCTION

38

1.7 The UCB transporters

The UCB transporters so far recognized belong to the ATP Binding Cassette

transporters (ABC) family. The family comprises a large membrane bound transporters

interacting with a wide range of substrates. They have been discovered in the late 1970 in

cells expressing a multidrug resistence phenotype, and so called MultiDrug Resistance and

Multidrug Resistance Associted protein (MDR or MRP). They are formed by a combination

of characteristic domains, including membrane spanning domains (TransMembrane Domain:

TMD) and cytoplasmatic ATP binding domains (Nucleotide Binding Domain: NBD) (Cole

and Deeley, 1998). A sequence present only in the ABC transporters and named ABC-

signature define the family of appurtenance. Despite the presence of numerous ABC family

transporters in blood brain interfaces, we are interested in Pgp (ABCb1) and Mrp1 (ABCc1)

because they have been involved in UCB transport. Both share the typical ABC transporters

molecular architecture (Fig 8) and show a unidirectional transport of their substrate (Borst and

Elferink, 2002).

TMD0

Pgp

NBD1 NBD2

TMD2TMD1

NH2

COOH

Mrp1

COOH

Fig 8: Schematic representation of Mrp1 and Pgp structure. Mrp1 (upper) is formed by 1531 amino acids (aa) and posses two glycosilation sites (28 KDa),for a total molecular weight of 190 KDa. Pgp (down) is formed by 1280 aa and posses twoglycosilation sites (30 KDa), for a total weight of 170 KDa. In humans Pgp is encoded only bythe MDR1 gene, that corresponds to the two rodents isoformes Mdr1a and Mdr1b.Abbreviations: TMD: Ttrans Membrane Domain; NBD : Nucleotide Binding Domain .

NH2

NBD1NBD2

TMD1 TMD2

TMD0

INTRODUCTION

39

1.7a Mrp1 (ABCc1)

MRP1 (Mulridrug Resistance associated Protein 1 or ABCc1; capital if referred to

humans, if not to animals) is the well-characterized transprter belongs to the ABC family. It

was discovered in the 1992, in a lungs cell line that showed doxorubicin resistance in absence

of MDR1 (Multi Drug Resistance) expression. MRP1/Mrp1 is ubiquously expressed on the

body and localize basolaterally (Cole et al., 1992). Its strong expression at the barriers (CP,

intestine, etc.) suggests a role in the body defence, by limiting the entry of exogenous, as well

as, blood burned potentially toxic compounds (Schinkel and Jonker, 2003). Also if difference

in substrate affinity between humans, rat and mice species have been reported, Mrp1

transports conjugated compound, the oxidized form of GSH (GSSG) (Leslie et al.,

2001),shows a high affinity for the inflammatory mediator leukotriene C4 (LTC4) and

probably plays a significant role in mediating immune responses (Haimeur et al., 2004). In

both human and rat the full length, glycosilated protein has 190 KDa, and display an

homology of around 93% (Gennuso et al., 2004).

The idea indicating in Mrp1 a key player in bilirubin toxicology became mainly from

our laboratory. The first evidence was done by transport studies performed on polarized

human BeWo (trophoblastic cell line) cells. These cells show a high apical (maternal facing)

expression of Mrp1 and a strong UCB release, characterized by a saturative kinetics and

inhibited by MK571 Mrp1 specific inhibitor (Pascolo et al., 2001). The result has been

subsequently confirmed in human placental trophoblast plasma membranes vesicles (Serrano

et al., 2002). Additionally the mortality in MEF (mouse embryonic fibroblast) cells derived

from Mrp1 KO mice was demonstrate to be higher if compared to MEF from Wild Type mice

after bilirubin treatment (Calligaris et al., 2006). In 2004, Rigato proved the role of this

transporter in the mediated transport of UCB using trasfected cells. The 3H-UCB transport in

plasma membranes vesicles was 3-5 times higher in Mrp1 stably trasfected MDCKII cells,

compared to Mrp2 tranfected and wild-type MDCK cells. The transport was increased by

GSH and inhibited by the Mrp1 inhibitor MK571. The transport of LTC4, a high affinity

substrate for Mrp1, was inhibited by UCB. The determined Km and Vmax for UCB Mrp1

mediated transport were 10±3 nM (Bf) and 100±13 pmolxmin-1x (mg of protein),

respectively (Rigato et al., 2004).

The evidence that the sensibility toward UCB inversely correlated with the Mrp1

expression, indicating a plausible role of the transporter in protecting neuronal cells, has been

INTRODUCTION

40

obtained by Falcao (Falcao et al., 2007). In fact, the maximal Mrp1 expression was present in

astrocytes at every time in culture respect the cultured neurons, with the latter more sensible

to bilirubin. Similarly the astrocytes showed the strongest up-regulation with the time in

culture, with the “younger” more sensible than the cells cultured for more days (Falcao et al.,

2007). Additionally, on cultured primary mice astrocytes, bilirubin reallocates Mrp1 from the

perinuclear region to the plasma membrane and, simultaneously, up regulate both mRNA and

protein expression (Gennuso et al., 2004).

1.7 b Pgp (ABCb1; MDR1; Mdr1a/b)

The Pgp is encodes in humans by the unique gene MDR1 in rodents the two Mdr1a

and Mdr1b isoforms are present, sharing an 85% of homology inside the rat (Borst and

Elferink, 2002); (Hooiveld et al., 2002). In both species, the full length and gliycosilate Pgp

shows an apparent molecular weight of 170 KDa (Hooiveld et al., 2002). The substrate

specificity is overlapped with these of Mrp1.

Also if UCB is reported a week substrate for Pgp (Jette et al., 1995a) several evidence

suggest its implication in the brain protection toward bilirubin. In vivo the UCB content in the

Mdr1a/b KO mice is 2 fold higher than in the Wild Type mice 10 min after intravenous UCB

infusion, without affecting the UCB clearance (Watchko et al., 2001; Watchko et al., 1998).

Similarly in younger-adult rats (P32-36 days old), in witch the Pgp expression in brain seems

to be similar to the adult level, the Pgp brain inflow is enhanced of the 24% by the Pgp

inhibitor erythromycin, and of the 236% by the Pgp inhibitor and bilirubin-albumin displacing

agent ceftriaxone (Hanko et al., 2003).

AIMS OF THE STUDY

41

AIMS OF THE STUDY

The brain is a peculiar organ in witch a fine homeostasis of the extra cellular fluid is

required for the neural cells to fulfil their complex physiological functions. This homeostasis

is achieved and controlled by several mechanisms, which primarily involve the specific

features of blood brain barriers, the BBB and the BCSFB, in controlling the exchanges

between the peripheral blood and the CNS.

Among efflux transporters presents on the BBI, Mrp1 and Pgp appear to be actively

involved in the control of the cerebral bioavailability of biologically active or toxic

compounds such as bilirubin.

Bilirubin in its unconjugated form (UCB) has been suggested to be a potent

antioxidant at low concentration while it seems to be extremely dangerous at higher

concentrations, causing encephalopathy in severely jaundiced neonates by damaging

astrocytes and neurons. Around 70% of children with kernicterus die within seven days, while

the 30% survivors usually suffer irreversible sequels, including hearing loss, paralysis of

upward gaze, mental retardation, and cerebral palsy with athetosis.

Despite the importance of Mrp1 and Pgp on BBI their pattern of expression and

cellular localization remains still unsettled. Based on these considerations

- the first aim of the thesis was clarify the relative protein expression of these

transporters at the two major BBI protecting the brain from toxic insults (Ia), and to identify

their post-natal developmental profile of expression and cellular localisation (Ib).

Many efforts have been made to understand how bilirubin enters the brain. The actual

data mainly concern the BBB, located at the endothelia of the MV, leading to the “free

bilirubin theory”. At the contrary no data have been provided about the role of the CPs and no

information is available on the contribute of the BCSFB in protecting the brain from the

bilirubin since no data about the Mrp1 and Pgp expression on BBI during the bilirubin

encephalopathy are available.

- The second aim of the thesis was investigate a relation between the high level of

blood bilirubin and Mrp1 and Pgp expression in brain barriers in vivo using the Gunn rat (II),

a model for Kerniterus and Crigler-Najjar Type I syndrome.

MATERIALS AND METHODS

42

1) Animals

Two different strains of rats were used in this Thesis. In the first task, we used adult

(60 days after the birth, P60), P9 and P2 Sprague-Dawley rats, that were purchased from

Harlan, Gannat, France.

In the second task we used adult (P60), P17, P9 and P2 SpdBlueGUNN:j-rats from our

colony. The animals, originally purchased from Harlan Illinois in 2006, were kept in the

animal facility of the Università degli Studi di Trieste.

All the experimental procedures were performed in compliance with the guidelines of

the European (86/609/EEC), French Ethical Committee (decree 87-848) and Italian (D.L.

116/92) lows.

2) Human samples

Two human cortical samples and three choroids plexuses samples were obtained at

autopsy from adult human within 24 hours after their death from accidents, according to the

ethical guidelines approved by INSERM. They were procured by the Service de

Neuropathologie, Hôpital Neurologique centralized via the Biological Resource Center

NeuroBioTec Banques, Lyon, France. All samples were kept at 4°C in Krebs-Ringer buffer

prior to processing or freezing. The choroidal nature of the samples was confirmed by

conventional histology.

3) Rat Tissue (Choroid Plexuses, Cortices and Cerebella) dissection

The animals were anaesthetized by ethyl-ether and decapitated, their brains rapidly

excised and placed at 4 °C in Krebs-Ringer buffer.

Under stereomicroscope, the choroids plexuses (4thV and LV CP) were dissected from each

anatomical district, pooled and rapidly frozen at -80 until use.

Then the cortices were isolated, cleaned from all apparent meninges and pooled, forming the

starting sample for the micro vessels isolation procedure.

For Gunn rats, also the cerebella (CLL) were dissected under stereomicroscope, and placed at

4 °C in Krebs-Ringer buffer for subsequent weight determination. The weigh of the dissected

Cx and CLL were calculated in g/animal, as difference of the tube weigh prior and after the

MATERIALS AND METHODS

43

tissue collection (pooled par age, and additionally par genotype and par gender for Gunn),

divided for the number of animals forming the sample.

4) Micro Vessels isolation procedure.

The protocol we used for the micro vessels (MV) isolation was described by Ghersi-

Egea (Ghersi-Egea et al., 1994). This protocol consists in different mechanical

homogenisation and separation of MV from contaminants steps, by dextran gradient and

passage through mesh sieve at decreasing pore diameter. At every developmental stage, the

meninges free cortices dissected from different animals were pooled to reach the amount of

tissue needed for the protocol. The protocol for the isolation of MV from 9-days old animals

(INSERM) and from 17-days old Gunn rats (CSF) was settled up by us by some modification

of the protocol used for adult animals.

4.a) Adult

The pooled cortices were minced in Krebs-Ringer buffer (KR) and homogenized by 5

strokes of A-type pestle in a Dounce-type glass-glass homogenizer. a) An aliquot of

homogenate for gamma glutamine transferase (γGT) measurement was token, then the

preparation was homogenized by 15 strokes of A pestle in 5 vol/g tissue of 1 % bovine serum

albumin-supplemented Krebs-Ringer buffer (1%-AKR buffer) and the homogenate was

filtered through a 500 µm-mesh sieve. b) The filtrate was diluted (1/1) with 1%-AKR buffer

and homogenized again with 25 strokes of B-type pestle. The homogenate was centrifuged at

1000 g for 10 minutes and the supernatant discarded. c) The pellet was suspended in Krebs-

Ringer buffer containing 17.5 % 70 KDa-dextran and centrifuged at 3000 g for 27 minutes in

order to discard the myelin retained at the surface of the gradient, thus the resulting pellet was

suspended in 1%-AKR buffer, filtered through 200 µm-mesh, then 74 µm-mesh sieves to

discard larger vessels. d) The filtrate was centrifuged at 125 g for 15 min and the resulting

pellet was suspended again and filtered on a 40 µm-mesh sieve. e) The micro vessel fraction

retained on the sieve was recovered in 0.1 %-AKR buffer, centrifuged at 125 g and suspended

in a small volume of the same buffer. All steps were carried at 4°C within 4 hours. The

quality and purity of the final preparations was visually jugged by contrast phase microscopy.

MATERIALS AND METHODS

44

The preparations were kept frozen at -80 °C until use for Western blot or enzymatic activity

analysis.

Human micro vessels were prepared using an identical procedure.

4.b) Micro vessels isolation from 9- and 17-days old rats

During the first week after the birth the network of cerebral micro vessels is still

forming and a complete basal lamina is not present. The P9 age, signs the end of this

sprouting period at the end of witch a basal lamina is present. Since a complete basal lamina

is a pre-requisite for maintaining the capillary integrity during the isolation procedure, P9

corresponds to the early post-natal age allowing to isolate micro vessels (Simionescu et al.,

1988; Caley and Maxwell, 1970). Similarly, at P17 the basal lamina is still arriving to the full

maturation (Simionescu et al., 1988), and some adaptations on the protocol settled-up for

adult MV isolation, are needed. Additional adaptations have been necessary for the transfer of

the protocols at the CSF. The improved modifications on the “adult” protocol are reported

above.

P9: INSERM: 8 animals per age. Homogenization obtained by a single step of 5 B-type

pestle strokes, followed by a filtration on 500 µm-mesh sieves. After

centrifugation at 1000 g for 10 minutes, the procedure resumed directly at

step (c).

CSF: 8 animals per age and gender. Homogenization obtained by a single step

of 3 B-type pestle strokes, followed by a filtration on 500 µm-mesh sieves.

After centrifugation at 1000 g for 10 minutes, the procedure resumed directly

at step (c).

P17: (only CSF) 6 animals per age, per gender. Homogenization obtained by a single step of 8

B-type pestle strokes, followed by a filtration on 500 µm-mesh sieves. After

centrifugation at 1000 g for 10 minutes, the procedure resumed directly at

step (c)

MATERIALS AND METHODS

45

5) Sample Homogenization and protein dosage

Whole CPs and cortices (meninges-free) dissected from different animals were pooled

and homogenized using a Dounce-type glass-glass homogenizer in homogenisation buffer

(HB) supplemented with a cocktail of protease inhibitors.

The total protein content in samples was measured by the Peterson method at the INSERM

(5a) and by the BicinConininc Acid Kit (BCA) at the CSF (5b). In both the amount of total

protein in the samples were calculate respect a standard curve done by serial dilution of the

albumin solution.

5a) Peterson method

The method is based on the combined oxidation of the peptidic bounds in alkaline

solution by the Cu ions (Biuret method), and the phospho-tungsten and phospho-molibden

reduction by the Folin Ciocalteau reactive. In brief: a) The samples were diluted in H2O to the

final volume of 100 μL. Similarly a standard curve was done by serial dilution of the

reference sample (BSA, 1mg/mL). b) The reaction solution (1 mL) was added to the samples

as well as to the standard solution, and the mix incubated 10 min at RT. c) 500 μL of Folin

reactive were added and the reaction product formation (blue) detected at 750 nm.

5b) BicinConinic Acid Kit (BCA)

The procedure has been performed following the producer instructions (Sigma-Aldrich

Procedure n° TPRO 562). Briefly: a) the samples were diluted in H2O and 100 μL of the

Solutions and instrumentations used for the described procedure (1-4).

Krebbs-Ringher Buffer (NaCl 135 mM; KCl 0.298 mM; CaCl2 2.2 mM; MgCl2 1.2 mM;

NaHCO3 6 mM; Hepes 10 mM; Glucose (D) monohydrate 5 mM, pH 7.4)

500, 200, 74 μm-mesh sieves (Merk_Netwell, Corning Inc, Corning, NY); 74 μm-mesh

sieves (BD Bioscience, Erembodegem, Belgium).

Centrifuge: INSERM_ Heraeus multifuge 3 L_R, equipped with a Heraeus 75006441K rotor

CSF_Beckmann Coulter Allegra 25R, equipped with a TS-5.1-500 rotor.

MATERIALS AND METHODS

46

reaction mix (Bicinconinic acid : Cu = 50 : 1) were added. b) after 30 min at 37 °C, the

reaction product (purple) was detected at 562 nm.

6) Spettrophotometric Enzymatic procedures for γGT activity measure.

The gamma glutamyl transferase (γGT) activity, a marker of micro vessel enrichment,

was determined in tissue homogenates as well as in micro vessels preparations following the

protocol described in Ghersi-Egea-94 (Ghersi-Egea et al., 1994). The γGT activity is highly

specific and strongly expressed on micro vessels, but not by the parenchyma it self or by the

larger vessels (Lawrenson et al., 1999). So, the Mv/Cx ratio in γGT activity allows judging

the enrichment of the preparation. In presence of glycilglycine and L-γ-glutamyl-3-carboxy-4-

nitroanilide as a substrate, the L-gamma-glutamil-glycylycine and the 3-carboxy-para-

nitroanilin were formed. The reaction may be followed at 405 nm. For more details, see

reference.

7) Enzymohistochemistry detection of γGT on brain slices.

The γGT activity was also revealed on brain sections by enzymohistochemistry, using

a previously published procedure (Satoh et al., 2005) modified as follows :

a) 10 μm-thick sections were collected from unfixed frozen brains and incubated in the

reaction buffer for 20 min at room temperature. B) Then the slices were washed with NaCl

0.85 % (saline), and c) incubated for 2 min in 0,2 M CuSO4. d) After wash with saline

solution and water, e) they were counter-staining with haematoxylin. In both procedures, the

Solutions and instrumentations used for the described procedure (5a-b).

Homogenistaion Buffer: sucrose 0.25 M; KH2PO4 40.2 mM, K2HPO4 9.8 mM; EDTA 1 mM;

DTT 0.1 mM; pH 7.4.

Protease inhibitor cocktel (Complete, Roche Diagnostics, Basel, CH)

Reaction solution: Na2CO3 21.6 mM; CuSO4 6.75 nM; potassium-sodium tartrate 22.5 nM;

NaOH 0.9 M; SDS 1.125%.

Folin reactive: (1/6 in H2O)

BCA kit: Sigma (B9643 and C2284)

MATERIALS AND METHODS

47

specificity of γGT activity was demonstrated using the specific inhibitor (0.5 and 1 mM in

spetcrophotometric, and 0.1 and 2mM in brain slices, on P9 and adult samples respectively).

8) SDS page and Western blot

The qualitative assessments of the anti-ABC transporters (MW: 170-190 KDa)

antibodies specificity was performed running the samples on 7% polyacrylamide gels

prepared following the Laemly protocol. All samples were diluted in Laemly loading buffer.

The quantification was performed by traditional chemiluminescence. To over-step the limits

in the procedure due to the absence of a linear relationship between the densitometry signal

on X-Ray film and the amount of protein of interest, we have developed the following

quantification procedure.

a) A 10% SDS-polyacrilamide gel to detect on the same membrane both the ABCs

transporters signal and the actins signal was used. By this way we avoided errors in

quantification due to different loading, different handling, different efficiency in transfer, etc.

b) On every gel a serial dilution of the reference sample was loaded, and the concentrations of

the first and second antibodies were chosen to obtain an efficiency curve as linear as possible

(Fig 13 and Fig 27). c) The amount of the samples under analysis loaded on the gel, was

evaluated to obtain a signal that may be contained in the linear part of the standard curve,

avoiding errors in quantification when the signal approach the plating parts of the curve. d) To

normalize the expression of ABCs transporters, we used the actins signal obtained on the

same gel. e) The transfer procedure was settled up to obtain a 100% efficiency jugged by the

Coomassie blue coloration (O/N) of the gel post transfer. Only the absence of any residual

Solutions and instrumentations used for the described procedure (6 and 7).

Reaction buffer in 6: TrisBase 120 mM; glycylglycine 120 mM; Triton X-100 500 μg/mL

Substrate: L-γ-glutamyl-3-carboxy-4-nitroanilide 73 mM.

Reaction buffer in 7: 0,57 mM γ-glutamyl-4-methoxy-naphtylamide, 3.3 mM

glycylglycine, 93.0 mM NaCl, 0.52 mM Fast Blue BB Salt, 22 mM

Tris-HCl, pH 7,6.

CuSO4: 0,2 M

Spectrophotometer: dual-beam Cary 100 spectrophotometer.

MATERIALS AND METHODS

48

band was jugged as a good transfer for subsequent quantification. f) To avoid the passing

trough the membrane of the lower proteins (actin 42KDa), the gel was cut before the transfer

and the upper part, containing the high MW ABCs (170-190 KDa), was transferred for 2h; the

lower part, containing the actin, for 1h. g) The transfer was controlled by Ponceau

h) The membranes were saturated in a blocking solution 1 hr, and then incubated overnight at

4°C with the first antibodies at the appropriate concentration in blocking solution. i) After

three rinses, the membranes were incubated with the respective horseradish peroxidase

conjugated secondary antibodies for 2 hr. l) rinsed tree times and developed using a

chemiluminescence procedure, according to the manufacturer's protocol, and visualized on X-

ray films.

Additional detection of very low signal (such as Pgp on CPs) was occasionally performed by

a high sensibility developer system, Super Signal West Femto (Pierce, Rockford, IL_code

34094).

9) Western blot quantification procedure

a) The optical density profiles were generated for each gel lane from the scans of X-

ray films, b) and the surface area of all the band of interest in the gel was measured in

Arbitrary Units (AU) (ImageQuant software, GE Healthcare Europe GmbH, France). c) A

standard curve was generated (Fig 13 and Fig 27) plotting the AU respect the μg of protein

loaded on the serial dilutions of the reference sample and the amount of the analysed protein

was calculated by interpolation from the non linear regression analysis curve before generated

(CurveExpert 1.3 software). d) The results were expressed as a percentage of the ABC protein

present in the reference sample, normalized for the protein load. The latter values were

controlled and adjusted using the actin signal revealed on the lower part of the same gel.

