Investigations of Hepatic Hemodynamics and Alterations in the NO-cGMP Pathway in an Animal Model of Liver Fibrosis / Cirrhosis Suggest PDE5 Inhibitors as Promising Adjunct in Portal Hypertension Therapy INAUGURALDISSERTATION zur Erlangung des Doktorgrades (Dr. rer. nat.) der Fakultät für Chemie und Pharmazie der Albert-Ludwigs-Universität Freiburg im Breisgau vorgelegt im Jahr 2018 von Denise Schaffner geboren in Breisach am Rhein
173
Embed
Investigations of Hepatic Hemodynamics and Alterations in ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Investigations of Hepatic Hemodynamics and Alterations in the
NO-cGMP Pathway in an Animal Model of Liver Fibrosis / Cirrhosis Suggest PDE5 Inhibitors as Promising Adjunct in Portal Hypertension Therapy
INAUGURALDISSERTATION
zur Erlangung des Doktorgrades (Dr. rer. nat.)
der Fakultät für Chemie und Pharmazie
der Albert-Ludwigs-Universität Freiburg im Breisgau
vorgelegt im Jahr 2018
von
Denise Schaffner
geboren in Breisach am Rhein
Vorsitzender des Promotionsausschusses: Prof. Dr. Stefan Weber
Dekan: Prof. Dr. Manfred Jung
Referentin: Prof. Dr. Irmgard Merfort
Korreferent: Prof. Dr. Wolfgang Kreisel
Drittprüfer: Prof. Dr. Andreas Bechthold
Datum der mündlichen Prüfung: 29. Juni 2018
“With man this is impossible,
but with God all things are possible”
- Matthew 19:26 -
Scientific Activities
Publications D. Schaffner, D. von Elverfeldt, P. Deibert, A. Lazaro, I. Merfort, L. Lutz, J. Neubauer, M.W. Baumstark, W. Kreisel, W. Reichardt:
“Phase-contrast MR flow imaging: A tool to determine hepatic hemodynamics in rats with a healthy, fibrotic, or cirrhotic liver” J Magn Reson Imaging, 46(5), 1526-1534 (2017)
D. Schaffner, A. Lazaro, P. Stoll, P. Deibert, I. Merfort, A. Schmitt-Gräff, M.W. Baumstark, L. Vauth, P. Hasselblatt, W. Kreisel:
„Analysis of the NO-cGMP pathway in experimental liver cirrhosis suggests phosphodiesterase 5 as potential target in portal hypertension therapy” Manuscript in preparation
Short Oral Presentations D. Schaffner, D. von Elverfeldt, P. Deibert, A. Lazaro, I. Merfort, L. Lutz, J. Neubauer, M.W. Baumstark, W. Kreisel, W. Reichardt:
“Effect of Chronic Thioacetamide Treatment on Hepatic Hemodynamic Parameters in Rats: Evaluation by Magnetic Resonance Imaging” United European Gastroenterology (UEG) week 2016, Vienna, Austria
D. Schaffner, A. Lazaro, P. Deibert, I. Merfort, A. Schmitt-Gräff, P. Hasselblatt, W. Kreisel:
„Störungen des NO-cGMP-Systems im Tiermodell einer Leberzirrhose: Implikationen für die Therapie der portalen Hypertonie beim Menschen“ Annual Meeting of the German Society of Gastroenterology, Digestive and Metabolic Diseases (Deutsche Gesellschaft für Gastroenterologie, Verdauungs- und Stoffwechselkrankheiten, DGVS) 2017, Dresden, Germany
Poster Presentations D. Schaffner, D. von Elverfeldt, P. Deibert, A. Lazaro, I. Merfort, L. Lutz, J. Neubauer, M.W. Baumstark, W. Kreisel, W. Reichardt:
“Phase-contrast Magnetic Resonance Flow Imaging: A Tool to Determine Hepatic Hemodynamics in Rats with a Healthy, Fibrotic, or Cirrhotic Liver” Annual Meeting of the International Society for Magnetic Resonance in Medicine (ISMRM) 2017, Honolulu, Hawaii, US
D. Schaffner, A. Lazaro, P. Deibert, M.W. Baumstark, I. Merfort, W. Kreisel:
“Investigation on Hepatic Hemodynamics in Animal Model of Liver Cirrhosis” Day of Research 2017, Faculty of Chemistry and Pharmacy, University Freiburg, Germany
D. Schaffner, A. Lazaro, P. Deibert, I. Merfort, A. Schmitt-Gräff, P. Hasselblatt, W. Kreisel:
“NO-cGMP Pathway Alterations may contribute to Portal Hypertension: Results of a Study in Rats with Liver Fibrosis/Cirrhosis” Annual Meeting of the American Association for the Study of Liver Diseases (AASLD), Liver Meeting 2017, Washington D.C., US
D. Schaffner, A. Lazaro, P. Deibert, I. Merfort, A. Schmitt-Gräff, P. Hasselblatt, W. Kreisel:
“Alterations of the NO-cGMP pathway in thioacetamide-induced liver fibrosis/cirrhosis in rats” United European Gastroenterology (UEG) week 2017, Barcelona, Spain
D. Schaffner, A. Lazaro, P. Deibert, I. Merfort, A. Schmitt-Gräff, P. Hasselblatt, W. Kreisel:
“The NO – cGMP Pathway in Experimental Liver Cirrhosis – Implications for Portal Hypertension Therapy” Day of Research 2017, Faculty of Medicine, University Hospital Freiburg, Germany
D. Schaffner, A. Lazaro, P. Hasselblatt, A. Schmitt-Gräff, M. Grosse-Perdekamp, I. Merfort, P. Deibert, W. Kreisel:
“Overexpression of Phosphodiesterase-5 in Liver Cirrhosis: A Rationale for Novel Therapy in Portal Hypertension” Digestive Disease Week® (DDW) 2018, Washington, D.C., US
Index 1. Summary .................................................................................................................... 1
2.1 The Liver - A Multifunctional Organ ........................................................................................................ 5
2.2 Hepatic Circulatory System ..................................................................................................................... 6
2.3 Regulatory Mechanisms of Hepatic Hemodynamics ........................................................................... 7
2.4 Regulatory Mechanisms of Hepatic Blood Flow ................................................................................... 7
2.5 Liver Cirrhosis ............................................................................................................................................ 9 2.5.1 Definition and Complications ........................................................................................................... 9 2.5.2 Epidemiology and Etiology ............................................................................................................ 10 2.5.3 Pathophysiology of Liver Fibrosis / Cirrhosis .............................................................................. 11 2.5.4 Pathophysiology of Portal Hypertension (PH) ............................................................................ 13
2.5.4.1 Cellular and Molecular Changes .......................................................................................... 14 2.5.5 Symptoms of Liver Cirrhosis and PH ........................................................................................... 17 2.5.6 Diagnosis and Classification of Liver Cirrhosis and PH ............................................................ 17 2.5.7 Therapy of Liver Cirrhosis and PH ............................................................................................... 21
2.5.7.1 NO – A Multifunctional Molecule .......................................................................................... 24 2.5.7.2 NO – Generation and Function ............................................................................................. 24 2.5.7.3 NO and NOS in the Pathophysiology of PH ....................................................................... 28 2.5.7.4 Strategies to Increase NO Availability and NO-cGMP Signaling ..................................... 29
2.5.8 PDE5 and PDE5 inhibitors ............................................................................................................ 31
3.1 Evaluation of the TAA Model ................................................................................................................. 38 3.1.1 General Remarks ............................................................................................................................ 38 3.1.2 Histological Assessment of the Degree of Liver Fibrosis .......................................................... 39 3.1.3 Mortality ............................................................................................................................................ 40
3.2 Noninvasive Hemodynamic Measurements ........................................................................................ 40 3.2.1 MR Assessment of the Degree of Liver Fibrosis ........................................................................ 42 3.2.2 Flow Velocity Patterns and Flow Curves ..................................................................................... 42 3.2.3 Hemodynamic Parameters ............................................................................................................ 43
3.3 Invasive Hemodynamic Measurements ............................................................................................... 46 3.3.1 Portal Flow Volume Rate ............................................................................................................... 47 3.3.2 Effect of Sildenafil on Hemodynamics ......................................................................................... 49 3.3.3 Effect of MAP on PVP ................................................................................................................... 55
12. Curriculum Vitae .................................................................................................. 159
Summary
1
1. Summary During the last 30 years phosphodiesterase 5 (PDE5) inhibitors had been
successfully integrated in the therapy of diseases with an underlying vascular
impairment, such as erectile dysfunction and pulmonary hypertension. Hence, the
use of PDE5 inhibitors is also considered as promising adjunct in the therapy of
portal hypertension (PH), one of the most crucial complications of liver cirrhosis, a
leading cause of death worldwide.