MATERIALS AND METHODS

49

Solutions and instrumentations used for the described procedure (8-9).

Running gel 10%: acrylamide 10%; bis-acylamide 0.3%; tris 0.375 M (pH 8.8); SDS 0.1%;

APS 0.1%; TEMED 0.004%.

Stacking gel: acrylamide 5%; bis-acylamide 0.15%; Tris 0.125 M (pH 6.8); SDS 0.1%; APS

0.1%; TEMED 0.1%.

Laemly loading buffer : Tris 62.5 mM; glycine 8%; SDS 1%; BBF 0.2%; 0.1 M DTT; pH 6.8.

Migration Buffer: 25 mM Tris base; 192 mM glycine; 0,1 % SDS

Transfer buffer: 20% methanol in 25 mM Tris base; 192 mM glycine; 0,1 % SDS

Destainer. 30% methanol, 10% acetic acid

Coomassie blue solution and Ponceau solution for protein detection (Sigma, St Lois, MO, USA)

Chemiluminiescence detection kit: Millipore, Billerica, MA, USA

X-Ray Films: Kodak Biomax Light (Kodak Rochester, NY, USA)

Anti-Mrp1 Ab

(both 1 μg/ml)

INSERM_ A23 (Alexis Biochemicals, Lausen, switzerland)

CSF_A23 produced in pour laboratory

Anti-Pgp Ab

(both 4.8 μg/ml)

INSERM_ C219 (Calbiochem, Darmstadt, Germany)

CSF_C219 (Abcam, Cambridge, UK)

Anti-Actins Ab Both INSERM and CSF_A2066 (Sigma, St Lois, MO, USA)

(1.2-0.6 μg/ml at the INSERM and CSF, respectively)

Anti-Rabbit and

anti-Mouse

HRP coonjugate

Abs

INSERM_both Jackson Immuno research (West Growe, PA, USA),

Both 0.8 μg/mL

CSF_both Sigma, St Lois, MO, USA

Anti-Rabbit: 0.3 μg/mL and anti-Mouse: 0.8 μg/mL

Blotting

Membrane

INSERM_ Protran, 40 µm pores

(Whatman Schleicher & Schuell, Dassel, Germany)

CSF_immunoblot PVDF membrane, 20 µm pores

(Bio-Rad Laboratories, Hercules, CA, USA)

Buffers used for

incubations with

the Abs

INSERM_5% non fat milk; 0.1% Tween20 in 12 mM NaHPO2; 150 mM

NaCl; pH 7.4

CSF_4% non fat milk; 0.1% Tween20 in 20 mM Tris base; 500 mM NaCl;

pH 7,5

MATERIALS AND METHODS

50

10) Immunohistochemistry

a) Rat brains dissected at either developmental stage were immediately snap-frozen,

and 10 µm-thick slices were cryo-sectionned. b) Sections were fixed at room temperature and

blocked for 1 hr at RT in blocking solution. c) Primary antibodies were added at the

appropriate concentration of in blocking solution and incubated overnight at 4°C. d) After

washing, the secondary anti-mouse or anti-rabbit antibodies were used in 0.3% Triton-

containing PBS for 2 hrs. e) the nuclei were stained with DAPI (0,1 µg/ml) in PBS for 10 min

at room temperature, then mounted.

Fluorescent immunoreactions were analysed by fluorescent microscopy using

INSERM_an Axioplan II microscope Zeiss (Le Pecq, France), equipped with the Digital

Camera F-View II and the software AnalySISauto, Olympus, Soft imaging system,

Münster, Germany in France.

CSF_a Nikon Eclipse TS100, equipped with the Digital Camera Nikon ELWD 0.3

(Nikon corporation, Kaganawa, Japan)

Charlotte Schmitt, at the INSERM, did the double staining for the localization of Pgp and

laminin on micro vessels. To detect the laminin 12.5 µg/ml of primary antibody was used, in

addition to the mentionned C219 Ab. For laminin, the slices were pre-treated with

hyaluronidase (4 U/ml, Sigma) at 37 °C, for 30 minutes before addition of the antibodies.

Solutions and instrumentations used for the described procedure (10).

Both INSERM and CSF

Fix PFA 1% on PBS for 30’’ (Pgp) and PFA 4% on PBS for 10 min (Mrp1) at RT.

Blocking 5% BSA, 5% NGS, 0.3% Triton in PBS

First Ab Final concentration 6.025 µg/ml (C219) or 1 µg/ml (A23) (same antibodies

than in Western blot) in Blocking solution

Second Ab INSERM and CSF_ both anti Rabbit and anti Mouse Alexa Fuor 488

(Moecular Probe), both 2 µg/ml

Mounting INSERM_ Fluoroprep (Biomerieux, Marcy l'Etoile, France)

CSF_ Vectashield hardest mounting medium (Vector Laboratories)

MATERIALS AND METHODS

51

11) Total RNA extraction

The CPs from each anatomical district of both Jj and jj Gunn rats were directly

collected in TriReagent (Sigma-Aldrich, Missouri, USA. T9424). The total RNA was

extracted according to manufacture’s instructions, briefly:

a) The samples were collected in 1 mL of tri-reagent b) and 200 μL of CHL were added, then

the samples were c) shacked for 15’’and d) incubated for 3 minutes at RT. e) The

homogenates were centrifuged at 10500rpm for 15’ at 4°C. f) The RNA, contained in the

upper aqueous phase was transferred in a cleaned tube and g) 500 μL of iospropanole were

added. h) The tube was gently inverted then incubated 10’ at RT. i) after 15’ at 10500rpm at

4°C centrifugation the l) pellet was washed with 1 mL of EtOH 75% and m) centrifuged again

at 10000rpm for 5’ at 4°C. n) Finally the supernatant was discarded and the pellet suspended

RNAse free water and stored at -80°C until analysis. The total RNA concentration and the

purity were assed by spectrophotometric analysis in a Beckman DU640. For each sample the

A260/A280 ratio comprised between 1.8 and 2.0 was considered as good RNA quality criteria.

The integrity of RNA was assed on standard 1% agarose/formaldehyde gel.

12) RNA retro transcription

Total RNA (1µg) was reverse-transcribed using the iScriptTMcDNA Synthesis kit

(Bio-Rad Laboratories, Hercules, CA, USA) according to manufacturer’s instructions. RT

was performed in a Thermal Cycler (Gene Amp PCR System 2400, Perkin–Elmer, Boston,

MA, USA) in agreement with the reaction protocol proposed by the manufacturer.

13) Real Time PCR

The Real Time quantitative PCR was performed in an i-Cycler IQ (Bio-Rad

Laboratories, Hercules, CA, USA). 18S, β-actin and GAPDH were used as housekeeping

genes.

PCR amplification was carried out in 25 µL reaction volume containing 25 ng of cDNA, 1 x

iQ SYBR Green Supermix (100 mM KCl; 40 mM Tris–HCl, pH 8.4; 0.4 mM each dNTP; 50

U/ml iTaq DNA polymerase; 6 mM MgCl2; SYBR Green I; 20 nM fluorescein; and

stabilizers) (Bio-Rad Laboratories, Hercules, CA, USA) and 250 nM gene specific sense and

MATERIALS AND METHODS

52

anti-sense primers and 100 nM primers for 18S. Standard curves using a “calibrator” were

prepared for each target and reference gene. In order to verify the specificity of the

amplification, a melt-curve analysis was performed, immediately after the amplification

protocol. Non-specific products of PCR were not found in any case. The relative

quantification was made using the Genex pre-settled excel sheet (Bio-Rad Laboratories,

Hercules, CA, USA), taking into account the efficiencies of individual genes. The results

were normalized to the house keeping genes and the initial amount of the template of each

sample was determined as relative expression versus one of the samples chosen as reference

(Mrp1: 4thV CP Jj), which is considered the 1x sample.

gene Primer sequence: forward Primer sequence: reverse Ampl. Length

Mrp1 ATGGTGTCAGTGGTTTAGG TGTGGGAAGAAGAGTTGC 111

Mdr1a GAGGGCGAGGTCAGTATC AATCATAGGCATTGGCTTCC 193

Mdr1b ACGGAACAGCAGACAAGAAC CGAACACCAGCATCAGGAG 189

Β-actin ATCGGCAATGAGCGGTTCC AGCACTGTGTTGGCATAGAGG 149

18S TAACCCGTTGAACCCCATT CCATCCAATCGGTAGTAGCG 150

GAPDH CCATCACCATCTTCCAGGAG CCTGCTTCACCACCTTCTTG 576

The primer pairs used were synthesized by Sigma Genosys Ltd., (London Road,

Pampisford, Cambridgeshire, CB2 4EF, UK) and were designed using the software Beacon

Designer 4.0 (PREMIER Biosoft International, Palo Alto, CA, USA).

14) Total Bilirubin in Serum (TBS) and albumin measure

Before the sacrifice, blood samples were token from a jugular puncture of each animal

used for the micro vessel preparation procedure. a) The blood collected in heparinized tubes,

b) was centrifuged at RT at 2500 rpm for 20’ to pellet the corpusolate part of the blood. c)

The supernatant was carefully collected, avoiding contamination by haemoglobin and

immediately freeze at –20°C until use. The samples were protected by the direct light to avoid

the bilirubin photo-degradation. The total serum bilirubin (TBS) and the total albumin content

in blood were quantified by routine hospital Kits and machinery (Boehringer-Mannheim Kit

1552414, Monza, Italy, based on the diazo-method, and Alb-Plus 11970909216 Roche-

MATERIALS AND METHODS

53

Hitachi, respectively (La Roche Ltd, Grenzacherstrasse, Switzerland)). The analysis were

performed by the Dr. Alan Rasini, within the collaboration of the Laboratorio Analisi dell’

Ospedale Infantile Burlo Garofalo Trieste, director Prof. Parco.

15) The calculated free bilirubin (cBf) in blood

As a routine method to detect the free bilirubin in the blood still not exist, and the Bf

seems to correlate better with BIND and kernicterus, we tried to calculate the amount of free

bilirubin in the blood of our animals. To do that, we used the following formula and the

published bilirubin-albumin in vivo affinity constants from Gunn rat pups.

This equation assumes independent binding of UCB at the to two sites on albumin: k1 and k2

are the binding constants for the first and second sites, respectively, with k1 given by the

binding constant values defined above and k2 equal to k1/15 (Ostrow et al., 2003b; Doumas

et al., 1985b; Doumas et al., 1985a; Ahlfors, 2001; Ahlfors and Shapiro, 2001).

In formula: TBB=Total Bilirubin in Serum (TBS; μM); Bf=free biliubin (μM); A=total

albumin in serum (μM); K1=affinity constant for the binding of bilirubin in the first site of the

albumin (33.6 μM-1), and K2 the affinity for the second site (1/15 of K1). The final equation as

been resolved by the Internet 3rd degree polynomial calculator

(http://utenti.quipo.it/base5/numeri/equasolutore.htm).

16) Statistical analysis

All the statistical analysis were performed by Excel, performing a paired T.test.

RESULTS

54

RESULTS

In the chapter the results are expressed following the tasks.

I) The Sprague-Dawley rat: the physiological model.

This first task of my thesis work is divided in two parts. In the first part, the relative

protein expression of the Mrp1 and Pgp transporters at the two major blood-brain interfaces

(BBI) has been quantified (Ia). In the second part, the Mrp1 and Pgp protein relative

expression and cellular localisation in the Blood-CerebroSpinal Fluid Barrier (BCSFB) and in

the Blood Brain Barrier (BBB), respectively, have been followed during the post-natal

development of the rats (Ib). The work has been performed on the Sprague-Dawley rats, a

classical laboratory animal, extensively used in research. I have done all the work here

described at the U842 (ex U433) of the INSERM (Institute Nationale pour la Santé Et la

Recherche Mèdicale) at the Faculté Laennec, during the 10-month stage during the DMM

PhD, I have spent in France.

Ia) Mrp1 and Pgp expression on adult brain barriers.

The brains excised from anestitized Adult animals (60 days post natal aged, P60) were

rapidly placed at 4°C in Krebbs Ringher buffer. The CPs from the LV and from the 4thV CP

were collected and pooled maintaining the separation in anatomical districts in order to

evaluate a possible difference between them. The micro vessels (MV) were isolated from

brain cortices cleaned from all apparent meninges. During the isolation of the micro vessels, a

sample of meninges free cortex was collected (Cx). In Fig 9, a typical preparation of the blood

brain barrier isolated from adult rats is showed. The choroid plexuses from the lateral

ventricles (A, left) are a thin, transparent structure containing the fenestrated blood vessels. In

the CPs dissected from the 4th ventricle (A, right), the fenestrated vessels are contained in a

more convolute, Y-like, structure that lines the ventricle and the overhanging cerebellum.

The quality of the MV preparation was assessed visually by phase contrast microscopy

(B). The preparation is formed prevalently by long, branched capillaries and micro vessels (5-

7 μm in diameter), as the elongated morphology of occasional red blood cells in the lumen of

the micro vessels confirms. No cellular debris, tissue and meninges remnants, or larger

vessels contaminants passed trough the sieves are presents. The samples were kept frozen at –

80 °C until use.

RESULTS

55

μmol

.min

-1.m

g pr

otei

n-1

To judge the MV enrichment of the preparations, we performed the γGT test. The γGT

enzymatic activity is strongly present in micro vessels and choroids plexuses, but not in larger

vessels (both arterioles and venules), in any vessels of the pia madre or in parenchyma

(Lawrenson et al., 1999). The γGT activity in the Cx samples (3.31±0.48 μmol.min-1.mg

protein-1) collected during the micro vessels isolation procedure, and the final MV

preparations (44.79±5.97 μmol.min-1.mg protein-1), was evaluated by spectrophotometer test

as described by Lojda (Lojda, 1981), and expressed as mean ± S.D. of more than 4

preparations (P ≤ 0.0007, paired T.test). The mean score we obtained for the MVs-Cx

preparation used in this work was 13.53 (C).

All together these data underline the quality of the adult preparations we prepared, also

if the presence of end-feets astrocytic processes remnants, strongly attached to the endothelial

membrane, couldn’t be excluded.

Fig 9: Characterization of blood brain interfaces isolated from adult rat. Choroid plexuses dissected from the lateral and forth ventricle (LV CP and 4thV CP) of adult rat, scale bar 1 mm (A). Adult microvessels (MV) preparation (B). The sample is mainly formed by capillaries and microvessels, with a diameter of 6 μm, approximately. The enrichment of the final MV preparation was jugged by the γGT activity (C). The 13.5 fold in the MV/Cx enzymatic activity, is indicative of a good final preparation.

100 μm

A

B

C

0

5

10

15

20

25

30

35

40

45

50

55γGT activity

Cx MV

LV 4thV

RESULTS

56

To analyse the relative protein expression of Mrp1 and Pgp in the brain preparations,

the samples were homogenised and the total protein amount determined.

The whole homogenates of 4th ventricle, lateral ventricle choroid plexuses (4thV CP, LV CP),

isolated micro vessels (MV) and cortices (Cx) preparations were used to evaluate the presence

of Mrp1 and Pgp transporters by Western blot with the anti-Mrp1 A23 poliyclonal antibody

(pAb) and the anti Pgp monoclonal antibody (mAb) C219, respectively.

The specificity of the antibodies was first checked running the samples and the

appropriate controls on 7% polyacrilamide gels (Fig 10, A and B, respectively).

The A23 Ab (anti-Mrp1) has been developed by Fernetti in our laboratory (Fernetti et al.,

2001). This Ab, produced in rabbit, reacts with both human and rodent Mrp1, including rats,

and recognises a C-terminal cytopasmatic epitope. The A23 Ab displays a high sensitivity for

rat samples, recognizing a major band at the expected molecular weight of 190 KDa.

Occasionally a faint band of about 60 KDa in MW was detected. This band likely corresponds

to a Mrp1 cleavage product as suggested from Fernetti (Fernetti et al., 2001) and corroborate

from the increasing intensity of the signal with freeze-thaw cycles (Fig 10, A). Additional

demonstration of the specificity of the signal was obtained by the use of a different anti-Mrp1

Ab, the Mrpr1. This Ab, prepared in rat, detected several additional bands, probably due to

the specie-cross reactions (data not shown).

The C219 Ab (anti-Pgp) was chosen because to date it is the only Ab guarante to react

with rodent Pgp, and the most utilized in the literature, allowing a comparison of the results

with those previously published. The C219 Ab recognises a repeated epitope, presents in the

central part and in the C- terminal of the transporters (Georges et al., 1990; Trambas et al.,

2001), allowing an easily detection of all potential cleavage products, additionally to the full-

length protein. The Pgp transporter was recognized as a single detectable band at 170 KDa, as

the expected molecular weigh. No apparent degradation occurred during the MV isolation

procedure (Fig 10, B).

RESULTS

57

By qualitative analysis, the Mrp1 and Pgp expression in the rat brain samples showed

a different pattern (Fig 11). Mrp1 was detected in all preparation investigated, albeit with

apparent difference in the amount of the protein expression. 2.5 μg/well of both CPs total

protein homogenate loaded on gels, produced a signal with an apparent intensity similar to 15

μg/well of MV and Cx total protein lysate (A).

Fig 10: Typical A23 (α-Mrp1) and C219 (α-Pgp) antibodies signals. The anti-Mrp1 A23 antibody detects a major 190 KDaband (choroidal tissue). Occasionally (such as shownhere) a faint 60 KDa band, corresponding to a cleavageproduct of Mrp1, is also observed (A). The C219 (antiPgp) antibody, detects a single and strong band of 170KDa on microvessels preparation (B). 4thV CP: Forth Ventricle Choroid Plexuses; MV:MicroVessels.

MV 2,5 μg

Pgp Mrp1

4thV CP 2,5 μg

A B

250

150

100

75

KDa

LV CP 4thV CP MV Cx Cx MV VL CP 4thL CP

Mrp1 Pgp

2.5 2.5 15 15 μg 2.5 30 50 50

KDa

Mrp1 Pgp

A B

Fig 11: Qualitative Western blot analysis of Mrp1 and Pgp expression in rat bloodbrain interfaces. The Mrp1 signal using the A23 antibody (A) gives strong bands on each sample, but theamount of total lysate loaded strongly differs (both CPs 2.5 �g/well, and MV and Cx, 15μg/well). The Pgp signal with the C219 antibody (B) shows a strong band on MV and Cx(2.5 and 30 μg/well, respectively), but no bands at the appropriate molecular weight could bedetected on CPs samples, respect to the high amount of protein loaded. LV CP : Lateral ventricle Choroid Plexuses; 4thV CP : Forth Venrticle Choroid Plexuses;MV: MicroVessels; Cx: Cortex.

RESULTS

58

Seldom the A23 antibody detected a 120 KDa band in Cx and MV samples. Using the

mouse Mrp1 KO and WT cortex homogenate, we demonstrate that this band is not a Mrp1

cleavage product. In fact the 120 KDa band is well detected in all samples loaded, wile the

Mrp1 190KDa specific signal is present in only in rat and WT mice, but disappears in the

Mrp1 KO sample (Fig 12).

The Pgp signal on 2.5 μg/well of MV, was apparently similar to the signal generated

from 30 μg/well on cortex sample by conventional chemiluminescence. No signal was

detected on CPs samples, despite the larger amount of proteins loaded (30 μg/well) (Fig 11,

B). Only by the use of high sensitive chemiluminescent kits, higher amount of CPs

homogenate loaded, and long exposure time, we were able to detect a faint band at 170 KDa.

Surprising, on CPs samples, the western blot for Pgp shown ad additional band at 100KDa.

This band on CPs was due to a specific cross-reactivity of the secondary antibodies, as

demonstrated by exposing the membrane at the secondary antibody alone. At least 4

secondary antibodies were tested (not shown).

Mouse Rat WT Mrp1-/- KDa 190 150 116

Fig 12: A23 staining for Mrp1 on ratand WT and Mrp1-/- mouse Cx. Wile the Mrp1 specific band is presentonly on rat and WT mouse, the A23 detecta band of about 120 KDa in all samples.The Western blot demonstrates that the 120KDa band is not related to Mrp1.

RESULTS

59

While the qualitative differences between barriers on Mrp1 and Pgp expression was

postulated in literature, our first goal was the quantification of the transporters expression in

barriers and parenchyma.

Between all samples analysed, the CPs and the MV displayed the maximal signal for

Mrp1 and Pgp, respectively, and were chosen as reference sample. The quantification of the

relative expression of the transporters on samples was performed by traditional

chemiluminescence.

To over-come the limit due to the no linear relationship between the densitometry

signal on X-Ray film and the amount of protein of interest, we developed a quantification

procedure based on standard curve (for details, see in Material and Methods).

On every gel a serial dilution of the reference sample was loaded, and the concentration of the

first and second antibodies were chosen to obtain an efficiency curve as linear as possible

(Fig 13). To normalize the expression of ABCs transporters, we used the actin signal obtained

on the same gel. After revealing by chemiluminescence, and densitometry signal collection,

the relative expression of the targeted protein (Mrp1 and Pgp) was calculated as interpolation

from a non-linear regression curve calculated on the standard sample (Fig 13), and normalized

for the actin signal, processed in the same way.