PH is associated with nitric oxide (NO) deficiency in the intrahepatic vasculature,
resulting in increased sinusoidal intrahepatic resistance. The latter is caused by a
mechanical and a functional component. However, up to now no drugs have been
approved to target the mechanical component, which occurs e.g. in the form of
fibrous connective tissue or regenerative nodules, responsible for around 70% of
increased intrahepatic resistance. The residual 30% is explained by the functional
component, which is determined by sinusoidal vasoreactivity. Impaired sinusoidal
vasoreactivty, in turn, can be caused by alterations in the key parameters of the nitric
nitric oxide-cyclic guanosine monophosphate (NO-cGMP) pathway, a regulator of
vascular tone. PDE5 is one of these key parameters involved in the NO-cGMP
pathway, initiating cGMP inactivation and thus leads to vasoconstriction. Hence,
pharmaceutical inhibition of PDE5 is a promising option to counteract sinusoidal
vasoconstriction and increased intrahepatic resistance. Initial preclinical and clinical
hemodynamic studies however, showed variable results considering the effect of
PDE5 inhibitors. Therefore, in this thesis the potential of PDE5 inhibitors in PH
therapy was further elucidated based on hemodynamic measurements and
biochemical investigations.
A rat model of thioacetamide (TAA)-induced liver fibrosis/cirrhosis was established
and noninvasive magnetic resonance (MR) measurements of hepatic and systemic
hemodynamics in rats with healthy, fibrotic or cirrhotic livers were performed. Liver
disease-induced changes in hemodynamic parameters, emphasizing on portal flow
volume rate, were determined. A significant decrease in portal flow volume rate was
observed in diseased rats, which was validated by subsequent invasive
hemodynamic measurements with a flow probe.
Summary
2
Moreover, dose-dependent effects of the PDE5 inhibitor sildenafil on hepatic and
systemic hemodynamics were investigated using pressure transducers. Acute effects
of administration of either sodium chloride, sildenafil 0.1 mg/kg or sildenafil 1.0 mg/kg
were compared. After high-dosage sildenafil administration (1.0 mg/kg), a trend
towards decreased portal venous pressure (PVP), a significant decrease in heart rate
(HR), and a nonsignificant decrease in mean arterial pressure (MAP) were found in
rats with cirrhotic livers. Hemodynamic data also revealed a significant effect of MAP
on PVP among all subgroups regardless of intervention, suggesting that changes in
systemic blood pressure may lead to changes in hepatic blood pressure.
Additionally, biochemical analyses of the key parameters in the NO-cGMP pathway
were conducted. Hepatic gene expression of the enzymes endothelial and inducible
NO synthase (eNOS, iNOS), soluble guanylyl cyclase subunit a1 and b1 (sGCa1,
sGCb1) and phosphodiesterase 5 (PDE5) was analyzed by qRT-PCR. An up-
regulation of iNOS and a significant overexpression of PDE5 in diseased rats were
observed. Enhanced levels of PDE5 protein expression were confirmed
immunhistochemically. Furthermore, serum cGMP concentrations from carotid
arterial blood samples were determined by ELISA. In diseased rats a slight decrease
was observed, whereas sildenafil administration (1.0 mg/kg) nearly renormalized
serum cGMP concentrations. Finally, studies were performed to evaluate whether the
hemodynamic measurement and the associated operative procedure affected gene
expression or serum cGMP concentrations. A significant decrease in eNOS gene
expression was detected.
In summary, the results of this study contribute to the general understanding of the
pathophysiology of PH and highlight the valuable potential of PDE5 inhibitors as
promising adjunct in PH therapy.
Summary
3
1. ZUSAMMENFASSUNG In den letzten 30 Jahren wurden Phosphodiesterase 5 (PDE5)-Inhibitoren erfolgreich
in die Therapie von Erkrankungen mit einer zugrundeliegenden vaskulären
Beeinträchtigung, wie z.B. erektile Dysfunktion und pulmonale Hypertonie, integriert.
Daher wird der Einsatz von PDE5-Inhibitoren auch in der Therapie der portalen
Hypertension (PH) als vielversprechender Zusatz angesehen. PH ist eine der
wesentlichsten Komplikationen der Leberzirrhose, eine der weltweit führenden
Todesursachen.
PH ist mit einem Stickstoffmonoxid (NO)-Mangel im intrahepatischen Gefäßsystem
assoziiert, was zu einem erhöhten sinusoidalen intrahepatischen Widerstand führt.
Letzteres wird durch eine mechanische und eine funktionelle Komponente
verursacht. Bis jetzt wurden jedoch keine Arzneimittel zugelassen, die auf die
mechanische Komponente abzielen, welche z.B. in Form von fibrösem Bindegewebe
oder regenerativen Knötchen auftritt und für etwa 70% des erhöhten intrahepatischen
Widerstandes verantwortlich ist. Die restlichen 30% erklären sich durch die
funktionelle Komponente, die durch sinusoidale Vasoreaktivität bestimmt wird. Eine
gestörte sinusoidale Vasoreaktivität kann wiederum durch Veränderungen in den
Schlüsselparametern der Stickstoffmonoxid-cyclisches Guanosinmonophosphat (NO-
cGMP)-Signalkaskade, einem Regulator des vaskulären Tonus, verursacht werden.
PDE5 ist einer dieser Schlüsselparameter in der NO-cGMP-Signalkaskade, der für
die Inaktivierung von cGMP verantwortlich ist und somit zur Vasokonstriktion führt.
Aus diesem Grund stellt die pharmazeutische Inhibierung von PDE5 eine
vielversprechende Option dar, um der sinusoidalen Vasokonstriktion und dem
erhöhten intrahepatischen Widerstand entgegenzuwirken. Erste präklinische und
klinische hämodynamische Studien zeigten hinsichtlich der Wirkung von PDE5-
Inhibitoren jedoch unterschiedliche Ergebnisse. Daher wurde in der vorliegenden
Arbeit das Potenzial von PDE5-Inhibitoren in der Therapie der PH auf der Grundlage
von hämodynamischen Messungen sowie biochemischen Analysen untersucht.
Es wurde ein Ratten-Modell der Thioacetamid (TAA)-induzierten Leberfibrose/-
zirrhose etabliert und nichtinvasive Magnetresonanz (MR)-Messungen der
hepatischen und systemischen Hämodynamik in Ratten mit gesunden, fibrotischen
oder zirrhotischen Lebern durchgeführt. Dadurch sollten die durch
Lebererkrankungen induzierten Veränderungen der hämodynamischen Parameter,
Summary
4
unter besonderer Berücksichtigung der portalen Volumenflussrate, bestimmt werden.
Bei erkrankten Ratten konnte eine signifikante Abnahme der portalen
Volumenflussrate, welche durch die nachfolgenden invasiven hämodynamischen
Messungen mit einer Strömungssonde bestätigt wurde, beobachtet werden.