μg/well

Fig 13: Standard curve derived from the optical density profile analysis of the ref. sample serial dilution. For each Western blot the relative expression of Pgp and Mrp1 is established usinga standard curve generated from a reference sample (capillary for Pgp and 4thVCPfor Mrp1) from the same batch. The curve here reported is an example of thetypical non-linear regression curve obtained by Curve Expert analysis of the opticaldensity profile evaluation (ImageQ) of the signal.

RESULTS

60

The result of the Mrp1 and Pgp expression in adult animals are exposed in Fig 14 and

Fig 15, respectively. All the results are representative as mean ± S.D. of 5 Western blot,

performed on different batch of samples, each obtained by a pool of 4 animals. As shown in

Fig 14, Mrp1is strongly expressed in the CPs with a surprising difference between the LV and

4thV CP. The 4thV CP (Ref.Sample) shows the highest Mrp1 relative expression while in the

LV CP the Mrp1 amount is of 60.21 ± 6.73 % in respect to the 4thV CP (statistically relevant

P ≤ 0.01, paired T.test). In MV and Cx, Mrp1 is less present, respectively the 5.18 ± 1.87 and

4.28 ± 1.69 % of the Mrp1 amount in the 4th V CP (both statistically different from the 4thV

CP: Ref.Sample, P ≤ 0.001, paired T.test). This similar value between MV and Cx suggest

that Mrp1 expression is not exclusive of the endothelial cells forming the MVs, but is

expressed in others parenchyma cells, such as astrocytes.

Fig 14: Mrp1 relative expression on adult rat. The results are expressed as mean ± S.D. of 5western blots, performed on different batch ofsamples. The relative amount of Mrp1 in samples isexpressed as % of the amount of Mrp1 in the Ref.Sample (4thV CP, arrowed). Abbreviations:Microvessels (MV) ; Cortices (Cx); Lateral VentricleChoroid Plexuses (LV CP) and Forth VentricleChoroid Plexuses (4thV CP). P≤0.05 (*); P≤0.01(**)and P≤0.001 (***); in paired T.Ttest. 0

10

20

30

40

50

60

70

80

90

100

th

(Ref.Samlpe)

5% 4%

60%

**

***

MV Cx LV CP 4 V CP

RESULTS

61

By contrast, the relative expression of Pgp was highest in the micro vessels

preparation (MV: Ref.Sample) (Fig 15). In the cortex, the amount of Pgp is of 6.84 ± 1.82 %

of the MV (statistically different from the Ref.Sample, P ≤ 0.001, paired T.test) and in CPs

the signal is quite not quantifiable, less than 0,5% of MVs.

The higher score for the MV/Cx ratio in Pgp expression (14.62) compared to the γGT

MV specific enzyme activity (13.53 ratio) (see Fig 10), indicates that the transportre is

strongly if not exclusively expressed by the micro vessels.

0

10

20

30

40

50

60

70

80

90

100

Pgp

Rel

.Exp

r (%

) Ref

.Sam

MV Cx LV CP 4thV CP (Ref.Samlpe)

Both ? 0.5%

***

6.7%

≤

Fig 15:Pgp relative expression on adult rat. The results are expressed as mean ± S.D. of 5 western blotseach performed on different batch of samples. The Pgp relative expression is calculated as % of the amount ofthe Pgp in the reference sample (MV, arrowed). Abbreviations:Microvessels (MV) ; Cortices (Cx); Lateral Ventricle ChoroidPlexuses (LV CP) and Forth Ventricle Choroid Plexuses (4thVCP). P≤0.05 (*); P≤0.01 (**) and P≤0.001 (***); in pairedT.Ttest.

RESULTS

62

Ib) Mrp1 and Pgp expression on rat brain barriers during the post-natal development

The second goal focused on the Mrp1 and Pgp post-natal developmental expression at

the BBI. Due to the minor dimensions, the samples were obtained pooling together the CPs

(VL CP and 4thV CP separately) dissected from 8 and 6 animals at 2-days and 9-days post

natal age, respectively, and subjected to the Mrp1 quantification as described before.

The Mrp1 developmental profile on CPs is shown in Fig 16. The Mrp1 relative

expression is higher in both CPs, with values similar to adult level since the earliest stage of

post-natal development. In LV CP a decreasing trend from higher value at the birth to lower,

adult levels seems to be present (P2: 114.9±29.16; P9: 98.11±42.84; P60: 60.21±6.73 % of

the Ref.Sample: 4thVCP adult, arrowed), also if the high standard deviation, claims caution in

the conclusions. In the 4thV CP, the relative expression is the 73.21±18.39 at P2 and

132.82±32.53% at P9, of the Ref.Sample. No statistical differences were present between the

tree developmental stages.

(Ref.Sample) post-natal age

Fig 16: Mrp1 relative expression in the LV CP and the 4thV CP in the post-natal development ofthe rat. Mrp1 expression in lateral (left) or 4th (right) ventricle CPs from the P2, P9 and adult (P60) rats. Theresults are expressed as percentage of the level in adult 4th ventricle CP (Ref.Sample, arrowed). Dataare expressed as mean ± S.D: of 4 western blots, from different batches of animals.

LV CP 4thV CP

0

20

40

60

80

100

120

140

160

180

P2 P9 P60 P2 P9 P60

RESULTS

63

While the dissection of CPs from brain could be performed at every age by the same

procedure than in adults, a basal lamina is a pre-requisite for the maintaining of the capillary

integrity during the isolation procedure. The micro vessels develop during the post natal

period, and the end of the “sprouting period” (P9-10) signs the early developmental stage in

witch a mature basal lamina is present (Caley and Maxwell, 1970; Simionescu et al., 1988),

and the first stage at witch MVs could be isolated with adequate purity and reproducibility.

The procedure used to isolate the MVs from the younger animals (P9) that we settled up, was

similar to the protocol used for adult animals (for details see M&M), except for the minor

dimensions of the cortices, needing to pool together 8 animals.

A typical preparation of MV isolated from P9 animals is shown in Fig 17. As

compared to the adult preparation (P60), the younger micro vessels appear shorter and larger

than in adults. Additionally, the preparation appears more heterogeneous, with few capillaries

than in the adult, and more micro vessels, probably reflecting the heterogeneity in cell

thickness and basal lamina development still observed at this early stage. No evident

contaminations are present, indicating a good isolation procedure set-up.

By the γGT activity test performed on P9 MV and Cx (B, left panel), a 7.8-fold

enrichment was observed (MV: 9.84±2.65; Cx : 1.26±0.12; P≤0.004, paired T.test). As this

was the first time that MV isolation procedure from a non-adult animal was performed, no

published value was available as reference. The different scale bar (compare left and right

panel in Fig 17 B) suggested a lower γGT activity in early postnatal stage when than in adult

preparations (Rehman et al., 2004).

To confirm this hypothesis we performed a γGT enzymatic activity directly on brain

slices (Satoh et al., 2005) (C). A lower in situ γGT activity associated with capillaries in 9-

day-old rats was detected; wile in the adult the γGT staining was strongly present on smaller

vessels and in CPs (compare left and right panel on C). (Satoh et al., 2005). Additionally, by

the use of γGT inhibitor acevicine we confirmed the specificity of both spectrophotometer and

enzymatic hystochementry test.

We suggest that the Cx/Mv γGT activity ratio we measured may be considered a good

enrichment parameter for judge the final micro vessels isolation preparations at P9.

RESULTS

64

Cx MV

For the relative quantification of Mrp1 and Pgp expression in post-natal developing

isolated MVs (Fig 18), we used an identical procedure used in the adult samples. The

reference sample, in this case, was the MV isolated from adult animals for both Mrp1 and Pgp

transporters. The post-natal developmental profile of relative expression indicates that the

100 μm

P60 P9

100 μm

A

P60

B

0

5

10

15

20

2530

35

40

45

50

55 γgt activity

�m

ol.m

in-1

.mg

prot

ein-1

0123456789

10111213 γgt activity

μmol

min

-1m

gpr

otei

n-1

P9 Cx MV

C

±

P60 ±

*±

P9 ±

Fig 17: Caracterisation om Microvessels (MV) isolated from P9 rats. The phase contrast microscopy of microvessel isolated from P9 rats, shows a major heterogeneity inthis preparation, in which capillaries, microvessels and some larger vessel are present (A). The γGTactivity (B) is lower than in adult preparation. In the enzymo-histochemical detection of γGT (C), thedifference in activity between the two ages is corroborated, with a quite total absence of the stainingon the P9 slice (±), and several MV detected on the adult preparation (±). By contrast, the signalassociated with the CPs is strong in both preparations and was almost completely abolished byaddition of 100 µM and 2 mM acivicin, respectively . The large vessels of the brain do not posses adetectable γGT activity.(*). Scale bar in A: 100 μm; in C: 200 µm.

RESULTS

65

Mrp1 amount on MV was slightly higher at P9 than on adult preparations (121.0±20.3% of

the adult MV, Ref.Sample) (A). It is important to underline that the reference sample

reported in the Fig 18, for the analysis of Mrp1 expression on BBB during development, is the

Mrp1 expression on adult MVs, not on adult 4thV CP, due to the high difference in Mrp1

expression between these two samples. When the Mrp1 expression in MVs at P9 is compared

to the Mrp1 expression on the CPs, the value is about below the 6% of the adult, confirming

the highest Mrp1amount in CPs also during the post-nalal development. By contrast (B) there

is a 4.6 fold increase in the Pgp relative expression between the isolated P9 micro vessels

(21.95±6.76% of the adult MV, Ref.Sample) and adult MVs (statistically different from the

Ref.Sample P ≤ 0.01, paired T.test). This clearly indicates that the Pgp expression develops

post-natally, and in the early post-natal development its activity on brain micro vessels may

be less important.

M V

0

20

40

60

80

100

120

140

160

Mrp

1 R

el.E

xpr.

(%) R

ef.S

a

P9 P60 (Ref.Sample) post-natal age (day)

0

50

100M V

0

1 0

2 0

3 0

4 0

5 0

6 0

7 0

8 0

9 0

1 0 0

Pgp

Rel

.Exp

r. (%

) Ref

.Sam

P9 P60 (Ref.Sample) post-natal age (day)

**

21.95%

A B

Fig 18: Mrp1 and Pgp relative expression on MV during the post-natal development. Mrp1 (A) and Pgp (B) protein levels in capillaries isolated from P9 and adult rat. The result are expressed as mean± S.D. of 3 Western blot, each from different batches of animals. The Mrp1 or Pgp relative expression in samples is expressed as % of the amount of the transporters in the adult preparations (Ref.Sample adult 4thV CP and MV, respectively; arrowed). Statistically different from adult, p<0.01 (paired t-test). The were obtained pooling toghether 8 and 4 animals at 9-days and 60-days old animals respectively

RESULTS

66

To normalize the results, and to control total protein amount loaded on the gels, the

actin signal was used in both adult and developing samples. During the post-natal

development, as indicated in the introducing chapter, the brain, the CPs and the vessels

develop, and are subjected to changes in morphology and volume. The doubt that, during this

development, the actin expression or the proportion of the actin in the total protein lysate

could change at same time was clarified experimentally. Several Western blot performed

confirmed the substantial stability of the actin signal during development, corroborating the

use of this protein also for tithe post-natal developmental analysis.

RESULTS

67

The clear difference in of Mrp1 and Pgp localisation obtained by Western blot was

confirmed by immunofluorescence analysis. The whole brains dissected at either

developmental stage were immediately frozen until use. Brain slices from each post-natal

developmental age were placed on the same glass, allowing processing simultaneously the

samples. The immunofluorescence for detecting Mrp1 is showed in Fig 19. Mrp1 is strongly

expressed in choroid plexuses of both ventricles (A-E). The transporter is located at the

basolateral side of the epithelial cells forming the Blood-CSF-barrier throughout the post-

natal development (A-E), thus at the blood side, as better underlined by the opposite, apical

ezryn signal (G). By these experimental conditions not additional staining on brain

parenchyma or in MV have been detected (stars in A, C, D), probably due to the lower

expression in these samples as revealed in western blot. But a Mrp1 staining has been

detected on isolated-cytospinned micro vessels (Fig 19, R) The Mrp1 localisation at the baso-

lateral side of the CP epithelium suggests that the transporters is functionally mature to export

molecules from the CSF (brain) to the blood since the early post natal age.

Fig 19: Immunohistochemical detection of Mrp1 in rat brain. A,C,D : Mrp1 immunodetection reveals a strong signal associated with CPs (arrows) of all adult (A), 9-day-old(C) and 2-day-old (D) animals. In the experimental conditions used, no signal was detected in the parenchyma(stars). B,E: The magnifications of adult and newborn CPs, respectively, highlight the basolateral localizationof Mrp1, clearly visualised by comparison with the typical apical brush border labelling of ezrin, as shown in2-day-old animal (G). Arrowhead in G: ependymal layer bordering the ventricle. F: negative control, run inabsence of primary antibody, is shown for a 2-day-old animals (arrow: choroid plexus). Nuclei are labelledwith Dapi.

A

*

B C

*

GFD

*

EP60-Mrp1-Dapi P9-Mrp1-Dapi

P2-Mrp1-Dapi

P60-Mrp1-Dapi

P2-Mrp1-Dapi P2-C.Neg-Dapi P2-Ezrin-Dapi

RESULTS

68

At the contrary (Fig 20), the Pgp signal is associated with the micro vessels in all brain

regions examined in both the adult (A) and P9 animals (B). In P2 animals (C), a staining for

Pgp on micro vessels is also presents, suggesting that Pgp is expressed before the complete

maturation basal lamina. No larger vessels (arterioles and venules) are stained. The signal on

P9 and P2 slices (B and C, respectively) is less evident and the vessels seem to be present in

minor density. This could be due to the minor expression of Pgp in vessels at the earliest post-

natal developmental stages, limiting the detection sensibility, or an effective minor density of

MVs, still developing (Caley and Maxwell, 1970; Simionescu et al., 1988).

The micro vessel structure, in witch the apical (blood facing) and basal side are highly

closed (less than 1 μm separates the two membranes), rends less immediate the localisation of

the Pgp transporter on BBB. The Pgp signal appears localized on luminal side of vessels (the

blood side), as revealed from the position of the nucleus with respect to the C219 staining (D,

F), in accord with its function in extruding molecules from the cell in the blood. This finding

is corroborated by the comparison with the ferritin receptor signal (compare D, F with G and

H). The ferritin receptor staining obtained by the Ox26 antibody, appears clearly at both side

of MVs, as the classical localisation reported in literature (Li et al., 1999) (G, H).The final

demonstration of the apical unique localisation of the Pgp staining was finally obtained by

double staining by Charlotte Schmitt (panels L -Q). In the adult MVs, the endothelial cell is

surrounded continuously by the basal lamina, closely contacting the basal side of the MV. In

adult as well as in younger animals, the opposite staining for Pgp and laminin, together with

the nuclei staining, reveals typical luminal localization for the transporter.

RESULTS

69

Fig 20: Pgp immunodetection ob brain slices and isolated MV. The Pgp signal is detectable in all P60 (A), P9 (B) and P2 (C) brain slices. From the early stages of the post-natal development to the adult life, the density of the Mv detected by the C219 antibody, seems increased (From C to D; P2, P9, P60). The Pgp appears localized at the luminal (blood facing) site of the endothelial cell, as showed by comparison with the nuclei staining (D, F: blue=Dapi, green_Pgp) or the transferin receptor (G,H). the latter is well know to localize at both sides. In the cartoon I, is represented a microvessel. The nuclei of the pericytes surrounding the MV, appear emerged and the basal lamina is present at both side. In L-P: double immunostainings of Pgp (red) and laminin (green) highlighting the luminal localization of Pgp in differentiated vessels of developing animals, similar to the capillary staining of adult animals (P). The luminal staining is best seen when the luminal and abluminalmembranes are separated by a nucleus (arrowhead). L, M, N: Pgp, Laminin, merge as typical staining of a capillary froma P2 rat. Merge for Pgp (red) and Laminin (green) as a typical staining of a capillary from a 9-day old animal (P, Q) and P60 animals (O). Nuclei are labelled with Dapi. In R, the signal for Mrp1 is detectable on isolatted-Cytospinned MV from the adult rat.

M L

N

O P Q

P9-Neg.Control-Dapi

P60-Pgp-Dapi

P60-Pgp-Dapi

P9-Pgp-Dapi P2-Pgp-Dapi

P2-Pgp-Dapi

P9-TRrl-Dapi P60-TRr-Dapi

P60-Mrp1-Dapi

P2-Pgp-Laminin-Dapi P2-Pgp-Laminin-Dapi

P2-Pgp-Laminin-Dapi

A B C

D E F

H I G

R

RESULTS

70

Inside the CPs, a diffuse unspecific-like signal is also present (Fig 21). This signal is

attributable to the secondary antibody, as expected from the similar observation in Western

blot.

Additional confirm of the absence of a relevant Pgp expression in CPs and of the

unspecificity of the signal detected by immunofluorescence was obtained loading primary

cultured epithelial cells, together with CPs homogenate and MV as positive control on

Western blot. A Pgp signal was absent in both cultured CP epithelial cells and CPs, but only

in CPs the unspecific band is still present.

The analyses of the ABCs transporters on human brain in physiological situation are,

obviously, rare. During the collaboration between the U842-INSERM and the Service de

Neuropathologie, Hôpital Neurologique centralized via the Biological Resource Center

NeuroBioTec Banques, Lyon, France, we had the opportunity to analyse human choroid

plexuses and isolate the MV from two-autoptical brain samples. The qualitative Western blot

on human BBB and CPs samples (Fig 22), performed by Charlotte Smith, confirmed the

mirroring expression of Pgp and Mrp1 in human blood brain barriers. In human samples,

Mrp1 is higher in CPs and weakly expressed on MV (see the amount of protein homogenate

loaded on western blot, A). The Pgp expression is strongest in MV and barely detectable in

CPs (B). A difference in expression is displayed for the two samples, each coming from a

different subject.

Fig 21: Immunohistochemical detection of Pgp in rat Choroid Plexuses. A diffuse, un-specific signal on CPs was revealed by the anti-Pgp antibody C219 (left). As expected from Western blot results on total lysate CPs, this signal is due to the secondary antibody alone (center). In the right panel, a specific Pgp reconducible signal is present only in the MV sample, nor in a total lysateobtained from choroidal epithelial cells primary culture (CPE) , nor in the total CP lysate (CP). In the latter (firstlane) the ≈ 100KDa cross-reactivity reconducible to the secondary antibody is shown.

Pgp-Dapi C.Neg-Dapi

RESULTS

71

Pgp

CP MV 4thV CP LVCP MV MV LVCP 4thV CP

Rat Human Rat Human

Mrp1 / MRP1

2.5 25 5 5 μg 5 10 30 17.5

KDa

190

KDa170

Fig 22: Western blot analysis of Mrp1/MRP1 and Pgp expression in blood-brain interfaces isolated from human brains. Typical Mrp1/MRP1 signals detected by A23 (left) and Pgp signal detected by C219 (right) using whole homogenates of human CPs and isolated MV are shown. Note the differences in the amount of proteins loaded on each line. Rat preparations are run on the first track of each panel to allow forcomparison. Mrp1/MRP1 is detected as a major band in both types of choroidal tissue, and with a much lower intensity in the microvessel preparation. Pgp is detected as a single and strong band inthe human microvessel preparation, while only a faint band can be visualized at the same molecularweight in both human choroidal preparations. CP: Choroid Plexuses, MV: MicroVessels ; 4thV CP : Forth Venrticle Choroid Plexuses ; LV CP : Lateral ventricle Choroid Plexuses

A B

RESULTS

72

II) The Gunn rat: the bilirubin implications on Mrp1 and Pgp expression on rat BBI.

The second task of my thesis work was performed at the Centro Studi Fegato. The

main aims was to quantify and compare the Mrp1 and Pgp relative protein expression on

blood brain interfaces in the well-established animal model for the hyperbilirubinemia and

kernikterus: the Gunn rat. Different aspects of the pathology have been studied, but, despite

the interest in how bilirubin could enter in the brain and the suggested role of Mrp1 and Pgp

in limiting the influx, no data are available on the expression of these transporters on the

blood brain barriers of the Gunn rats.

Breading the homozygous jj male and heterozygous Jj female litters was produced. To

compare the genotypes, breading was synchronized to obtain a sufficient number of littermate

pooled puppies for each genotype, gender and for post-natal age (P ± 1 day), the collected

samples were analysed simultaneously.

We choose different postnatal ages to analyse the transporters expression on the BBI,

during the progression of the pathology. P2 (2 days after the birth) was selected as the earliest

developmental stage after the birth that allows to distinguishing between rat genotypes. The

bilirubin level in the blood of homozygous mutant jj rats needs 24-48 h to accumulate in

tissue and produce a yellow discoloration of the animals (see Fig 3). Then, P9 was selected as

the first post-natal stage in which we were able to isolate MV with the protocol settled up on

Sprague-Dawley rats at the INSERM. The P17 postnatal stage was chosen as the

developmental stage in which the total serum bilirubin (TBS) picks in the Gunn rat and

corresponds to the developmental stage from witch most of the data contained in literature are

described. Additionally, the P60 postnatal age animals to follow the effects of

hyperbilirubinemia until the adult life were used.

The Gunn rat originated by a spontaneous mutation of a Wistar colony in 1934 (Gunn,

1938). Since the first half of the XX century, different natural strains of Gunn rats diffuse in

the laboratory around the world, where the colonies were maintained by breading brothers and

sisters. In addition news strains have been created by the transfer of the mutation in genetic

background different from the Wistars (Sprague-Dawley, non albuminemic, etc.).