Zudem wurden dosisabhängige Effekte des PDE5-Inhibitors Sildenafil auf die
hepatische und systemische Hämodynamik mittels Drucksensoren untersucht. Die
Effekte einer Verabreichung von entweder Natriumchlorid, Sildenafil 0,1 mg/kg oder
Sildenafil 1,0 mg/kg wurden verglichen. Nach Verabreichung von hoch-dosiertem
Sildenafil (1,0 mg/kg) wurde bei Ratten mit zirrhotischen Lebern ein Trend zu
verringertem Pfortaderdruck (PVP), eine signifikante Abnahme der Herzfrequenz
(HR) und eine nicht signifikante Abnahme des mittleren arteriellen Blutdrucks (MAP)
beobachtet. Zudem wurde anhand der hämodynamischen Daten bei allen
Untergruppen, unabhängig von der Intervention, ein signifikanter Effekt des MAP auf
den PVP ermittelt. Dies deutet darauf hin, dass Veränderungen des systemischen
Blutdrucks zu Veränderungen des hepatischen Blutdrucks führen können.
Darüber hinaus wurden biochemische Analysen der Schlüsselparameter der NO-
cGMP-Signalkaskade durchgeführt. Die hepatische Genexpression der Enzyme
endotheliale und induzierbare NO-Synthase (eNOS, iNOS), lösliche Guanylyl-
Cyclase-Untereinheit a1 und b1 (sGCa1, sGCb1) und Phosphodiesterase 5 (PDE5)
wurde mittels qRT-PCR analysiert. Dabei zeigte sich eine Hochregulation von iNOS
und eine signifikante Überexpression von PDE5 bei erkrankten Ratten. Letzteres
wurde durch immunhistochemische Untersuchungen der PDE5-Proteinexpression
validiert. Außerdem wurden Serum-cGMP-Konzentrationen aus Blutproben der
Halsschlagader mittels ELISA bestimmt. In erkrankten Ratten wurde eine leichte
Abnahme beobachtet. Die Verabreichung von Sildenafil (1,0 mg/kg) führte dagegen
fast zu einer Renormierung der Serum-cGMP-Konzentrationen. Abschließend wurde
untersucht, ob die hämodynamische Messung und der damit verbundene operative
Eingriff die Genexpression oder Serum-cGMP-Konzentrationen beeinflussten.
Hierbei wurde eine signifikante Abnahme der eNOS-Genexpression nachgewiesen.
Zusammenfassend tragen die Ergebnisse dieser Studie zum allgemeinen
Verständnis der Pathophysiologie der PH bei und verdeutlichen das Potenzial von
PDE5-Inhibitoren als vielversprechenden Zusatz in der Therapie der PH.
Introduction
5
2. Introduction 2.1 The Liver - A Multifunctional Organ The liver is the largest gland in the human organism, and the second largest organ
after the skin 1. It is segmented into lobes, reddish-brown in color, and has a soft
consistency. Its central location in the upper-right portion of the abdomen, beneath
the diaphragm and to the right of the stomach, points out its importance for life.
The basic architectural unit of the liver is the hepatic lobule 2, where multiple
essential metabolic, detoxifying, and synthesizing processes take place:
• breaking down nutrients and turning them into energy
• storing glycogen, vitamins, iron and other essential chemicals
• controlling blood composition, i.e. levels of lipids, amino acids and glucose
• detoxifying potentially harmful substances, e.g. drugs and alcohol
• clearing the blood of particles and infections, e.g. toxins and bacteria
• converting ammonia to urea
• synthesizing immunologically active cells, plasma proteins and numerous
hormones
• synthesizing bile to digest lipids
• controlling blood clotting and repair of damaged tissues
To fulfill these tasks, the liver, together with its circulatory system and the associated
biliary duct, has evolved many structural and physiological features that underpin the
broad spectrum of critical functions. One major feature is functional liver tissue, which
encompasses at least seven different cell types. Among those, hepatocytes are the
major parenchymal cells, whereas sinusoidal endothelial cells (SECs),
cholanigocytes, as well as immunologically active cells such as hepatic stellate cells
(HSCs), Kupffer cells (KCs), natural killer cells (NKs) and lymphocytes of different
phenotypes are non-parenchymal cells 3. The most numerous cells are hepatocytes,
comprising 70-85% of the liver tissue 1,4. Other unique features of the liver are its
capacity for self-regeneration and its complex dual circulatory system 2.
Introduction
6
2.2 Hepatic Circulatory System The circulatory system of the liver is supplied by two distinct circulatory routes: the
hepatic artery and the portal vein 5. Each route provides blood of differing
compositions: the hepatic artery delivers well-oxygenated blood, accounting for 25%
of hepatic blood, whereas the residual 75% are supplied by the portal vein, which
delivers deoxygenated, but nutrient-rich blood 6–8. Both routes enter the liver via the
portal tracts, which are components of the hepatic lobules, and finally merge in the
sinusoids (Figure 1). The latter are a specialized network of intrahepatic blood
vessels, representing the hepatic microcirculation system and resembling systemic
capillaries 4.
Sinusoids, which are considered to be the functional vascular unit of the liver, are
composed of SECs, KCs, and HSCs 9,10. SECs form a loose physical barrier between
the blood circulating within the sinusoids and hepatocytes lining the sinusoids 10.
SECs and hepatocytes are in turn separated by the so-called “space of Disse”, where
HSCs are located 4. KCs are mainly located in the sinusoidal lumen, but they can
also make direct contact with the hepatocytes 11.
Figure 1: Schematic diagram of a portal tract (left) and a hepatic sinusoid (right) Original source: Y .Iwakiri et al. 2014: “Vascular pathobiology in chronic liver disease and cirrhosis –
Current status and future dicrections” (https://doi.org/10.1016/j.jhep.2014.05.047)
This article was published under the terms of the Creative Commons Attribution-NonCommercial-No
Derivatives License (CC BY NC ND).
SECs are highly specialized endothelial cells unique to their location 12. In contrast to
other endothelial cells, SECs lack a continuous endothelial lining and exhibit a
fenestration, which makes them the most permeable endothelial cells of the
mammalian organism 12. This “sinusoidal gap” most likely serves to facilitate the
This static component is superimposed by the dynamic component, which is based
on locally acting regulatory mechanisms and regulatory mechanisms that adjust the
current hemodynamic status to the demands of the organism as a whole 17.
2.4 Regulatory Mechanisms of Hepatic Blood Flow As already mentioned, 75% of the hepatic blood is supplied by the portal vein. But
portal blood flow is in fact simply the sum of outflows of splanchnic organs, which
means that the liver is not capable of regulating portal blood flow directly 7,5.
Introduction
8
However, to counteract acute or chronic changes in portal blood flow, the liver
evolved several interrelated regulatory mechanisms, which primarily influence blood
flow to extrahepatic splanchnic organs 5. As a result, a constant hepatic blood flow-
to- liver mass ratio can be ensured under physiological conditions. The regulatory
mechanisms have been elucidated by Lautt 5 and are summarized in the following:
The first mechanism is vascular compliance, which is based on the physical principle
of a volume-pressure relationship. In general, vascular compliance describes the
extent to which the volume of the vessel passively changes with changes in
pressure. The vessel volume itself is controlled by vasodilation or vasoconstriction.
Thus, a decreased portal flow is followed by a passive decrease in intrahepatic
pressure and furthermore a passive blood extrusion from the huge hepatic blood
reservoir into the central venous circulatory system. Thereby, cardiac output is
increased, which leads to an elevation of blood flow in the splanchnic arteries that
feed the portal venous system. As a consequence the initial flow deficit is, at least
partially, buffered.