RESULTS

73

For these reasons the serum bilirubin levels differ significantly between different

researches group working with Gunn rats (see in introduction) and for these reasons, our

colony was first characterized by measuring the level of bilirubin in the serum of each

sacrificed animal at every experimental post-natal stage. Additionally the serum albumin

concentration during development was also measured. The TBS and albumin determination

was performed by the Dr. Alan Rasini from the Laboratorio Analisi dell’ Ospedale Burlo

Garofalo, in Trieste.

In jj rats, the activity of the UGT1A1 is absent and the bilirubin in blood (TBS: Total

Bilirubin in Serum) rises due to the loss of conjugation and consequent lack in the

elimination. The TBS, expressed in μmol/liter, in homozygous jj rats is several times higher

than in Jj genotype during all the post-natal developmental period analysed, as shown by Fig

23 and Table 1. At P2, the TBS is about of 230 μM and pick at the 250 μM at P9 in jj

animals, values that were more than 3 times higher that on Jj littermates at the same post-natal

ages. While the total bilirubin amount in serum decreases quickly to normal adult level in Jj

rats (P17: 7.47 ± 3.34 μM; P60: 2.86±0.54 μM) due to the rapid maturation of the UGT1A1

activity still present, the decrease in homozygous rats is very slow and the TBS never reached

the Jj values, remaining around the 80 μM at P60.

TBS - cBf

0

50

100

150

200

250

300

0 10 20 30 40 50 60

post-natal age (day)

μM

0

100

200

300

400

500

600

nM

Jj TBS

jj TBS

Jj CBf

jj CBf

Fig 23: Total Serum Bilirubin (TBS) and calculated free Bilirubin (cBf) The Total Bilirubin in Serum (TBS) on Gunn animalsis expressed in μM. In jellow the TBS on homozygpus jjhyperbilirubinemic rats, in black the heterozygous Jj animals. The calculated Free bilirubin (CBf) is expressed innM. In Red the homozygpus jj hyperbilirubinemic rats, in gray the heterozygous Jj animals. The resultsrepresents the mean ± S.D. of more than 16 animals for each age and genotype.

RESULTS

74

The concentration of albumin in serum is lower at the birth (P2: 150 μM) and

increases during post-natal development (P60: 600 μM) (Fig 24). Between genotypes no

differences are present. In our study, the bilirubin-albumin ratio is significantly higher than

the unit at P2 and P9 post-natal age on jj rat (Table 1, in red). At the later developmental stage

on jj rats, and during all the postnatal developmental period analysed in the Jj rats, the ratio is

lower than the 0.5.

TBS (μM) B/A ratio Calculated Bf (nM)

jj Jj

jj/Jj TBS ratio

Mean Alb (μM) jj Jj jj Jj

jj/Jj TBS ratio

P2-3 235.04±13.09 63.99±3.89 3.67 142.12±30 1.47 0.51 503±0.28 0.17±0.01 29.42

P9-10 249.03±5.07 81.12±2.02 3.07 172.44±5.49 1.47 0.46 255±0.052 0.17±0.004 14.54 P17-18 216.86±26.51 7.47±3.34 29.03 417.39±41.68 0.55 0.017 0.29±0.035 0.05±0.002 53.06

P60 82.33±1.9 2.86±0.54 28.77 594.94±14.62 0.14 0.005 0.04±0.001 0.001±0.0003 32.13

Serum Albumin

0

100

200

300

400

500

600

700

0 10 20 30 40 50 60post-natal age (day)

μM

Jjjjmean

Fig 24: comparison of the blood albumin content in Jj and jj Gunn rats during the post-nataldevelopment. No differences in total serum albumin content (μM) are detected between the heterozygous (Jj) andhomozygous hyperbilirubinemic (jj) Gunn rats. Blood samples were collected before sacrifice. Theresults represents the mean ± S.D. of more than 16 animals for each age and genotype (2, 9, 17 and60 ± 1 days after the birth).

Tab 1: Total Bilirubin, albumin, and calculated free bilirubin in the serum of the jj and Jj Gunn rats. The values and the ratio between jj and Jj TBS are shown (green). By the mean (jj plus Jj of the albumin content inblood, the bilirubin/albumin (B/A) ratio was calculated for each post-natal age. In red, the ratio over the unitshowed by the P2 and P9 jj rats, are underlined (white). The free bilirubin in the blood was calculated from the twoabove mentioned values (TBS and albumin) (yellow).

RESULTS

75

As discussed in the introduction, the free bilirubin, rather than the TBS, seems to

better correlate with the kernicterus. No routine method to measure the Bf in blood is

available. When the stechyometric bilirubin albumin ratio exceeds the unit, the free billirubin

(Bf) increases and diffuses to tissue. By measuring the albumin content in the blood and the

published albumin–bilirubin affinity constant for Gunn rats, we was able to calculate the free

bilirubin in blood (cBf) (For reference and formula, see in Materials and Methods)

As shown in Fig 23 and table 1, while the cBf in Jj rats is below the 17 nM during all

the period analysed, in jj animals the cBf reaches the highest value of 500 nM at P2 (30 times

more than in age-matched Jj rats), decreases to 250 nM at P9 (15 times more) and then,

quickly drops to 30 nM at P17. In the jj and Jj adult animal the values are above the 4 and 0.1

nM, respectively. Thus, concerning the cBf, since P17 the values in jj animals are similar to

those in Jj animals at both P2 and P17 days after the birth, values not sufficient to develop any

apparent dysfunction or damage in the Jj animals.

In the Gunn rat, an evident effect of the high levels of bilirubin is the hypoplasia of the

cerebellum in jj animals. We used this parameter to follow the progression of the pathology.

The weigh of the cerebelli and cortex were calculated as the difference of the tube weigh prior

and after the cerebella collection, divided for the number of animals forming the sample

(g/animal). The samples were obtained pooling together the animals according to genotype,

per age per tissue and per gender. The sex were analysed separately to detect a possible

difference maybe explaining the higher incidence of kernicterus in males (Cannon et al.,

2006).

RESULTS

76

As shows Fig 25, the values, expressed as the means ± S.D., of more than 32 animals

per each post-natal age analysed, reveal the absence of differences in Cx development

between jj and Jj genotypes (A). No sex difference was observed (B). The transient mild-

hyperbuilirubinemia present in Jj rats at P9-10, is not sufficient to produce any macroscopical

effect on CNS development of Gunn animals, when compared with Cx samples dissected

from Wistar rats (not shown).

0,0

0,2

0,4

0,6

0,8

1,0

1,2

0 10 20 30 40 50 60

post-natal age (day)

g/an

ima

F Jj

F jj

M Jj

M jj

Cortex weight

0,0

0,2

0,4

0,6

0,8

1,0

1,2

0 10 20 30 40 50 60

gr/a

nim

jj M+F

Jj M+F

B

A

Fig 25: Post-natal development of the Coretx (Cx) weight on Jj and jj rats. In the upper panel (A) the development of the Cx weight of jj and Jj Gunns rats is compared, in the lower thecomparison has been made by separated gender (B). In both any difference is revealed. The dissected cortices havebeen pooled (per age, per genotype, and per gender). The values are expressed as mean ± S.D. of more than 32animals for each age (A). Colours: Black Jj ; Red jj; Blue: male; Pink: female. Continuous line: jj genotype;sketched lines: Jj genotype.

RESULTS

77

By contrast, the hypoplasia of the cerebellum (Fig 26) is present since P9 in jj

hyperbilirubinemic Gunn rats witch show a 25% weight loss (statistically relevant, P= 0.003

on paired T.test) as compared to the non-hyperbilirubinemic heterozygous controls. The loss

in weight increases by the days, reaching the 37 and 47% at P17 and P60 respectively (A)

(statistically relevant ,both P< 0.001 on paired T.test). When the analysis was performed on

gender pooled samples, no sex specific difference was detected (B). The dramatic hypoplasia

occurring in the cerebelli is well shown by the image (C), representing the whole brains

freshly dissected from a 17 days old Jj and jj Gunn rats. In the image, the yellow discoloration

of the jj brain, still present, is appreciable.

A

0,00

0,05

0,10

0,15

0,20

0,25

0,30

0 10 20 30 40 50 60post-natal age (day)

F Jj

M Jj

F jj

M jj

Cerebelli weight

0,00

0,05

0,10

0,15

0,20

0,25

0,30

0 10 20 30 40 50 60

jj M+F

Jj M+F

B

C

5 mm

jj Jj

Fig 26: Post-natal development ofthe cerebellum (CLL) weight onJj and jj rats. The upper panel shows thedevelopment of CLL hypoplasia inthe jj animals (A). At P9-10 the jjCLL weight is the 75% of the age-matched Jj animals, then decrease atthe 63 and 53% at P17-18 and P60,respectively. The difference in CLLdimensions and brain discolorationat P17, are visible in C. Nodifferences are present when genderpooled samples (age- and genotypematched), are analysed (B). Thevalues are expressed as mean +/-St.Dev. of more than 32 animals foreach age and genotype. P≤0.05 (*);P≤0.001 (**) and P≤0.0001 (***);in T.Ttest.

**

***

***

RESULTS

78

One of the goals of the work performed on Gunn rats was to evaluate if bilirubin is

able to modulate the Mrp1 and Pgp expression on Blood Brain Interfaces.

Based on the indications obtained at the INSERM on Sprague-Dawley rats, the

relative protein expression of Mrp1 was studied in the CPs, and Pgp in isolated MV.

The CPs were dissected from all post-natal developmental age, and the MV were

isolated from P9, P17 and adult (P60) animals.

The protocol for the isolation of MV from P17 animals was settled up at the CSF, by

modification of the protocol used for the P9 animals.

To allow quantification we set up a reference curve by the use of the SHY5Y cell line

in witch the Mrp1 and Pgp expression was first checked. The quantification procedure was

performed as described in the first section. At the end the reference sample were

mathematically changed from the SH to the 4thV CP P60 Jj (for Mrp1) and MV P60 Jj (for

Pgp), to maintain the same way in the results exposition. The relative amount of the

transporter in sample was expressed as means ± S.D. of 3-4 Western blot for each post-natal

developmental age, performed on different samples each formed by several animals pooled

together.

In Fig 27 is showed a typical western blot for hMRP1/rMrp1 and actins, as example.

The antibodies used recognize specifically bands of the expected molecular weight (190 KDa

for Mrp1 and 42 KDa for actins) on both human and rat samples. No unspecific bands were

detected.

RESULTS

79

SHY5Y Jj CP jj CP LV 4thV MW LV 4thV 30 15 5 2.5 5 5 5 5 μg

KDa

190

42

Mrp1 Act

A

S = 0.00000000r = 1.00000000

X Axis (units)

Y A

xis

(uni

ts)

0.1 2.8 5.5 8.3 11.0 13.7 16.410.00

539.43

1068.87

1598.30

2127.73

2657.17

3186.60

DensitometryArbitrary Units

3200

2670

2200

1670

1060

640

02.5 5 15 30 μg

B

Fig 27: Representative western blot for Mrp1 staining on Gunn CPs and SHY5Y cells (Ref.Sample). The relative quantification of the transporters amount (Mrp1 or Pgp, here Mrp1), has been performed running on the same gel a reference sample (SHY5Y) and both genotypes of the samples in analysis (A). Then the relative expression has been calculated from the non-linear regression curves obtained plotting the densitometry scanned values respect the total protein amount of the Ref:Sample loaded in the calibration serial dilution (B). The values have been normalized for the actins signal developed on the lower part of the same gel (A). The panel A, shows a typical staining for Mrp1 detected by the A23 antibody. On both Ref.Sample and CPs samples, the antibody detect a defined, unique band. No un-specific bands, or degradation products are visible.

RESULTS

80

The CPs obtained from Gunn rats dissection not revealed differences between the two

genotypes at every post-natal age studied, by visual examination under stereomicroscope

vision (data not shown).

On LV CP at P2 the Mrp1 relative expression (Ref.sample 4V CP Jj P60) between

genotypes is identical, then the Mrp1 amount in the homozygous hyperbilirubinemic animals

decreases (Fig 28 left). The amount of the decrease between genotypes is around the 25%,

33% and 30% of the age matched counterpart at P9, P17 and P60, respectively, reaching the

statistical relevance from P17 (both P17 and P60, P≤0.02, paired T.test). In both genotypes no

statistical developmental differences are detectable (left A). The analysis of gender specific

samples, revealed an identical pattern of expression for Mrp1 between males and females (left

B).

In the 4thV CP (Fig 28 right) the reduction on the Mrp1 expression in

hyperbilirubinemic rats is higher and is precocious. While the developmental Mrp1relative

expression on Jj rats feebly increases reaching the adult level at P17, in jj the decrease is rapid

and dramatic. Since P9 the Mrp1 amount in the jj rats is about the 50% of the age matched

normobilirubinemic animals (statistically relevant, P≤ 0.01 on paired T.test) and stays around

the 50% of the value observed in the Jj control also at P17 and P60 (statistically relevant, P≤

0.001 and 0.01, respectively) (Fig 28 right A). No gender differences were detected, as

showed in Fig 28 right B.

The comparison of the Mrp1 relative expression in the same genotype (compare Fig

28 left and right), shows a significant difference between CPs from the 4th and the lateral

ventricles in Jj animals at P60, when the LV CP Mrp1 amount is around the 70% of that

measured in the 4thV CP (statistically relevant, P<0.05 on paired T.test).

As discussed below, the relative amount of the transporters in samples has been made

vs. the SHY5Y cells, and then mathematically changed in order to refer the values in a similar

way with the Srague-Dawley report. By this way we shown a 7% of variability between the

different 4thV CP batch dissected from the Jj P60 animals.

RESULTS

81

Fig 28: Mrp1 relative expression in the LV and 4thV CP of the Jj and jj Gunn rats. The results are expressed as mean ± S.D. of more than 4 Western blot par age, from different samples (A). Thesamples were obtained pooling (par tissues, par age, par gender) the CPs dissected from 3, 4, 3 and 4 animalsfor each post-natal age (P2, P9, P17 and P60 ± 1 day, respectively). The Mrp1 relative expression on samplesis expressed as % of the amount of Mrp1 in the Ref. Sample (P60, Jj, arrowed). In B, the results of genderspecific expression of Mrp1 are shown. P≤0.05 (*); P≤0.01 (**); P≤0.001 (***) in paired T.Test. The per centrepresents the amount of Mrp1 in jj rats respect the age matched Jj genotype.

0

20

40

60

80

100

120

140

P2-3 P9-10 P17-18 P60post-natal age (day)

Mrp

1 R

el.E

xpr.

(%) R

ef.S

am

F Jj

M Jj

F jj

M jj

0

20

40

60

80

100

120

140

P2-3 P9-10 P17-18 P60

post-natal age (day)

Mrp

1 R

el.E

xpr (

%) R

ef.S

am

F Jj

M Jj

F jj

M jj

0

20

40

60

80

100

120

140

P2-3 P9-10 P17-18 P60post-natal age (day)

Mrp

1 R

el. E

xpr.

(

Jj

jj

0

20

40

60

80

100

120

140

P2-3 P9-10 P17-18 P60post-natal age (day)

Mrp

1 R

el. E

xpr.

(

Jj

jj

LV CP 4thV CP

***

***

** * *

52%

45% 50%

75%

67%70%

A

B

RESULTS

82

We investigated the down regulation of Mrp1 by Real-time PCR to understand the

mechanisms involved in this event. The CPs from both LV and 4thV CP, and both genotypes,

were dissected from P9 animals and homogenized in Tri-reagent. The relative expression

(4thV CP Jj ref. Sample) of Mrp1, Mdr1a and 1b (together Pgp) mRNA was then analysed

(Fig 29).

Ct

0 5 10 15 20 25 30 35 40

cicles

GAPDH 18S beta-ActinMrp2 Mrp3 Mdr1bMdr1a Mrp1

Mrp1

Mdr1a

Mdr1b

Beta-actin beta-acin

GAPDH

18S

Strong expression Ct Very low expression

Fig 29: Treshold cycles in real-time PCR for the analysed mRNA (Mrp1, Mdr1a andMdr1b) and house keeping genes (Beta-actin, 18S and GAPDH in CPs. Analysis of the mRNA relative expression of the Mrp1 and Mdr1a/1b (togheter Pgp) genes inthe CPs of the P9 animals (female, each sample 4 animals). By the Ct, the highest abundanceof the Mrp1 mRNA in the CPs is pointed out. Both the Mdr1a and Mdr1b, the two rodentisoforms coding for the correspondent MDR1 in humans, are weakly expressed. For thequantification, tree house keeping genes have been employed, each corresponding to adifferent functions in the cell.

RESULTS

83

The results show that while the Mrp1 protein expression in jj genotype is strongly

down-regulate, the mRNA is not modified (Fig 30). Suggesting a post-transcriptional

mechanism in the protein down regulation of Mrp1 expression.

Mrp1

0,0

0,2

0,4

0,6

0,8

1,0

1,2

1,4

1,6

Jj jj Jj jj

LV CP 4th V CP

Fig 30: Real time PCR analysis of the Mrp1 mRNA. The mRNA relative expression of the Mrp1 shows no significantdifferences between genotypes, indicating that the down-regulation of theMrp1 protein expression is post-trascriptional. For the quantification, the relative abundance of the Mrp1 mRNA, isexpressed as ratio of the Mrp1 mRNA calculated amount in the 4thV CP Jj(=1, reference sample, arrowed) Tree house keeping genes (beta-actin; 18Sand GAPDH), each corresponding to a different functions in the cell, havebeen employed to normalize the results. The results are expressed of mean+/- standard deviation of 2 TR-q-PCR (on different samples).

RESULTS

84

Since we detected a low protein Pgp expression at the CPs of Sprague-Dawley rats,

and the Mdr1a/b mRNA was detected by real time PCR in Gunn rats indicating the presence

of this transporter, we evaluated the Pgp protein relative expression on Gunn CPs (Fig 31).

The CPs are strongly involved on waste potentially toxic compounds from the CNS.

The system plays in both direction, from the tissue to the blood, and via the CSF from the LV,

to the 4thV CP and the cisterna magna, and finally out of the brain. In Gunn rats, the down-

regulation of Mrp1 on the baso-lateral side of the CPs, may be compensated by the up-

regulation of Pgp on the apical side of CPs, in order to waste bilirubin out of the CNS by CSF

flow.

As shown in Fig 31, the Pgp protein expression is barely detectable. At P2 and at P9

post-natal ages, no signal has been detected despite the high amount of total CPs protein

loaded (Fig 32). At P17 only a fain band is detectable and at P60, when Pgp shows the

maximal amount in both jj and Jj CPs, the relative expression was below the 1% of the

reference sample (MV P60 Jj). No differences between CPs or genotypes have been detected.

RESULTS

85

30 15 5 2.5 30 30 MW 30 30 5 ug/well

SHY5Y: St.Curve Lv jj 4jj Lv Jj 4 Jj MV

Pgp

Act

Fig. 32: Western blot for Pgp on Choroid Plexuses samples from jj and Jj Gunn rats. Despite the large amount of choroid plexuses total lysate loaded (30 μg/well), Pgp is barelydetectable. Conversely the in the MV (5 μg/well) the Pgp signal is strong, as well as ispresent in the standard curve.

4thV CP

0,0

0,2

0,4

0,6

0,8

1,0

P2-3 P9-10 P17-18 P60post-natal age (day)

Pgp

% e

xpre

ss.R

ef.s

am

jjJj

LV CP

0,0

0,2

0,4

0,6

0,8

1,0

P2-3 P9-10 P17-18 P60post-natal age (day)

Pgp

% e

xpre

ss. R

ef.s

am

jjJj

N.D. N.D. N.D. N.D.

Fig 31: Pgp relative expression on Jj and jj Gunn rats Choroid Plexuses during the post-nataldevelopment. The results are expressed as mean ± S.D. of 2 Western blot from different samples. The relative expression ofPgp in CPs is reported as % of the amount of Pgp in the Ref. Sample (MV P60, Jj). Albeit the high amount of CPtotal protein loaded in Western blot, the Pgp amount in CPs is very low, and rests under the 1% of the Pgpamount in the MVs (Ref.Sample: MV P60, Jj). Abbreviations: Non detectable (ND).

RESULTS

86

The localization of Mrp1 in Gunn rat brain slices was performed to study a possible

re-allocation of the transporter in hyperbilirubinemic animals (Gennuso et al., 2004).

Cryothome cutted slices from freshly isolated brain at the same post-natal age were

used. On every slice, both genotypes were present, allowing a direct comparison of the Mrp1

staining. A strong Mrp1 signal is detectable on CPs from both anatomical districts (LV and

4thV CP) (Fig 33) and in both genotypes is localized at the baso-lateral side of the epithelial

cells during all post-natal development. No differences between genotypes have been detected

and any signal is present on parenchyma or in the negative control.

RESULTS

87

Fig 33: Immunofluorescence for Mrp1 on Gunn ratbrain slices during the post-natal development. The Mrp1 signal (green) is localized at the baso-lateralside (→, the blood face) of the choroid plexuses of bothJj (A, B, C and D) and jj (E, F, G, and H) genotypes atevery post-natal age examined (from P2 to P60, fromleft to the right). In the parenchyma any signal isrevealed (*) (see in A, B, E, F, G), as well as nobackground is detectable in the negative controlobtained omitting the first antibody (I). The Dapinuclear staining (blue) fails to reveals apoptotic figuresin the CPs. The image L, evidences the position of a LVCP (‡) respect to the ventricle, containing the cerebro-spinal fluid (CSF) (§), the parenchyma (eht dna(ependyma (arrowed) that lines the ventricle cavity,separating the CNS tissue from the CSF.

A B C D

E F G H

I L

P2 Jj Mrp1-Dapi P9 Jj Mrp1-Dapi P17 Jj Mrp1-Dapi P60 Jj Mrp1-Dapi

P2 jj Mrp1-Dapi P9 jj Mrp1-Dapi P17 jj Mrp1-Dapi P60 jj Mrp1-Dapi

RESULTS

88

BBB analysis.

As mentioned below, the MVs have been isolated from Jj and jj Gunn rats by a

protocol similar to the used at the INSERM, maintaining a good quality of the final

preparation as shown in Fig 34. While the Pgp is the marker of MVs, the micro vessels

preparations were used as reference sample for the relative quantification of this transporter.

The results are expressed as means ± S.D. of the values derived from more than 3 Western

blot from each age, performed with different samples, each formed pooling together par age

and gender (8, 6 and 4 animals at P9, 17 and 60, respectively).

P60-Female-jj

P60-Female-JjC

BA

D

Fig 34: Adult MicroVessels freshly isolated from Jj and jj Gunn rats. In phase contrast microscope imaging of microvessels, the MV isolated from adultJj (A, B) and jj (C, D) are shown. The preparation are formed mainly by microvessels; in the panel A, same contamination is visible (↑). The jj preparation is less abundant, but any contamination is present (C). In B and D, magnification of the Jj and jj samples. the smaller diameter of the microvessels is appreciable due tothe elongated form took by the red blood cells inside the MV lumen. Scale bar 100 μm in A, B; 25 μm in B, D.

RESULTS

89

The Pgp expression in jj animals at every postnatal age analysed is up regulated (Fig

35). The up-regulation was already present at P9 when the amount of the transporter in jj rats

is of 195.5% than in the age-matched Jj (statistically different from value obtained in age-

matched Jj rats; P < 0.001), reaching Pgp amount similar to these found in P17 aged Jj

animals. At P17 and P60, the amount of Pgp in the Mv isolated from the hyperbilirubinemic

animals is about of 130-140% respect the age matched Jj rats (at P60 statically different from

value obtained in age-matched Jj rats; P < 0.05).

The Pgp relative expression in both jj and Jj isolated MV increases about of 10 fold

and 5 fold from P9 and P17 to P60, respectively.

MV

0

20

40

60

80

100

120

140

160

P9-10 P17-18 P60post-natal age (day)

(%) P

gp R

el.E

xpr.R

ef. S

am

Jjjj

195.5%

141.3%

131.8%

***

*

Fig. 35: Pgp relative expression on Gunn rat MicroVessels. The results are expressed as mean ± S.D. of 3 Western blot performed ondifferent batch of samples. The Pgp amount in the sample is expressed as% of the Pgp amount in the Ref.Sample, MV P60, arrowed. The %indicated in the figure represents the amount of Pgp in jj rats respevt the Jjage-matched animals. The samples were obtained pooling (par tissues, parage, par gender) the MV isolated from 8, 6 and 4 animals for each post-natal age (P9, P17 and P60 ±1 day, respectively). P≤0.05 (*); P≤0.01 (**);P≤0.001 (***) in paired T.Test.

RESULTS

90

Similarly, the expression of Pgp detected on the whole cortices (Cx) is up-regulated at

every post-natal age analysed. The Fig 36, shows a different pattern in the development as

compared to MV, characterized by a low difference between the tree ages. This pattern is

similar between the two genotypes. However, the amount of Pgp in whole Cx is low (see Fig

37), confirming that Pgp is expressed exclusively in the endothelial cells forming the MVs.

Cx

0

1

2

3

4

5

6

7

P9-10 P17-18 P60post-natal age (day)

(%) P

gp R

el.E

xpr.R

ef. S

am

Jjjj

120.3%

167.2%

141.94%

**

**

Fig. 36: Pgp relative expression on Gunnrat Cortex. The results are expressed as mean ± S.D. of 3Western blot performed on different batch ofsamples. The Pgp amount in the sample isexpressed as % of the Pgp amount in the MVP60 Jj (Ref.Sample). The % indicated in the figure represents theamount of Pgp in jj rats respevt the Jj age-matched animals. The samples were obtainedpooling (par tissues, par age, par gender) theMV isolated from 8, 6 and 4 animals for eachpost-natal age (P9, P17 and P60 ±1 day,respectively). P≤0.05 (*); P≤0.01 (**);P≤0.001 (***) in paired T.Test. Both P9 and P17: statistically different fromvalue obtained in age-matched control,P≤0.01 in paired T.test.

Cx MV Cx MV

Pgp Act S2

KDa

170

42

30 4 30 4 μg

Jj jj

Fig. 37: Typical western blot staining for Pgp onCx and MV of Jj and jj Gunn rats. The Pgp is strongest expressed in the MV of bothgenotypes, and only a faint band could be detectedon the Cx, despite the difference in protein amountloaded. Abbreviations: Cortex (Cx), MicroVessels(MV), Heterozygous Gunn rats (Jj), homozygushyperbilirubinemic Gunn rats (jj).

RESULTS

91

On brain slices (Fig 38), the Pgp signal rests confined to the micro vessels. No signal

was detectable in others brain structures or in negative control. About the localization of Pgp

on micro vessels, the fragility of the tissues, especially in the jj genotype, and the

impossibility to perfuse animals due to the antigen sensibility of the C219 Ab, ender difficult

to define the Pgp localisation on MV, although the Pgp appears to be localized.

P17_jj_Pgp_Dap P60_jj_Pgp_Dap

P17_Nj_Pgp_Dap P60_Nj_Pgp_Dap

Fig 38: Immuno-hysochemistry for Pgp on Gunn rat brain slices. On both genotypes the Pgp staining is specific for the micro vessels. No signal was detected on parenchyma.

DISCUSSION

92

I) SPD

The blood–brain barrier and the blood–cerebrospinal-fluid barrier function together to

isolate the brain from circulating drugs, toxins, and xenobiotics. Both present a large surface

area for the exchanges, due to the extensive network developed by MVs, and the anatomical

organization in numerous villi and apical microvillus-bearing membranes of the specialized

epithelial cells (Keep and Jones, 1990) in BCSFB.

The barrier phenotype derives from the presence of tight junctions (TJ) avoiding the

paracellular diffusion of the molecules, and the presence of transporters located on both blood

or CNS facing sides of the barriers, controlling the passage of the lipophylic molecules, that

could cross the cellular membranes.

Among those, the ATP-binding cassette transporters Pgp and Mrp1 appear to be key

players in preventing access, or increasing elimination of numerous endo- and xenobiotics

from brain (Schinkel and Jonker, 2003; Wijnholds et al., 2000). But there is a limited

knowledge about their distribution and function of at the BBI.

Ia) Mrp1 and Pgp relative expression on adult BBI

By quantitative Western blot, we have demonstrated a “mirroring expression” of those

transporters between the two brain barriers. The choroid plexuses are characterized by a

strong presence of Mrp1 with a difference between the LV and the 4thV CP, the former being

less expressed than the last, on the contrary the Pgp relative protein expression is virtually

absent.

By contrast Pgp is highly expressed in adult endothelial cells forming the MVs, where

the expression of Mrp1 is 15-20 fold lower that in CPs.

The immunofluorescence analysis on brain slices revealed that both ABC proteins are

localized at the blood side: Mrp1 is localized baso-lateral in the CPs (BCSFB). Wile Pgp is

apical on the micro vessels forming the BBB. This is the correct localization to avoid the

entry, or increase the efflux from the brain, of the potentially toxic compounds.

Previously Mrp1, has been detected in CP epithelium (Rao et al., 1999; Wijnholds et

al., 2000) and also in cultured brain endothelial cells and neuronal cell lines. The Mrp1

protein, mRNA and activity, in vitro, have been shown to be up regulated during culture

DISCUSSION

93

(Regina et al., 1998; Sugiyama et al., 2003; Miller et al., 2000). In rat freshly isolate MV,

both Mrp1 mRNA and protein, were detected.

The quantification of Mrp1 protein level and the immunohistochemical data presented

in this thesis shows unambiguously that in the rat the CPs is the main site of blood-facing,

Mrp1-dependent cellular efflux from the brain. Additionally, our results underline, for the

first time, a difference in Mrp1 expression between CPs, which may suggest a different role

for this organ on the two anatomical districts. Usually, the published data are referred to all

CPs pooled together or, when not specified, to the LV CP alone. It is impossible to speculate

about the meaning of the difference we have shown and additional studies are clearly needed.

While Mrp1 seems not to be specifically expressed by endothelial cells forming the

MVs but that other cells, such as astrocytes and oligodentrocytes, may express the transporter

(Gennuso et al., 2004; Mercier et al., 2004).

By contrast, Pgp seems to be expressed in manner almost exclusively on the

endothelial cells, as suggested by the comparison of the MV/Cx ratio for Pgp expression and

γGT test, and the specificity of the immunofluorescence signal for MVs in brain slices.

While the Pgp expression on the capillaries forming the Pgp is well documented

(Thiebaut et al., 1989; Jette et al., 1995a; Regina et al., 1998), its localization is controversial

(Beaulieu et al., 1997; Mercier et al., 2004; Virgintino et al., 2002; Bendayan et al., 2002;

Golden and Pardridge, 1999; Pardridge et al., 1997). The morphology of micro vessels, with

the two leaflets closed, renders difficult to distinguish between the apical and basal

localization. In our work, the double staining for Pgp and laminin indicates clearly an apical

localization for Pgp. In humans and rat, the Pgp protein expression has been reported on CPs,

localized by Immunohistochemisry at the apical side, facing the CSF (Rao et al., 1999). In our

preparations, however, not quantifiable levels of Pgp were detected by Western blot, and any

specific staining for Pgp was present in CPs, indicating that the major sites of Pgp-dependent

brain efflux do not include this barrier, and may explain why clear functional evidence for

Pgp activity at the BCSFB has not been produced.

This specific pattern of Pgp and Mrp1 localization may be related to the characteristics

of their respective substrates, metabolism and environment. Both transport proteins are multi

specific. However, typical Pgp substrates are usually rather lipophilic (Leslie et al., 2005;

Schinkel and Jonker, 2003) and the MV are surrounded by the lipid rich nature of brain

DISCUSSION

94

parenchyma which comprises oligodendrocyte-derived myelin sheathes, and neuronal and

glial cellular membranes. So the main role of Pgp would be to prevent the otherwise likely

accumulation of potentially deleterious lipid soluble toxins into the brain.

Conversely, the choroidal epithelium is directly contacting the CSF, harbors high

levels of Mrp1, and display the higher conjugating and detoxifying activity in the brain

(Strazielle et al., 2004; Strazielle and Ghersi-Egea, 1999; Ghersi-Egea and Strazielle, 2001).

Mrp1 at the blood side of the CPs, may limit the access of blood-borne amphiphilic

compounds, eliminate the conjugates-metabolites, and control the central bioavailability of

endogenous compounds, such as the leukotriene C4 inflammatory mediator, diffusing them

out of the CNS.

Data concerning the ABCs transporters on human brain in physiological situation are,

obviously, rare (Nies et al., 2004). Our results on human BBB and CPs samples, suggest that

the mirroring expression of Pgp and Mrp1 in blood brain interfaces observed in rat, may be

true also for humans. Concerning the major variability in the expression Pgp expression on

human CPs samples, we have to consider that the ABCs transporters are known as multi-drug

resistance transporters involved in resistance against medical compounds. Medical treatments

during the life or before death could modulate the expression.

Ib) The Mrp1 and Pgp expression during the post-natal development of the rat

With respect to the post-natal developmental profile, in our work the Pgp post-natal

expression is referred at the isolated micro vessels. The amount of Pgp increase 5 fold from

the 9-day-old animals compared to the content found in capillaries from adult animals, and by

immunofluorescence a evident signal is present still P2. These results are in agreement with

previous data, obtained in whole brain of mice and rats (Matsuoka et al., 1999; Tsai et al.,

2002). The Pgp staining appears already associated with the luminal membrane, and the

density of the stained MV seems to increase during all the post-natal development. This

observation is corroborated by a study of Caley, that reported as the capillary density is lower

in developing than in adult animals (Caley and Maxwell, 1970).

All together these results suggests that the Pgp mediated protection at the BBB might

not to be so efficient at the early post-natal stage than in adult life.

DISCUSSION

95

With respect to Mrp1, no evaluation of its developmental profile in brain has been

reported so far. The Mrp1 amount in both CPs is high since the early post-natal age. Although

the reproducibility of the data claims to caution, the decreasing trends, suggest a major role of

CPs in the early-embryonic age rather than in adult brain. The results are not fully surprising.

The CPs develops precociously (from E13 in rat), and are involved in the maintenance of the

brain homeostasis and in driving the development of the CNS (Dziegielewska et al., 2001;

Strazielle et al., 2005). Probably a fully functional activity of CPs is needed since the early

stages of the life, and that could comprise also a precocious expression of the CPs associated

machinery (Ghersi-Egea et al., 2006) and transporters. In developing brain, Mrp1 may

participate in transporting some compound involved in the modulation of this structural

development or helping in maintaining the brain homeostasis (redox state, energy) during

these steps. As bilirubin is the physiological Mrp1 substrate with the highest affinity so far

reported (Rigato et al., 2004), our data also raise the interesting hypothesis that this

transporter in CP is involved in neuroprotection against excessive accumulation of bilirubin in

the brain during physiological neonatal jaundice.

DISCUSSION

96

II) The Gunn rat: the bilirubin implications on the Mrp1 and Pgp expression at the BBI

Historically the studies concerning the bilirubin entry into central nervous system have

focused on the blood brain barrier, based on the theory (Diamond and Schmid, 1966; Aono et

al., 1989; Hanko et al., 2003; Takahashi et al., 1984), that only the free unconjugated

bilirubin, the part of UCB exceeding the binding capacity of the serum albumin, was able to

cross the cell membranes (Brodersen, 1980; Ostrow et al., 1994; Zucker et al., 1999; Roca et

al., 2006).

Similarly the strong expression of Pgp at the BBB has been reported as the key factors

in protecting from the bilirubin influx into the brain (Hanko et al., 2003; Watchko et al., 2001;

Watchko et al., 1998).

It was suggested that the high incidence of kericterus in infants, especially in pre-term

newborns, was related to a lower efficiency in the Pgp mediated protection on blood brain

barrier.

Our results in Gunn rats show that the Pgp expression in MV of jj animals is up-

regulated at every post-natal age analysed, but the entity of this up regulation do not seems to

be sufficient to confer protection until P17 when the amount of the transporter in both

genotypes is about 5 times lower than in the adult animals. This post-natal developmental

profile of Pgp expression is in agreement with that previously reported in normal rats

(Matsuoka et al., 1999), and in mice (Tsai et al., 2002).

According to the data, we suggest that the Pgp alone is not able to efficiently protect the CNS

from the bilirubin flow, at least until the adult levels, reported to occurs around the 40 days in

the rat (Matsuoka et al., 1999).

This conclusion is in agreement with the observation that the UCB in the central

nervous system is higher at P10 compared to the P21 (Roger et al., 1996). Interesting, in older

animals (P32-36) traded with displacing agents and/or Pgp inhibitors before bilirubin

infusion, the localization of the maximal H3-bilirubin concentration did not correlate with the

typical brain areas damaged in kernicterus (Hanko et al., 2003), and was not explained by a

different tissue clearance ability (Hansen and Cashore, 1995).

All these experiments have been done in Sprague-Dawley rats in witch the

bilirubinemia was normal until the infusion of the pigment. Although, this is the best model to

study the permeability properties of the brain interfaces, it does not take in account the

DISCUSSION

97

possible effects of a chronic exposition to high bilirubin levels on the blood brain barriers

properties, as in the Gunn rat model.

In fact, we have shown for the first time a strong down regulation of the Mrp1 protein

expression in the blood cerebrospinal fluid barrier. Since P9 the amount of Mrp1 in the jj

BCSFB drops, in the 4thV CP, to the about 50% than in the age matched Jj littermates. In the

LV CP the decrease is less marked, but in any case the Mrp1 expressions in both jj CPs is

strongly impaired.

Also if the contributes of each barrier in limiting the brain inflow is difficult to asses,

we assume that the down regulation of Mrp1 in CPs, strongly impairing the Mrp1 barrier

mediated properties at the BSFB, is overlapped by the absence of Pgp mediated protection at

the immature BBB, contributing in additive manner to the bilirubin entry in the brain. This

might be responsible, at least in part, to the localization of the phenomena.

A positive significant relationship between serum and cerebrospinal fluid bilirubin

concentrations has been reported (Rodriguez Garay and Scremin, 1971). From the CSF, the

bilirubin can easily diffuse into the tissue crossing the ependyma and accumulate in the peri-

ventricular regions. In fact, in kernicterus the more damaged brain regions are located in

proximity of the ventricles, correlating with the exponential decrease in radiolabel bilirubin

content respect to the distance from the ventricular surface in jj Gunn rats reported in H3-

bilirubin ventricular infusion experiments. In the same experiment, the out-flow of the

bilirubin from the CSF was founded significantly impaired in jj Gunn rats (Rodriguez Garay

and Scremin, 1971), suggesting a difference in the outflow and accumulation property in

Gunn animals respect others normal rat strains (Hanko et al., 2003). This may truly be

explained by the Mrp1 strong down-regulation we founded. Conversely no reallocation of the

transporter at the BCSF barrier was evidenced by immunoluorescence.

As we reported, normally the Mrp1 protein amount is maximal in the 4thV CP (100%),

respect the 70% in LV CP of Jj Gunn, and 60% of Sprague-Dawley animals. In jj rats the

decrease of the Mrp1 amount in CPs is strongest in the 4thV CP (~50%) respect the LV CP

(~30%). One possible explanation point out on the anatomy of the ventricular system. The 4th

ventricle is at the end of the ventricular cavities and is possible that the bilirubin here

accumulates prior to sort from the CNS, reaching their maximal amount and displaying the

maximal effect.

DISCUSSION

98

By the same reasoning also the maximal bilirubin concentration founded in the

cerebellum (Diamond and Schmid, 1966) of 18.9 μg bilirubin/gr tissue respect the 10.7 and

3.0 μg/g tissue in the brainstem and cortex (Cannon et al., 2006) may be explained.

In the experiments performed by Cannon, the total bilirubin content in the brain was

higher on jj animals respect to the Jj counterparts, due to the different TBS (7.3 vs 0.15

mg/dL; ratio 48.7). In the latter animals (Jj normo bilirubinemic animals) no difference in

bilirubin concentrations between brain regions was founded, either in saline nor in displacing

agents treated animals, differently from the jj genotype (Cannon et al., 2006). This claims to

the concept of the “free bilirubin”, considered a better predictor parameter in bilirubin

induced neurological dysfunction and kernicterus, respect the TBS (Ahlfors and Shapiro,

2001; Ahlfors, 2001; Ostrow et al., 2003b; Calligaris et al., 2007).

The maximal inflow of bilirubin is present in animals treated with displacing agents

(Cannon et al., 2006), at every post-natal age examined, and this effect is greater than in

animals treated with Pgp and/or Mrp1 inhibitors alone (Hanko et al., 2003). Roger argue that

the different bilirubin-albumin (B/A) ratio may be one possible alternative explanation for the

higher BBB bilirubin permeability on younger animals after the infusion of an identical

amount of UCB at both P10 and P21 post-natal ages, corresponding to a B/A ratio about 15

and 1.7, respectively (Roger et al., 1996). Experimental evidences do not support this

hypothesis.

Similarly in younger-adult rats (P32-36 days old), in witch the Pgp expression in brain

seems to be similar to the adult level, the Pgp brain inflow is enhanced of the 24% by the Pgp

inhibitor erythromycin, and of the 236% by the Pgp inhibitor and bilirubin-albumin displacing

agent ceftriaxone (Hanko et al., 2003).

In our study, the cBf in jj animals show the highest value of 500-250 nM till P9, then

quickly drops to values 10 times lower (P17), reaching concentrations similar to the P2-9 old

Jj rats (Fig 23). The latter, represents a bilirubin concentration not sufficient to cause any

evident dysfunction in the Jj genotype thus we suggest in the 20-500 nM the range of the

potentially toxic amount of free bilirubin in blood. It is supposable that in the CNS the

amount of free bilirubin might be inferior, also if the impaired barrier properties and a

possible reservoir effect have to be take in consideration.

Maybe the purported neurotoxic threshold for Bf in vitro, until now reported in the 71-

770 nM (Ostrow et al., 2003b), might have to be revised in the lower values.

DISCUSSION

99

In fact, concentrations of 40 nM of Bf show effects on primary culture of astrocytes

(Gennuso et al., 2004). Additionally, data recently produced in our laboratory indicate that

cultured neurons display cytotoxicy from 10 nM of Bf. And Bf of 20 nmol/L (1.2 μg/dL) in

blood was hypothesized to be a risk threshold for bilirubin toxicity in term infants based on

the highest Bf measured after titrating cord sera (Wennberg, unpublished data).

The highest level of (calculated) free bilirubin is attended in the first week of life.

During this period both the impaired growth of the cerebellum and the Mrp1 down-regulation

on CPs take place, indicating in the first week of life the determinant period for BIND in rats.

As the expression of Mrp1, this is the first time, at the best of my knowledge that was

reported as inhibited in jj rats. The mechanism involved in Mrp1 down regulation is still

unknown. The mRNA analysis shows no differences between genotypes in both ventricles,

suggesting a post-transcriptional mechanism. At the same time, Pgp expression on MV and

Cx is enhanced.

One possible hypothesis concerns the inflammatory properties of bilirubin.

According with this observation, the pigment is able to trigger pro-inflammatory cytokine

release from cultured microglia (Gordo et al., 2006) and astrocytes (Fernandes et al., 2006).

While a short exposure the pro inflammatory cytokines triggers the down regulation of both

function and protein expression of Pgp in MV (Hartz et al., 2006; Bauer et al., 2005), recent

data suggest that a chronic exposure of brain capillaries to TNF-α increases both transport

activity and the transporter protein expression (Bauer et al., 2007). But the inflammation in

vivo has been denied (Ahdab-Barmada, 2000).

In contrast to inflammation, cellular stress (such as reactive oxygen species),

unambiguously up-regulates the Pgp transporter (Bauer et al., 2005; Samoto et al., 1994; Felix

and Barrand, 2002). In agreement, both the expression and the Pgp activity have been

founded increased in the oxidative stress situation triggered by the GSH depletion on micro

vessels endothelial cells (Hong et al., 2006).

Bilirubin seems able to induce the glutamate release from both cell type, microglia and

astrocytes (Fernandes et al., 2004; Gordo et al., 2006; Silva et al., 1999), and causes an

impairment of the redox state of the cell (Rodrigues et al., 2002b). In agreement, the

glutamate treatment of cultured endothelial cells induces reactive oxygen species formation

and up-regulation of Pgp by a NMDA receptor mediated mechanism (Zhu and Liu, 2004;

DISCUSSION

100

Bauer et al., 2007). But the bilirubin ability to cause glutamate neurotoxicity has been

recently denied both in vitro and in vivo (Tiribelli and Ostrow, 2005).

By contrast, no data are available for the Mrp1 on BBI. The unique reported down-

regulation of Mrp1 has been done by a COX2 specific inhibitor, colecoxib. But in this study,

the effect is both at the mRNA and at the protein level (Kang et al., 2005). The cellular

signalling pathway is unknown and the only common feature between bilirubin and

colecoxib, seems to be the inhibition of p38 MAP Kinases (Ollinger et al., 2005). Unlikely,

others authors reports an activating properties of bilirubin on p38 (Fernandes et al., 2006).

In conclusion, the complexity of the molecular mechanisms and the paucity of the data

concerning the specific pathway involved in bilirubin-induced neurotoxicity needs to be

clarified and investigated.

Finally, Garay pointed out the idea that UCB entry in the central nervous system from

the blood may occurs through the ventricular system (Rodriguez Garay and Scremin, 1971).

In line with this indication, and expanding the reasoning of all the compounds that could

potentially enter in the CNS via the BCSFB, the protracted down-regulation of Mrp1 may

assume important consequences, especially if what we have seen in rats takes place also in

humans. We may hypotheses that in babies that have suffered BIND or in CN type I patients,

the Mrp1 related protection at the BBI, remains at length reduced as well as the transport of

physiological Mrp1 substrates may be altered (LC4, GSH- and gluthatyon conjugates,

medical compounds, etc).

CONCLUSIONS

101

Collectively, the data obtained from the two models studied during my thesis, have

permitted to define:

1) A mirroring pattern of expression for Mrp1 and Pgp transporters at the two principal

blood brain interfaces in adult animals. With the BBB characterized by Pgp, and the

BCSFB by Mrp1.

2) This differential expression is not modified by the hyperbilirubinemia.

3) Both transporters are localized at the blood side of the respective barriers. Mrp1 is baso-

lateral on choroid plexuses cells, and Pgp apical on MVs, and the localization is well

defined all long the postnatal development.

4) The hyperbilirubinemic condition does not modify the Mrp1 localization on BCSFB

witch remains baso-lateral during all the post-natal development. Similarly, the Pgp

apical localization on MVs doesn’t seems to be modified. But the technical troubles

rend necessary additional data.

5) The Pgp expression is exclusive of the endothelial cells forming the MVs, and is

increased in jj Gunn rats.

6) Conversely the Mrp1 transporter is present also in parenchyma cells (probably

astrocytes and oligodendrocytes) with the higher expression on CPs, where a difference

between the LV and the 4thV CP has been evidenced.

7) We showed a different post-natal developmental profile of expression for the two

transporters. Pgp increase from P9 to P60 on BBB, while Mrp1 is present early and at

levels similar or slightly higher than in adults on BCSFB.

8) The bilirubin toxic effect on Mrp1 expression in CPs during development leads to a

marked and precocious decrease of the transporter expression and this down-regulation

appears to be post-translational. The Pgp developmental profile of expression is not

significantly modified, although the Pgp amount in jj animals is increased.

All together these results indicate that the two barriers differs.

A) The BBB is surrounded by the lipid rich environment of the parenchyma and develops

post-natally, in agreement with the Pgp substrate preference for lipid compounds and the

strong increase of expression from the early post-natal stage to the adult life. The Pgp

mediated defence is not so efficient in the early post-natal age than in adult.

CONCLUSIONS

102

B) The CPs are located between two fluids, the blood and the CSF, they are important for the

development of the brain since the embryonic period and posses the maximal phase II

metabolic activity in brain. The Mrp1 mediated protection at the BCSFB is developed

since the birth and may be involved in the transport of same compounds important in the

development of the brain, or in the maintenance of its redox state, or in transporting

conjugate metabolites (LC4, GSH end redox state balance; GS- sulpha- gluthatyon

conjugates).

C) During the hyperbilirubinemia the Pgp offered protection is not sufficiently modulate until

P17, when the cBf is elevated and could cross the barriers accumulating in brain, and its

accumulation may be facilitated by the simultaneously Mrp1 strong impairment.

Additionally, this deficit hosts rests at length and may be have consequences in the Mrp1

substrate transport ((LC4, GSH end redox state balance; GS- sulfo- gluthathyon

conjugates).

Acknowledgments

103

First, I wish to thank the Professor Claudio Tiribelli, my supervisor, for the opportunity to

work in his laboratory at the Centro Studi Fegato, and for opening my mind (and my life) to

the world.

The Dr. Jean-François Ghersi-Egea, my external supervisor, Charlotte Schmitt of the Blood

Brain Interfaces team (“Interfaçes Sang-Cerveaux”- INSERM U842, Faculté de Médecine

Laennec, Rue Guillaume Paradin, 69008 Lyon , France), and Nathalie Strazielle (Brain-i, 34

Rue du Dr Bonhomme, 69008 Lyon, France ). My profound gratitude goes to all you for the

guidance, support and help for the hard work we have done together during the year I spent in

Lyon.

I wish to thank the Dr. Michelle Fèvre-Montange and the Service de Neuropathologie,

Hôpital Neurologique centralized via the Biological Resource Center NeuroBioTec Banques,

Lyon, France, that provided the human samples.

My gratitude and thanks to the Prof. Stefano Gustincich, my tutor, for the time you spent in

advising me and for the logistical support. Thanks to Marta Biagioli and Milena Pinto, for

your help.

I would like to thank the Dr. Alan Rasini and the Prof. Sergio Parco from the S.C. Laboratorio

Analisi Cliniche IRCCS Burlo Garofolo, for the kind support and for the bilirubin and

albumin measure on the “Gunn rat project”. I hope this may be the starting point for a fruitful

and last long scientific collaboration.

I would like to thank Paola Zarattini and Marco Stebel, to have filled all my exorbitant

demand of Gunn animals.

Finally, thanks to all my collegues: Alessia Bortolussi, Alejandro Aranda, Francesco Fazzari,

Cristina Belllarosa, Cristina Zennaro, Elena Boscolo, Giulia Bortolussi, Graciela Mazzone,

Sandra Leal, Lorella Pascolo, Lucia Corich, Maria Gabriela Mediavilla, Maryam Kazemi,

Natalia Rosso, Riccarda Delfino, Sabrina Corsucci, Vittorio Di Maso, Sebastian Calligaris,

Andrea Berengeno, Mohamed Qaisiya, Norberto Chavez and Leslye Roca-Burgos.

REFERENCE LIST

104

Reference List

Abbott,N. (2002). Astrocyte-endothelial interactions and blood-brain barrier permeability. J. Anat. 200, 527.

Abbott,N.J. and Bundgaard,M. (1992). Electron-dense tracer evidence for a blood-brain barrier in the cuttlefish Sepia officinalis. J. Neurocytol. 21, 276-294.

Abbott,N.J., Ronnback,L., and Hansson,E. (2006). Astrocyte-endothelial interactions at the blood-brain barrier. Nat. Rev. Neurosci. 7, 41-53.

Ahdab-Barmada (2000). The neuropathology of kernicterus:definition and debate. In Neonatal jaudice, M.J.Maisels and J.F.Watchko, eds., pp. 75-88.

Ahlfors,C.E. (2001). Bilirubin-albumin binding and free bilirubin. J. Perinatol. 21 Suppl 1, S40-S42.

Ahlfors,C.E. and Shapiro,S.M. (2001). Auditory brainstem response and unbound bilirubin in jaundiced (jj) Gunn rat pups. Biol. Neonate 80, 158-162.

Akin,E., Clower,B., Tibbs,R., Tang,J., and Zhang,J. (2002). Bilirubin produces apoptosis in cultured bovine brain endothelial cells. Brain Res. 931, 168-175.

Al-Sarraf,H. and Philip,L. (2003). Increased brain uptake and CSF clearance of 14C-glutamate in spontaneously hypertensive rats. Brain Res. 994, 181-187.

Amit,Y. and Brenner,T. (1993). Age-dependent sensitivity of cultured rat glial cells to bilirubin toxicity. Exp. Neurol. 121, 248-255.

Aono,S., Semba,R., Sato,H., and Kashiwamata,S. (1989). Mode of bilirubin deposition in the cerebellum of developing jaundiced Gunn rats. Biol. Neonate 55, 119-123.

Banks,W.A., Kastin,A.J., Brennan,J.M., and Vallance,K.L. (1999). Adsorptive endocytosis of HIV-1gp120 by blood-brain barrier is enhanced by lipopolysaccharide. Exp. Neurol. 156, 165-171.

Baranano,D.E., Rao,M., Ferris,C.D., and Snyder,S.H. (2002). Biliverdin reductase: a major physiologic cytoprotectant. Proc. Natl. Acad. Sci. U. S. A 99, 16093-16098.

Bauer,B., Hartz,A.M., Fricker,G., and Miller,D.S. (2005). Modulation of p-glycoprotein transport function at the blood-brain barrier. Exp. Biol. Med. (Maywood. ) 230, 118-127.

Bauer,B., Hartz,A.M., Pekcec,A., Toellner,K., Miller,D.S., and Potschka,H. (2007). Seizure-Induced Upregulation of P-glycoprotein at the Blood-Brain Barrier through Glutamate and COX-2 Signaling. Mol. Pharmacol.

Bauer,H.C., Bauer,H., Lametschwandtner,A., Amberger,A., Ruiz,P., and Steiner,M. (1993). Neovascularization and the appearance of morphological characteristics of the blood-brain barrier in the embryonic mouse central nervous system. Brain Res. Dev. Brain Res. 75, 269-278.

REFERENCE LIST

105

Beaulieu,E., Demeule,M., Ghitescu,L., and Beliveau,R. (1997). P-glycoprotein is strongly expressed in the luminal membranes of the endothelium of blood vessels in the brain. Biochem. J. 326 ( Pt 2), 539-544.

Begley,D.J. (2004). Delivery of therapeutic agents to the central nervous system: the problems and the possibilities. Pharmacol. Ther. 104, 29-45.

Bendayan,M. (2002). Morphological and cytochemical aspects of capillary permeability. Microsc. Res. Tech. 57, 327-349.

Bendayan,R., Lee,G., and Bendayan,M. (2002). Functional expression and localization of P-glycoprotein at the blood brain barrier. Microsc. Res. Tech. 57, 365-380.

Berezowski,V., Landry,C., Dehouck,M.P., Cecchelli,R., and Fenart,L. (2004). Contribution of glial cells and pericytes to the mRNA profiles of P-glycoprotein and multidrug resistance-associated proteins in an in vitro model of the blood-brain barrier. Brain Res. 1018, 1-9.

Betz,A.L., Firth,J.A., and Goldstein,G.W. (1980). Polarity of the blood-brain barrier: distribution of enzymes between the luminal and antiluminal membranes of brain capillary endothelial cells. Brain Res. 192, 17-28.

Betz,A.L. and Goldstein,G.W. (1980b). The basis for active transport at the blood-brain barrier. Adv. Exp. Med. Biol. 131, 5-16.

Betz,A.L. and Goldstein,G.W. (1980a). Transport of hexoses, potassium and neutral amino acids into capillaries isolated from bovine retina. Exp. Eye Res. 30, 593-605.

Borlongan,C.V., Skinner,S.J., Geaney,M., Vasconcellos,A.V., Elliott,R.B., and Emerich,D.F. (2004). CNS grafts of rat choroid plexus protect against cerebral ischemia in adult rats. Neuroreport 15, 1543-1547.

Borst,P. and Elferink,R.O. (2002). Mammalian ABC transporters in health and disease. Annu. Rev. Biochem. 71, 537-592.

Breen,C.M., Sykes,D.B., Baehr,C., Fricker,G., and Miller,D.S. (2004). Fluorescein-methotrexate transport in rat choroid plexus analyzed using confocal microscopy. Am. J. Physiol Renal Physiol 287, F562-F569.

Brightman,M.W. and Reese,T.S. (1969). Junctions between intimately apposed cell membranes in the vertebrate brain. J. Cell Biol. 40, 648-677.

Brito, M. A. The human RBC as a model of unconjugated bilirubin binding and toxicity. New Trends in Brain Research , 1-38. 2006. Nova Science Publishers, Inc. Ref Type: Generic

Brito,M.A., Brondino,C.D., Moura,J.J., and Brites,D. (2001). Effects of bilirubin molecular species on membrane dynamic properties of human erythrocyte membranes: a spin label electron paramagnetic resonance spectroscopy study. Arch. Biochem. Biophys. 387, 57-65.

Brito,M.A., Silva,R.F., and Brites,D. (2002). Bilirubin induces loss of membrane lipids and exposure of phosphatidylserine in human erythrocytes. Cell Biol. Toxicol. 18, 181-192.

REFERENCE LIST

106

Brito,M.A., Silva,R.M., Matos,D.C., da Silva,A.T., and Brites,D.T. (1996). Alterations of erythrocyte morphology and lipid composition by hyperbilirubinemia. Clin. Chim. Acta 249, 149-165.

Brodersen, R. Binding of bilirubin to albumin and tissues. Monset-Couchard, M. and Minkowski, A. Eds. Basel S. Karger. Physiological and biochemical basis for perinatal medicine. Levine Conference,[First International Meeting, Paris, 1979], 144-152. 1981. Ref Type: Conference Proceeding

Brodersen,R. (1980). Binding of bilirubin to albumin. CRC Crit Rev. Clin. Lab Sci. 11, 305-399.

Caley,D.W. and Maxwell,D.S. (1970). Development of the blood vessels and extracellular spaces during postnatal maturation of rat cerebral cortex. J. Comp Neurol. 138, 31-47.

Calligaris,S., Cekic,D., Roca-Burgos,L., Gerin,F., Mazzone,G., Ostrow,J.D., and Tiribelli,C. (2006). Multidrug resistance associated protein 1 protects against bilirubin-induced cytotoxicity. FEBS Lett. 580, 1355-1359.

Calligaris,S.D., Bellarosa,C., Giraudi,P., Wennberg,R.P., Ostrow,J.D., and Tiribelli,C. (2007). Cytotoxicity is predicted by unbound and not total bilirubin concentration. Pediatr. Res. 62, 576-580.

Cannon,C., Daood,M.J., O'Day,T.L., and Watchko,J.F. (2006). Sex-specific regional brain bilirubin content in hyperbilirubinemic Gunn rat pups. Biol. Neonate 90, 40-45.

Chalasani,N., Chowdhury,N.R., Chowdhury,J.R., and Boyer,T.D. (1997). Kernicterus in an adult who is heterozygous for Crigler-Najjar syndrome and homozygous for Gilbert-type genetic defect. Gastroenterology 112, 2099-2103.

Chodobski,A., Szmydynger-Chodobska,J., Epstein,M.H., and Johanson,C.E. (1995). The role of angiotensin II in the regulation of blood flow to choroid plexuses and cerebrospinal fluid formation in the rat. J. Cereb. Blood Flow Metab 15, 143-151.

Choudhuri,S., Cherrington,N.J., Li,N., and Klaassen,C.D. (2003). Constitutive expression of various xenobiotic and endobiotic transporter mRNAs in the choroid plexus of rats. Drug Metab Dispos. 31, 1337-1345.

Chowdhury,J., Huang,T.J., Kesari,K., Lederstein,M., Arias,I.M., and Roy-Chowdhury,N. (1991). Molecular basis for the lack of bilirubin-specific and 3-methylcholanthrene-inducible UDP-glucuronosyltransferase activities in Gunn rats. The two isoforms are encoded by distinct mRNA species that share an identical single base deletion. J. Biol. Chem. 266, 18294-18298.

Chowdhury,J.R., Kondapalli,R., and Chowdhury,N.R. (1993). Gunn rat: a model for inherited deficiency of bilirubin glucuronidation. Adv. Vet. Sci. Comp Med. 37, 149-173.

Cohen,G., Livovsky,D.M., Kapitulnik,J., and Sasson,S. (2006). Bilirubin increases the expression of glucose transporter-1 and the rate of glucose uptake in vascular endothelial cells. Rev. Diabet. Stud. 3, 127-133.

REFERENCE LIST

107

Cole,S.P., Bhardwaj,G., Gerlach,J.H., Mackie,J.E., Grant,C.E., Almquist,K.C., Stewart,A.J., Kurz,E.U., Duncan,A.M., and Deeley,R.G. (1992). Overexpression of a transporter gene in a multidrug-resistant human lung cancer cell line. Science 258, 1650-1654.

Cole,S.P. and Deeley,R.G. (1998). Multidrug resistance mediated by the ATP-binding cassette transporter protein MRP. Bioessays 20, 931-940.

Conlee,J.W. and Shapiro,S.M. (1997). Development of cerebellar hypoplasia in jaundiced Gunn rats: a quantitative light microscopic analysis. Acta Neuropathol. 93, 450-460.

Cordon-Cardo,C., O'Brien,J.P., Casals,D., Rittman-Grauer,L., Biedler,J.L., Melamed,M.R., and Bertino,J.R. (1989). Multidrug-resistance gene (P-glycoprotein) is expressed by endothelial cells at blood-brain barrier sites. Proc. Natl. Acad. Sci. U. S. A 86, 695-698.

CRIGLER,J.F., Jr. and NAJJAR,V.A. (1952). Congenital familial nonhemolytic jaundice with kernicterus. Pediatrics 10, 169-180.

Dallas,S., Zhu,X., Baruchel,S., Schlichter,L., and Bendayan,R. (2003). Functional expression of the multidrug resistance protein 1 in microglia. J. Pharmacol. Exp. Ther. 307, 282-290.

DAVSON,H. (1963). The cerebrospinal fluid. Ergeb. Physiol 52, 20-72.

Decleves,X., Regina,A., Laplanche,J.L., Roux,F., Boval,B., Launay,J.M., and Scherrmann,J.M. (2000). Functional expression of P-glycoprotein and multidrug resistance-associated protein (Mrp1) in primary cultures of rat astrocytes. J. Neurosci. Res. 60, 594-601.

Deeken,J.F. and Loscher,W. (2007). The blood-brain barrier and cancer: transporters, treatment, and Trojan horses. Clin. Cancer Res. 13, 1663-1674.

Diamond,I. and Schmid,R. (1966). Experimental bilirubin encephalopathy. The mode of entry of bilirubin-14C into the central nervous system. J. Clin. Invest 45, 678-689.

Dore,S. and Snyder,S.H. (1999). Neuroprotective action of bilirubin against oxidative stress in primary hippocampal cultures. Ann. N. Y. Acad. Sci. 890, 167-172.

Doumas,B.T., Kwok-Cheung,P.P., Perry,B.W., Jendrzejczak,B., McComb,R.B., Schaffer,R., and Hause,L.L. (1985a). Candidate reference method for determination of total bilirubin in serum: development and validation. Clin. Chem. 31, 1779-1789.

Doumas,B.T., Wu,T.W., Poon,K.C., and Jendrzejczak,B. (1985b). Chemical nature of a synthetic bilirubin conjugate and its reactivities in the total and direct reactions by the Jendrassik-Grof method. Clin. Chem. 31, 1677-1682.

Drion,N., Lemaire,M., Lefauconnier,J.M., and Scherrmann,J.M. (1996). Role of P-glycoprotein in the blood-brain transport of colchicine and vinblastine. J. Neurochem. 67, 1688-1693.

Drummond,G.S. and Kappas,A. (1984). An experimental model of postnatal jaundice in the suckling rat. Suppression of induced hyperbilirubinemia by Sn-protoporphyrin. J. Clin. Invest 74, 142-149.

REFERENCE LIST

108

Dziegielewska,K.M., Ek,J., Habgood,M.D., and Saunders,N.R. (2001). Development of the choroid plexus. Microsc. Res. Tech. 52, 5-20.

Dziegielewska,K.M., Evans,C.A., Malinowska,D.H., Mollgard,K., Reynolds,M.L., and Saunders,N.R. (1980). Blood-cerebrospinal fluid transfer of plasma proteins during fetal development in the sheep. J. Physiol 300, 457-465.

elAwady,M., Chowdhury,J.R., Kesari,K., van,E.H., Jansen,P.L., Lederstein,M., Arias,I.M., and Chowdhury,N.R. (1990). Mechanism of the lack of induction of UDP-glucuronosyltransferase activity in Gunn rats by 3-methylcholanthrene. Identification of a truncated enzyme. J. Biol. Chem. 265, 10752-10758.

Emerich,D.F., Skinner,S.J., Borlongan,C.V., Vasconcellos,A.V., and Thanos,C.G. (2005). The choroid plexus in the rise, fall and repair of the brain. Bioessays 27, 262-274.

Emi,Y., Ikushiro,S., and Iyanagi,T. (1995). Drug-responsive and tissue-specific alternative expression of multiple first exons in rat UDP-glucuronosyltransferase family 1 (UGT1) gene complex. J. Biochem. (Tokyo) 117, 392-399.

Falcao,A.S., Bellarosa,C., Fernandes,A., Brito,M.A., Silva,R.F., Tiribelli,C., and Brites,D. (2007). Role of multidrug resistance-associated protein 1 expression in the in vitro susceptibility of rat nerve cell to unconjugated bilirubin. Neuroscience 144, 878-888.

Falcao,A.S., Fernandes,A., Brito,M.A., Silva,R.F., and Brites,D. (2006). Bilirubin-induced immunostimulant effects and toxicity vary with neural cell type and maturation state. Acta Neuropathol. (Berl) 112, 95-105.

Felix,R.A. and Barrand,M.A. (2002). P-glycoprotein expression in rat brain endothelial cells: evidence for regulation by transient oxidative stress. J. Neurochem. 80, 64-72.

Fernandes,A., Falcao,A.S., Silva,R.F., Gordo,A.C., Gama,M.J., Brito,M.A., and Brites,D. (2006). Inflammatory signalling pathways involved in astroglial activation by unconjugated bilirubin. J. Neurochem. 96, 1667-1679.

Fernandes,A., Silva,R.F., Falcao,A.S., Brito,M.A., and Brites,D. (2004). Cytokine production, glutamate release and cell death in rat cultured astrocytes treated with unconjugated bilirubin and LPS. J. Neuroimmunol. 153, 64-75.

Fernetti,C., Pascolo,L., Podda,E., Gennaro,R., Stebel,M., and Tiribelli,C. (2001). Preparation of an antibody recognizing both human and rodent MRP1. Biochem. Biophys. Res. Commun. 288, 1064-1068.

Furuse,M., Furuse,K., Sasaki,H., and Tsukita,S. (2001). Conversion of zonulae occludentes from tight to leaky strand type by introducing claudin-2 into Madin-Darby canine kidney I cells. J. Cell Biol. 153, 263-272.

Gao,B., Hagenbuch,B., Kullak-Ublick,G.A., Benke,D., Aguzzi,A., and Meier,P.J. (2000). Organic anion-transporting polypeptides mediate transport of opioid peptides across blood-brain barrier. J. Pharmacol. Exp. Ther. 294, 73-79.

REFERENCE LIST

109

Genc,S., Genc,K., Kumral,A., Baskin,H., and Ozkan,H. (2003). Bilirubin is cytotoxic to rat oligodendrocytes in vitro. Brain Res. 985, 135-141.

Gennuso,F., Fernetti,C., Tirolo,C., Testa,N., L'Episcopo,F., Caniglia,S., Morale,M.C., Ostrow,J.D., Pascolo,L., Tiribelli,C., and Marchetti,B. (2004). Bilirubin protects astrocytes from its own toxicity by inducing up-regulation and translocation of multidrug resistance-associated protein 1 (Mrp1). Proc. Natl. Acad. Sci. U. S. A 101, 2470-2475.

Georges,E., Bradley,G., Gariepy,J., and Ling,V. (1990). Detection of P-glycoprotein isoforms by gene-specific monoclonal antibodies. Proc. Natl. Acad. Sci. U. S. A 87, 152-156.

Ghersi-Egea,J.F., Leininger-Muller,B., Cecchelli,R., and Fenstermacher,J.D. (1995). Blood-brain interfaces: relevance to cerebral drug metabolism. Toxicol. Lett. 82-83, 645-653.

Ghersi-Egea,J.F., Leninger-Muller,B., Suleman,G., Siest,G., and Minn,A. (1994). Localization of drug-metabolizing enzyme activities to blood-brain interfaces and circumventricular organs. J. Neurochem. 62, 1089-1096.

Ghersi-Egea,J.F., Minn,A., and Siest,G. (1988). A new aspect of the protective functions of the blood-brain barrier: activities of four drug-metabolizing enzymes in isolated rat brain microvessels. Life Sci. 42, 2515-2523.

Ghersi-Egea,J.F. and Strazielle,N. (2001). Brain drug delivery, drug metabolism, and multidrug resistance at the choroid plexus. Microsc. Res. Tech. 52, 83-88.

Ghersi-Egea,J.F., Strazielle,N., Murat,A., Jouvet,A., Buenerd,A., and Belin,M.F. (2006). Brain protection at the blood-cerebrospinal fluid interface involves a glutathione-dependent metabolic barrier mechanism. J. Cereb. Blood Flow Metab 26, 1165-1175.

Golden,P.L. and Pardridge,W.M. (1999). P-Glycoprotein on astrocyte foot processes of unfixed isolated human brain capillaries. Brain Res. 819, 143-146.

Goldstein,G.W. and Betz,A.L. (1983). Recent advances in understanding brain capillary function. Ann. Neurol. 14, 389-395.

Gordo,A.C., Falcao,A.S., Fernandes,A., Brito,M.A., Silva,R.F., and Brites,D. (2006). Unconjugated bilirubin activates and damages microglia. J. Neurosci. Res. 84, 194-201.

Gourley,G.R. (1997). Bilirubin metabolism and kernicterus. Adv. Pediatr. 44, 173-229.

Grojean,S., Koziel,V., Vert,P., and Daval,J.L. (2000a). Bilirubin induces apoptosis via activation of NMDA receptors in developing rat brain neurons. Exp. Neurol. 166, 334-341.

Grojean,S., Koziel,V., Vert,P., and Daval,J.L. (2000b). Bilirubin induces apoptosis via activation of NMDA receptors in developing rat brain neurons. Exp. Neurol. 166, 334-341.

Gunn,C.H. (1938). Hereditary acholuric jaundice in a new mutant strain of rats. J. Hered. 29, 137-139.

REFERENCE LIST

110

Haimeur,A., Conseil,G., Deeley,R.G., and Cole,S.P. (2004). The MRP-related and BCRP/ABCG2 multidrug resistance proteins: biology, substrate specificity and regulation. Curr. Drug Metab 5, 21-53.

Hanko,E., Tommarello,S., Watchko,J.F., and Hansen,T.W. (2003). Administration of drugs known to inhibit P-glycoprotein increases brain bilirubin and alters the regional distribution of bilirubin in rat brain. Pediatr. Res. 54, 441-445.

Hansen,T.W. (1994c). Bilirubin in the brain. Distribution and effects on neurophysiological and neurochemical processes. Clin. Pediatr. (Phila) 33, 452-459.

Hansen,T.W. (1994b). Bilirubin in the brain. Distribution and effects on neurophysiological and neurochemical processes. Clin. Pediatr. (Phila) 33, 452-459.

Hansen,T.W. (1994a). Bilirubin in the brain. Distribution and effects on neurophysiological and neurochemical processes. Clin. Pediatr. (Phila) 33, 452-459.

Hansen,T.W. and Cashore,W.J. (1995). Rates of bilirubin clearance from rat brain regions. Biol. Neonate 68, 135-140.

Hansen,T.W., Mathiesen,S.B., and Walaas,S.I. (1996). Bilirubin has widespread inhibitory effects on protein phosphorylation. Pediatr. Res. 39, 1072-1077.

Hansen,T.W., Mathiesen,S.B., and Walaas,S.I. (1997). Modulation of the effect of bilirubin on protein phosphorylation by lysine-containing peptides. Pediatr. Res. 42, 615-617.

Hartz,A.M., Bauer,B., Fricker,G., and Miller,D.S. (2006). Rapid modulation of P-glycoprotein-mediated transport at the blood-brain barrier by tumor necrosis factor-alpha and lipopolysaccharide. Mol. Pharmacol. 69, 462-470.

Hirrlinger,J., Konig,J., and Dringen,R. (2002). Expression of mRNAs of multidrug resistance proteins (Mrps) in cultured rat astrocytes, oligodendrocytes, microglial cells and neurones. J. Neurochem. 82, 716-719.

Hong,H., Lu,Y., Ji,Z.N., and Liu,G.Q. (2006). Up-regulation of P-glycoprotein expression by glutathione depletion-induced oxidative stress in rat brain microvessel endothelial cells. J. Neurochem. 98, 1465-1473.

Hooiveld,G.J., Heegsma,J., van Montfoort,J.E., Jansen,P.L., Meijer,D.K., and Muller,M. (2002). Stereoselective transport of hydrophilic quaternary drugs by human MDR1 and rat Mdr1b P-glycoproteins. Br. J. Pharmacol. 135, 1685-1694.

Iyanagi,T., Watanabe,T., and Uchiyama,Y. (1989). The 3-methylcholanthrene-inducible UDP-glucuronosyltransferase deficiency in the hyperbilirubinemic rat (Gunn rat) is caused by a -1 frameshift mutation. J. Biol. Chem. 264, 21302-21307.

Jacobsen,J. (1969). Binding of bilirubin to human serum albumin - determination of the dissociation constants. FEBS Lett. 5, 112-114.

REFERENCE LIST

111

Jagt,D.L., Wilson,S.P., Dean,V.L., and Simons,P.C. (1982). Bilirubin binding to rat liver ligandins (glutathione S-transferases A and B). Relationship between bilirubin binding and transferase activity. J. Biol. Chem. 257, 1997-2001.

Jette,L., Murphy,G.F., Leclerc,J.M., and Beliveau,R. (1995a). Interaction of drugs with P-glycoprotein in brain capillaries. Biochem. Pharmacol. 50, 1701-1709.

Jette,L., Pouliot,J.F., Murphy,G.F., and Beliveau,R. (1995b). Isoform I (mdr3) is the major form of P-glycoprotein expressed in mouse brain capillaries. Evidence for cross-reactivity of antibody C219 with an unrelated protein. Biochem. J. 305 ( Pt 3), 761-766.

Johansson,P.A., Dziegielewska,K.M., Ek,C.J., Habgood,M.D., Liddelow,S.A., Potter,A.M., Stolp,H.B., and Saunders,N.R. (2006). Blood-CSF barrier function in the rat embryo. Eur. J. Neurosci. 24, 65-76.

Johnson,J.A., Hayward,J.J., Kornguth,S.E., and Siegel,F.L. (1993). Effects of hyperbilirubinaemia on glutathione S-transferase isoenzymes in cerebellar cortex of the Gunn rat. Biochem. J. 291 ( Pt 2), 453-461.

Kadakol,A., Ghosh,S.S., Sappal,B.S., Sharma,G., Chowdhury,J.R., and Chowdhury,N.R. (2000). Genetic lesions of bilirubin uridine-diphosphoglucuronate glucuronosyltransferase (UGT1A1) causing Crigler-Najjar and Gilbert syndromes: correlation of genotype to phenotype. Hum. Mutat. 16, 297-306.

Kang,H.K., Lee,E., Pyo,H., and Lim,S.J. (2005). Cyclooxygenase-independent down-regulation of multidrug resistance-associated protein-1 expression by celecoxib in human lung cancer cells. Mol. Cancer Ther. 4, 1358-1363.

Karssen,A.M., Meijer,O.C., van,d.S., I, Lucassen,P.J., de Lange,E.C., De Boer,A.G., and de Kloet,E.R. (2001). Multidrug resistance P-glycoprotein hampers the access of cortisol but not of corticosterone to mouse and human brain. Endocrinology 142, 2686-2694.

Katoh-Semba,R. and Kashiwamata,S. (1980). Interaction of bilirubin with brain capillaries and its toxicity. Biochim. Biophys. Acta 632, 290-297.

Keep,R.F. and Jones,H.C. (1990). A morphometric study on the development of the lateral ventricle choroid plexus, choroid plexus capillaries and ventricular ependyma in the rat. Brain Res. Dev. Brain Res. 56, 47-53.

Keino,H., Sato,H., Semba,R., Aono,S., Aoki,E., and Kashiwamata,S. (1985). Mode of prevention by phototherapy of cerebellar hypoplasia in a new Sprague-Dawley strain of jaundiced Gunn rats. Pediatr. Neurosci. 12, 145-150.

Kemper,E.M., Cleypool,C., Boogerd,W., Beijnen,J.H., and van,T.O. (2004). The influence of the P-glycoprotein inhibitor zosuquidar trihydrochloride (LY335979) on the brain penetration of paclitaxel in mice. Cancer Chemother. Pharmacol. 53, 173-178.

Keshavan,P., Schwemberger,S.J., Smith,D.L., Babcock,G.F., and Zucker,S.D. (2004). Unconjugated bilirubin induces apoptosis in colon cancer cells by triggering mitochondrial depolarization. Int. J. Cancer 112, 433-445.

REFERENCE LIST

112

Knott,G.W., Dziegielewska,K.M., Habgood,M.D., Li,Z.S., and Saunders,N.R. (1997). Albumin transfer across the choroid plexus of South American opossum (Monodelphis domestica). J. Physiol 499 ( Pt 1), 179-194.

Kusuhara,H., Suzuki,H., Naito,M., Tsuruo,T., and Sugiyama,Y. (1998). Characterization of efflux transport of organic anions in a mouse brain capillary endothelial cell line. J. Pharmacol. Exp. Ther. 285, 1260-1265.

Lawrenson,J.G., Reid,A.R., Finn,T.M., Orte,C., and Allt,G. (1999). Cerebral and pial microvessels: differential expression of gamma-glutamyl transpeptidase and alkaline phosphatase. Anat. Embryol. (Berl) 199, 29-34.

Lee,G., Schlichter,L., Bendayan,M., and Bendayan,R. (2001). Functional expression of P-glycoprotein in rat brain microglia. J. Pharmacol. Exp. Ther. 299, 204-212.

Leggas,M., Adachi,M., Scheffer,G.L., Sun,D., Wielinga,P., Du,G., Mercer,K.E., Zhuang,Y., Panetta,J.C., Johnston,B., Scheper,R.J., Stewart,C.F., and Schuetz,J.D. (2004). Mrp4 confers resistance to topotecan and protects the brain from chemotherapy. Mol. Cell Biol. 24, 7612-7621.

Leslie,E.M., Deeley,R.G., and Cole,S.P. (2005). Multidrug resistance proteins: role of P-glycoprotein, MRP1, MRP2, and BCRP (ABCG2) in tissue defense. Toxicol. Appl. Pharmacol. 204, 216-237.

Leslie,E.M., Deeley,R.G., and Cole,S.P. (2001). Toxicological relevance of the multidrug resistance protein 1, MRP1 (ABCC1) and related transporters. Toxicology 167, 3-23.

Li,J.Y., Sugimura,K., Boado,R.J., Lee,H.J., Zhang,C., Duebel,S., and Pardridge,W.M. (1999). Genetically engineered brain drug delivery vectors: cloning, expression and in vivo application of an anti-transferrin receptor single chain antibody-streptavidin fusion gene and protein. Protein Eng 12, 787-796.

Liebner,S., Kniesel,U., Kalbacher,H., and Wolburg,H. (2000). Correlation of tight junction morphology with the expression of tight junction proteins in blood-brain barrier endothelial cells. Eur. J. Cell Biol. 79, 707-717.

Lin,S., Yan,C., Wei,X., Paul,S.M., and Du,Y. (2003). p38 MAP kinase mediates bilirubin-induced neuronal death of cultured rat cerebellar granule neurons. Neurosci. Lett. 353, 209-212.

Lippoldt,A., Kniesel,U., Liebner,S., Kalbacher,H., Kirsch,T., Wolburg,H., and Haller,H. (2000a). Structural alterations of tight junctions are associated with loss of polarity in stroke-prone spontaneously hypertensive rat blood-brain barrier endothelial cells. Brain Res. 885, 251-261.

Lippoldt,A., Liebner,S., Andbjer,B., Kalbacher,H., Wolburg,H., Haller,H., and Fuxe,K. (2000b). Organization of choroid plexus epithelial and endothelial cell tight junctions and regulation of claudin-1, -2 and -5 expression by protein kinase C. Neuroreport 11, 1427-1431.

Liu,X.D. and Liu,G.Q. (2001). P glycoprotein regulated transport of glutamate at blood brain barrier. Acta Pharmacol. Sin. 22, 111-116.

REFERENCE LIST

113

Lojda,Z. (1981). Proteinases in pathology. Usefulness of histochemical methods. J. Histochem. Cytochem. 29, 481-493.

Loscher,W. and Potschka,H. (2005). Blood-brain barrier active efflux transporters: ATP-binding cassette gene family. NeuroRx. 2, 86-98.

Maroti,Z., Katona,M., Orvos,H., Nemeth,I., Farkas,I., and Turi,S. (2007). Heme oxygenase-1 expression in premature and mature neonates during the first week of life. Eur. J. Pediatr. 166, 1033-1038.

Matsuoka,Y., Okazaki,M., Kitamura,Y., and Taniguchi,T. (1999). Developmental expression of P-glycoprotein (multidrug resistance gene product) in the rat brain. J. Neurobiol. 39, 383-392.

Mayer,M. (2000). Association of serum bilirubin concentration with risk of coronary artery disease. Clin. Chem. 46, 1723-1727.

McDonald,J.W., Shapiro,S.M., Silverstein,F.S., and Johnston,M.V. (1998). Role of glutamate receptor-mediated excitotoxicity in bilirubin-induced brain injury in the Gunn rat model. Exp. Neurol. 150, 21-29.

Mercier,C., Masseguin,C., Roux,F., Gabrion,J., and Scherrmann,J.M. (2004). Expression of P-glycoprotein (ABCB1) and Mrp1 (ABCC1) in adult rat brain: focus on astrocytes. Brain Res. 1021, 32-40.

Miller,D.S., Nobmann,S.N., Gutmann,H., Toeroek,M., Drewe,J., and Fricker,G. (2000). Xenobiotic transport across isolated brain microvessels studied by confocal microscopy. Mol. Pharmacol. 58, 1357-1367.

Mustafa,M.G. and King,T.E. (1970). Binding of bilirubin with lipid. A possible mechanism of its toxic reactions in mitochondria. J. Biol. Chem. 245, 1084-1089.

Nag,S. (2003). Blood-brain barrier permeability using tracers and immunohistochemistry. Methods Mol. Med. 89, 133-144.

Nagaoka,S. and Cowger,M.L. (1978). Interaction of bilirubin with lipids studied by fluorescence quenching method. J. Biol. Chem. 253, 2005-2011.

Ngai,K.C., Yeung,C.Y., and Leung,C.S. (2000). Difference in susceptibilities of different cell lines to bilirubin damage. J. Paediatr. Child Health 36, 51-55.

Nies,A.T., Jedlitschky,G., Konig,J., Herold-Mende,C., Steiner,H.H., Schmitt,H.P., and Keppler,D. (2004). Expression and immunolocalization of the multidrug resistance proteins, MRP1-MRP6 (ABCC1-ABCC6), in human brain. Neuroscience 129, 349-360.

Odell,G.B. (1959). Studies in kernicterus. I. The protein binding of bilirubin. J. Clin. Invest 38, 823-833.

Ollinger,R., Bilban,M., Erat,A., Froio,A., McDaid,J., Tyagi,S., Csizmadia,E., Graca-Souza,A.V., Liloia,A., Soares,M.P., Otterbein,L.E., Usheva,A., Yamashita,K., and Bach,F.H. (2005). Bilirubin: a natural inhibitor of vascular smooth muscle cell proliferation. Circulation 112, 1030-1039.

REFERENCE LIST

114

Ostrow, J. D. Bile pigments and jaundice. 1987. Marcel Dekker,Inc. Ref Type: Generic

Ostrow,J.D., Mukerjee,P., and Tiribelli,C. (1994). Structure and binding of unconjugated bilirubin: relevance for physiological and pathophysiological function. J. Lipid Res. 35, 1715-1737.

Ostrow,J.D., Pascolo,L., Brites,D., and Tiribelli,C. (2004b). Molecular basis of bilirubin-induced neurotoxicity. Trends Mol. Med. 10, 65-70.

Ostrow,J.D., Pascolo,L., Brites,D., and Tiribelli,C. (2004a). Molecular basis of bilirubin-induced neurotoxicity. Trends Mol. Med. 10, 65-70.

Ostrow,J.D., Pascolo,L., Shapiro,S.M., and Tiribelli,C. (2003a). New concepts in bilirubin encephalopathy. Eur. J. Clin. Invest 33, 988-997.

Ostrow,J.D., Pascolo,L., and Tiribelli,C. (2003b). Reassessment of the unbound concentrations of unconjugated bilirubin in relation to neurotoxicity in vitro. Pediatr. Res. 54, 926.

Ostrow,J.D., Pascolo,L., and Tiribelli,C. (2002). Mechanisms of bilirubin neurotoxicity. Hepatology 35, 1277-1280.

Ostrow,J.D. and Tiribelli,C. (2003). Bilirubin, a curse and a boon. Gut 52, 1668-1670.

Oude Elferink,R.P. and Zadina,J. (2001). MDR1 P-glycoprotein transports endogenous opioid peptides. Peptides 22, 2015-2020.

PAPPENHEIMER,J.R., HEISEY,S.R., and JORDAN,E.F. (1961). Active transport of Diodrast and phenolsulfonphthalein from cerebrospinal fluid to blood. Am. J. Physiol 200, 1-10.

Pardridge,W.M. (2003). Blood-brain barrier genomics and the use of endogenous transporters to cause drug penetration into the brain. Curr. Opin. Drug Discov. Devel. 6, 683-691.

Pardridge,W.M. (2007). Blood-brain barrier genomics. Stroke 38, 686-690.

Pardridge,W.M., Golden,P.L., Kang,Y.S., and Bickel,U. (1997). Brain microvascular and astrocyte localization of P-glycoprotein. J. Neurochem. 68, 1278-1285.

Pardridge,W.M., Sakiyama,R., and Fierer,G. (1983). Transport of propranolol and lidocaine through the rat blood-brain barrier. Primary role of globulin-bound drug. J. Clin. Invest 71, 900-908.

Pascolo,L., Fernetti,C., Garcia-Mediavilla,M.V., Ostrow,J.D., and Tiribelli,C. (2001). Mechanisms for the transport of unconjugated bilirubin in human trophoblastic BeWo cells. FEBS Lett. 495, 94-99.

Potschka,H., Fedrowitz,M., and Loscher,W. (2003). Multidrug resistance protein MRP2 contributes to blood-brain barrier function and restricts antiepileptic drug activity. J. Pharmacol. Exp. Ther. 306, 124-131.

REFERENCE LIST

115

Potschka,H. and Loscher,W. (2001). In vivo evidence for P-glycoprotein-mediated transport of phenytoin at the blood-brain barrier of rats. Epilepsia 42, 1231-1240.

Rao,V.V., Dahlheimer,J.L., Bardgett,M.E., Snyder,A.Z., Finch,R.A., Sartorelli,A.C., and Piwnica-Worms,D. (1999). Choroid plexus epithelial expression of MDR1 P glycoprotein and multidrug resistance-associated protein contribute to the blood-cerebrospinal-fluid drug-permeability barrier. Proc. Natl. Acad. Sci. U. S. A 96, 3900-3905.

Reese,T.S. and Karnovsky,M.J. (1967). Fine structural localization of a blood-brain barrier to exogenous peroxidase. J. Cell Biol. 34, 207-217.

Regina,A., Koman,A., Piciotti,M., El,H.B., Center,M.S., Bergmann,R., Couraud,P.O., and Roux,F. (1998). Mrp1 multidrug resistance-associated protein and P-glycoprotein expression in rat brain microvessel endothelial cells. J. Neurochem. 71, 705-715.

Rehman,K.K., Ayesha,Q., Khan,A.A., Ahmed,N., and Habibullah,C.M. (2004). Tyrosine aminotransferase and gamma-glutamyl transferase activity in human fetal hepatocyte primary cultures under proliferative conditions. Cell Biochem. Funct. 22, 89-96.

Rigato,I., Ostrow,J.D., and Tiribelli,C. (2005). Bilirubin and the risk of common non-hepatic diseases. Trends Mol. Med. 11, 277-283.

Rigato,I., Pascolo,L., Fernetti,C., Ostrow,J.D., and Tiribelli,C. (2004). The human multidrug-resistance-associated protein MRP1 mediates ATP-dependent transport of unconjugated bilirubin. Biochem. J. 383, 335-341.

Rizzi,M., Caccia,S., Guiso,G., Richichi,C., Gorter,J.A., Aronica,E., Aliprandi,M., Bagnati,R., Fanelli,R., D'Incalci,M., Samanin,R., and Vezzani,A. (2002). Limbic seizures induce P-glycoprotein in rodent brain: functional implications for pharmacoresistance. J. Neurosci. 22, 5833-5839.

Roca,L., Calligaris,S., Wennberg,R.P., Ahlfors,C.E., Malik,S.G., Ostrow,J.D., and Tiribelli,C. (2006). Factors affecting the binding of bilirubin to serum albumins: validation and application of the peroxidase method. Pediatr. Res. 60, 724-728.

Rodrigues,C.M., Sola,S., and Brites,D. (2002a). Bilirubin induces apoptosis via the mitochondrial pathway in developing rat brain neurons. Hepatology 35, 1186-1195.

Rodrigues,C.M., Sola,S., Brito,M.A., Brites,D., and Moura,J.J. (2002b). Bilirubin directly disrupts membrane lipid polarity and fluidity, protein order, and redox status in rat mitochondria. J. Hepatol. 36, 335-341.

Rodrigues,C.M., Sola,S., Silva,R., and Brites,D. (2000). Bilirubin and amyloid-beta peptide induce cytochrome c release through mitochondrial membrane permeabilization. Mol. Med. 6, 936-946.

Rodrigues,C.M., Sola,S., Silva,R.F., and Brites,D. (2002c). Aging confers different sensitivity to the neurotoxic properties of unconjugated bilirubin. Pediatr. Res. 51, 112-118.

Rodriguez Garay,E.A. and Scremin,O.U. (1971). Transfer of bilirubin- 14 C between blood, cerebrospinal fluid, and brain tissue. Am. J. Physiol 221, 1264-1270.

REFERENCE LIST

116

Rodriguez-Baeza,A., Reina-de la,T.F., Poca,A., Marti,M., and Garnacho,A. (2003). Morphological features in human cortical brain microvessels after head injury: a three-dimensional and immunocytochemical study. Anat. Rec. A Discov. Mol. Cell Evol. Biol. 273, 583-593.

Roger,C., Koziel,V., Vert,P., and Nehlig,A. (1995). Mapping of the consequences of bilirubin exposure in the immature rat: local cerebral metabolic rates for glucose during moderate and severe hyperbilirubinemia. Early Hum. Dev. 43, 133-144.

Roger,C., Koziel,V., Vert,P., and Nehlig,A. (1996). Autoradiographic mapping of local cerebral permeability to bilirubin in immature rats: effects of hyperbilirubinemia. Pediatr. Res. 39, 64-71.

Rubaltelli,F.F. (1993). Bilirubin metabolism in the newborn. Biol. Neonate 63, 133-138.

Sage,M.R. (1982). Blood-brain barrier: phenomenon of increasing importance to the imaging clinician. AJR Am. J. Roentgenol. 138, 887-898.

Samoto,K., Ikezaki,K., Yokoyama,N., and Fukui,M. (1994). P-glycoprotein expression in brain capillary endothelial cells after focal ischaemia in the rat. Neurol. Res. 16, 217-223.

Sanchez del Pino,M.M., Hawkins,R.A., and Peterson,D.R. (1995). Biochemical discrimination between luminal and abluminal enzyme and transport activities of the blood-brain barrier. J. Biol. Chem. 270, 14907-14912.

Satoh,K., Takahashi,G., Miura,T., Hayakari,M., and Hatayama,I. (2005). Enzymatic detection of precursor cell populations of preneoplastic foci positive for gamma-glutamyltranspeptidase in rat liver. Int. J. Cancer 115, 711-716.

Schiff,D., Chan,G., and Poznansky,M.J. (1985). Bilirubin toxicity in neural cell lines N115 and NBR10A. Pediatr. Res. 19, 908-911.

Schinkel,A.H. and Jonker,J.W. (2003). Mammalian drug efflux transporters of the ATP binding cassette (ABC) family: an overview. Adv. Drug Deliv. Rev. 55, 3-29.

Schinkel,A.H., Wagenaar,E., Mol,C.A., and van,D.L. (1996). P-glycoprotein in the blood-brain barrier of mice influences the brain penetration and pharmacological activity of many drugs. J. Clin. Invest 97, 2517-2524.

Schmid,R., Axelrod,J., HAMMAKER,L., and SWARM,R.L. (1958). Congenital jaundice in rats, due to a defect in glucuronide formation. J. Clin. Invest 37, 1123-1130.

Schutta,H.S. and Johnson,L. (1969). Clinical signs and morphologic abnormalities in Gunn rats treated with sulfadimethoxine. J. Pediatr. 75, 1070-1079.

Sedlak,T.W. and Snyder,S.H. (2004). Bilirubin benefits: cellular protection by a biliverdin reductase antioxidant cycle. Pediatrics 113, 1776-1782.

Serrano,M.A., Bayon,J.E., Pascolo,L., Tiribelli,C., Ostrow,J.D., Gonzalez-Gallego,J., and Marin,J.J. (2002). Evidence for carrier-mediated transport of unconjugated bilirubin across plasma membrane vesicles from human placental trophoblast. Placenta 23, 527-535.

REFERENCE LIST

117

Servedio,V., d'Apolito,M., Maiorano,N., Minuti,B., Torricelli,F., Ronchi,F., Zancan,L., Perrotta,S., Vajro,P., Boschetto,L., and Iolascon,A. (2005). Spectrum of UGT1A1 mutations in Crigler-Najjar (CN) syndrome patients: identification of twelve novel alleles and genotype-phenotype correlation. Hum. Mutat. 25, 325.

Seubert,J.M., Darmon,A.J., El-Kadi,A.O., D'Souza,S.J., and Bend,J.R. (2002). Apoptosis in murine hepatoma hepa 1c1c7 wild-type, C12, and C4 cells mediated by bilirubin. Mol. Pharmacol. 62, 257-264.

Shapiro,S.M. (2002). Somatosensory and brainstem auditory evoked potentials in the Gunn rat model of acute bilirubin neurotoxicity. Pediatr. Res. 52, 844-849.

Shelby,M.K., Cherrington,N.J., Vansell,N.R., and Klaassen,C.D. (2003). Tissue mRNA expression of the rat UDP-glucuronosyltransferase gene family. Drug Metab Dispos. 31, 326-333.

Siakotos,A.N. and Rouser,G. (1969). Isolation of highly purified human and bovine brain endothelial cells and nuclei and their phospholipid composition. Lipids 4, 234-239.

Silva,R., Mata,L.R., Gulbenkian,S., Brito,M.A., Tiribelli,C., and Brites,D. (1999). Inhibition of glutamate uptake by unconjugated bilirubin in cultured cortical rat astrocytes: role of concentration and pH. Biochem. Biophys. Res. Commun. 265, 67-72.

Silva,R.F., Rodrigues,C.M., and Brites,D. (2002). Rat cultured neuronal and glial cells respond differently to toxicity of unconjugated bilirubin. Pediatr. Res. 51, 535-541.

Simionescu,M., Ghinea,N., Fixman,A., Lasser,M., Kukes,L., Simionescu,N., and Palade,G.E. (1988). The cerebral microvasculature of the rat: structure and luminal surface properties during early development. J. Submicrosc. Cytol. Pathol. 20, 243-261.

Sisodiya,S.M., Heffernan,J., and Squier,M.V. (1999). Over-expression of P-glycoprotein in malformations of cortical development. Neuroreport 10, 3437-3441.

Sisodiya,S.M., Lin,W.R., Harding,B.N., Squier,M.V., and Thom,M. (2002). Drug resistance in epilepsy: expression of drug resistance proteins in common causes of refractory epilepsy. Brain 125, 22-31.

Sisodiya,S.M., Lin,W.R., Squier,M.V., and Thom,M. (2001). Multidrug-resistance protein 1 in focal cortical dysplasia. Lancet 357, 42-43.

Soontornmalai,A., Vlaming,M.L., and Fritschy,J.M. (2006). Differential, strain-specific cellular and subcellular distribution of multidrug transporters in murine choroid plexus and blood-brain barrier. Neuroscience 138, 159-169.

Soorani-Lunsing,I., Woltil,H.A., and Hadders-Algra,M. (2001). Are moderate degrees of hyperbilirubinemia in healthy term neonates really safe for the brain? Pediatr. Res. 50, 701-705.

Stewart,P.A. and Hayakawa,E.M. (1987). Interendothelial junctional changes underlie the developmental 'tightening' of the blood-brain barrier. Brain Res. 429, 271-281.

REFERENCE LIST

118

Stewart,P.A. and Hayakawa,K. (1994). Early ultrastructural changes in blood-brain barrier vessels of the rat embryo. Brain Res. Dev. Brain Res. 78, 25-34.

Strauss,K.A., Robinson,D.L., Vreman,H.J., Puffenberger,E.G., Hart,G., and Morton,D.H. (2006). Management of hyperbilirubinemia and prevention of kernicterus in 20 patients with Crigler-Najjar disease. Eur. J. Pediatr. 165, 306-319.

Strazielle,N., Belin,M.F., and Ghersi-Egea,J.F. (2003). Choroid plexus controls brain availability of anti-HIV nucleoside analogs via pharmacologically inhibitable organic anion transporters. AIDS 17, 1473-1485.

Strazielle,N. and Ghersi-Egea,J.F. (1999). Demonstration of a coupled metabolism-efflux process at the choroid plexus as a mechanism of brain protection toward xenobiotics. J. Neurosci. 19, 6275-6289.

Strazielle,N. and Ghersi-Egea,J.F. (2005). Factors affecting delivery of antiviral drugs to the brain. Rev. Med. Virol. 15, 105-133.

Strazielle,N., Khuth,S.T., and Ghersi-Egea,J.F. (2004). Detoxification systems, passive and specific transport for drugs at the blood-CSF barrier in normal and pathological situations. Adv. Drug Deliv. Rev. 56, 1717-1740.

Strazielle,N., Mutin,M., and Ghersi-Egea,J.F. (2005). [The choroid plexuses: a dynamic interface between the blood and the cerebrospinal fluid]. Morphologie. 89, 90-101.

Sugiyama,D., Kusuhara,H., Lee,Y.J., and Sugiyama,Y. (2003). Involvement of multidrug resistance associated protein 1 (Mrp1) in the efflux transport of 17beta estradiol-D-17beta-glucuronide (E217betaG) across the blood-brain barrier. Pharm. Res. 20, 1394-1400.

Sugiyama,D., Kusuhara,H., Shitara,Y., Abe,T., Meier,P.J., Sekine,T., Endou,H., Suzuki,H., and Sugiyama,Y. (2001). Characterization of the efflux transport of 17beta-estradiol-D-17beta-glucuronide from the brain across the blood-brain barrier. J. Pharmacol. Exp. Ther. 298, 316-322.

Takagishi,Y. and Yamamura,H. (1989). Purkinje cell abnormalities and synaptogenesis in genetically jaundiced rats (Gunn rats). Brain Res. 492, 116-128.

Takahashi,M., Sugiyama,K., Shumiya,S., and Nagase,S. (1984). Penetration of bilirubin into the brain in albumin-deficient and jaundiced rats (AJR) and Nagase analbuminemic rats (NAR). J. Biochem. 96, 1705-1712.

Thiebaut,F., Tsuruo,T., Hamada,H., Gottesman,M.M., Pastan,I., and Willingham,M.C. (1989). Immunohistochemical localization in normal tissues of different epitopes in the multidrug transport protein P170: evidence for localization in brain capillaries and crossreactivity of one antibody with a muscle protein. J. Histochem. Cytochem. 37, 159-164.

Thomas,S.A. and Segal,M.B. (1998). The transport of the anti-HIV drug, 2',3'-didehydro-3'-deoxythymidine (D4T), across the blood-brain and blood-cerebrospinal fluid barriers. Br. J. Pharmacol. 125, 49-54.

REFERENCE LIST

119

Tiribelli,C. and Ostrow,J.D. (2005). The molecular basis of bilirubin encephalopathy and toxicity: report of an EASL Single Topic Conference, Trieste, Italy, 1-2 October, 2004. J. Hepatol. 43, 156-166.

Tishler,D.M., Weinberg,K.I., Hinton,D.R., Barbaro,N., Annett,G.M., and Raffel,C. (1995). MDR1 gene expression in brain of patients with medically intractable epilepsy. Epilepsia 36, 1-6.

Tomaro,M.L. and Batlle,A.M. (2002). Bilirubin: its role in cytoprotection against oxidative stress. Int. J. Biochem. Cell Biol. 34, 216-220.

Torok,M., Huwyler,J., Gutmann,H., Fricker,G., and Drewe,J. (2003). Modulation of transendothelial permeability and expression of ATP-binding cassette transporters in cultured brain capillary endothelial cells by astrocytic factors and cell-culture conditions. Exp. Brain Res. 153, 356-365.

Trambas,C., Wang,Z., Cianfriglia,M., and Woods,G. (2001). Evidence that natural killer cells express mini P-glycoproteins but not classic 170 kDa P-glycoprotein. Br. J. Haematol. 114, 177-184.

Tsai,C.E., Daood,M.J., Lane,R.H., Hansen,T.W., Gruetzmacher,E.M., and Watchko,J.F. (2002). P-glycoprotein expression in mouse brain increases with maturation. Biol. Neonate 81, 58-64.

Uhr,M., Holsboer,F., and Muller,M.B. (2002). Penetration of endogenous steroid hormones corticosterone, cortisol, aldosterone and progesterone into the brain is enhanced in mice deficient for both mdr1a and mdr1b P-glycoproteins. J. Neuroendocrinol. 14, 753-759.

van Aubel,R.A., Smeets,P.H., Peters,J.G., Bindels,R.J., and Russel,F.G. (2002). The MRP4/ABCC4 gene encodes a novel apical organic anion transporter in human kidney proximal tubules: putative efflux pump for urinary cAMP and cGMP. J. Am. Soc. Nephrol. 13, 595-603.

Van Bree,J.B., De Boer,A.G., Danhof,M., and Breimer,D.D. (1992). Drug transport across the blood--brain barrier. I. Anatomical and physiological aspects. Pharm. Weekbl. Sci. 14, 305-310.

Virgintino,D., Robertson,D., Errede,M., Benagiano,V., Girolamo,F., Maiorano,E., Roncali,L., and Bertossi,M. (2002). Expression of P-glycoprotein in human cerebral cortex microvessels. J. Histochem. Cytochem. 50, 1671-1676.

Vitek,L., Kotal,P., Jirsa,M., Malina,J., Cerna,M., Chmelar,D., and Fevery,J. (2000). Intestinal colonization leading to fecal urobilinoid excretion may play a role in the pathogenesis of neonatal jaundice. J. Pediatr. Gastroenterol. Nutr. 30, 294-298.

Volk,B., Hettmannsperger,U., Papp,T., Amelizad,Z., Oesch,F., and Knoth,R. (1991). Mapping of phenytoin-inducible cytochrome P450 immunoreactivity in the mouse central nervous system. Neuroscience 42, 215-235.

Watchko,J.F., Daood,M.J., and Hansen,T.W. (1998). Brain bilirubin content is increased in P-glycoprotein-deficient transgenic null mutant mice. Pediatr. Res. 44, 763-766.

Watchko,J.F., Daood,M.J., Mahmood,B., Vats,K., Hart,C., and hdab-Barmada,M. (2001). P-glycoprotein and bilirubin disposition. J. Perinatol. 21 Suppl 1, S43-S47.

REFERENCE LIST

120

Weisiger,R.A., Ostrow,J.D., Koehler,R.K., Webster,C.C., Mukerjee,P., Pascolo,L., and Tiribelli,C. (2001). Affinity of human serum albumin for bilirubin varies with albumin concentration and buffer composition: results of a novel ultrafiltration method. J. Biol. Chem. 276, 29953-29960.

Wennberg,R.P. (2000). The blood-brain barrier and bilirubin encephalopathy. Cell Mol. Neurobiol. 20, 97-109.

Wennberg,R.P., Ahlfors,C.E., Bhutani,V.K., Johnson,L.H., and Shapiro,S.M. (2006). Toward understanding kernicterus: a challenge to improve the management of jaundiced newborns. Pediatrics 117, 474-485.

Wijnholds,J., deLange,E.C., Scheffer,G.L., van den Berg,D.J., Mol,C.A., van,d., V, Schinkel,A.H., Scheper,R.J., Breimer,D.D., and Borst,P. (2000). Multidrug resistance protein 1 protects the choroid plexus epithelium and contributes to the blood-cerebrospinal fluid barrier. J. Clin. Invest 105, 279-285.

Wolburg,H., Wolburg-Buchholz,K., Kraus,J., Rascher-Eggstein,G., Liebner,S., Hamm,S., Duffner,F., Grote,E.H., Risau,W., and Engelhardt,B. (2003). Localization of claudin-3 in tight junctions of the blood-brain barrier is selectively lost during experimental autoimmune encephalomyelitis and human glioblastoma multiforme. Acta Neuropathol. 105, 586-592.

Wolburg,H., Wolburg-Buchholz,K., Liebner,S., and Engelhardt,B. (2001). Claudin-1, claudin-2 and claudin-11 are present in tight junctions of choroid plexus epithelium of the mouse. Neurosci. Lett. 307, 77-80.

Yamada,N., Sawasaki,Y., and Nakajima,H. (1977). Impairment of DNA synthesis in Gunn rat cerebellum. Brain Res. 126, 295-307.

Zhang,Y., Han,H., Elmquist,W.F., and Miller,D.W. (2000). Expression of various multidrug resistance-associated protein (MRP) homologues in brain microvessel endothelial cells. Brain Res. 876, 148-153.

Zhu,H.J. and Liu,G.Q. (2004). Glutamate up-regulates P-glycoprotein expression in rat brain microvessel endothelial cells by an NMDA receptor-mediated mechanism. Life Sci. 75, 1313-1322.

Zlokovic,B.V. and Apuzzo,M.L. (1998). Strategies to circumvent vascular barriers of the central nervous system. Neurosurgery 43, 877-878.

Zucker,S.D., Goessling,W., and Hoppin,A.G. (1999). Unconjugated bilirubin exhibits spontaneous diffusion through model lipid bilayers and native hepatocyte membranes. J. Biol. Chem. 274, 10852-10862.

Reprint of the Papers

121

Reprint of the Papers

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150