Another well-described regulatory mechanism of the liver is the hepatic arterial buffer
response (HABR) 7,18. The key player in the HABR is adenosine, a potent vasodilator
of the hepatic artery. Adenosine is constantly secreted into the space of Mall that
surrounds the terminal branches of the hepatic artery and the portal vein before they
finally merge in the liver sinusoids. If portal flow is decreased, adenosine
accumulates, resulting in dilation of hepatic artery being stimulated. The induced
increase in hepatic arterial flow into the portal vein buffers changes in portal flow on
total hepatic flow. In former publications Lautt (et al) described the HABR to be
capable of compensating a 25% to 60% decrease in portal blood flow 19,20, however,
in a more current publication, he stated that the hepatic arterial buffer capacity is
challenging to quantify 5. Interestingly, the HABR only works unidirectionally, since a
decrease of hepatic arterial flow does not induce an elevation of portal flow 21,7.
Accumulation of adenosine also indirectly mediates the activation of the hepatorenal
reflex via hepatic afferent nerves. This reflex induces a decrease in renal output and
fluid retention, thereby leading to an increase in blood volume, venous return, cardiac
output, and ulitmately splanchnic blood flow.
Introduction
9
Moreover, the liver has a unique way of counteracting severe vasoconstriction.
Looking at the hepatic artery, vasoconstriction leads to decreased hepatic arterial
flow. The portal vein in comparison, responds to local intrahepatic vasoconstriction
by an increase in portal venous pressure (PVP) with no alterations in portal flow,
since portal flow is controlled by the outflow of the splanchnic organs. In addition to
adenosine, nitric oxide (NO) is also a potent vasodilator and antagonist to
vasoconstrictors. NO-induced vasodilation of the portal vein as well as the hepatic
artery occurs when intrahepatic vasoconstriction enhances shear stress. In contrast,
adenosine-induced vasodilation occurs only when vasoconstriction is more systemic
and causes a decrease in portal flow. Its vasodilatory effects, however, seem to be
limited to the hepatic artery.
If all these compensatory mechanisms are not sufficient to maintain hepatic blood
flow homeostasis, in the last resort liver mass is adapted to match the blood demand.
Therefore, hepatocyte proliferation is induced when portal flow is elevated, whereas
hepatocyte apoptosis is induced when portal flow is reduced.
Considering this massive compensatory machinery, it becomes obvious how
significant an adequate hepatic blood flow is to sustain liver function. Nevertheless,
the occurrence of hemodynamic disturbances and vascular insult, e.g. in association
with liver cirrhosis, cannot be excluded 6.
2.5 Liver Cirrhosis 2.5.1 Definition and Complications Liver cirrhosis is a serious chronic liver disease. Its pathogenesis describes a
prolonged and creeping progress characterized by fibrosis development, or scarring,
and structural modifications of the liver architecture 22–24. Secondary to liver cirrhosis,
the occurrence of impaired liver function as well as a characteristic vascular disorder,
namely portal hypertension (PH), is likely 25,26. Impaired liver function leads to
increased blood values of bilirubin and ammonia, and decreased blood values of
albumin, cholinesterase and clotting factors. Along with PH, further complications
emerge, such as ascites, esophageal or gastric varices, variceal bleeding,
spontaneous bacterial peritonitis, or dysfunction of other organs 27–29. The latter can
appear in form of the hepatorenal syndrome, hepatopulmonal syndrome,
Introduction
10
portopulmonal hypertension, cirrhotic cardiomyopathy or hepatic encephalopathy 28,29. Hence, in advanced stages of liver cirrhosis not only is there the risk of needing
a liver transplantation, but the risk of morbidity and mortality also increases
immensely 30,28,23,31.
2.5.2 Epidemiology and Etiology Liver cirrhosis is one of the most frequent chronic liver diseases worldwide that
appears in rich as well as in poor nations 32,33,24. With more than one million deaths
per year (data from 2010), it is the 14th most common cause of death worldwide 34,35.
In central Europe, it is in fact the fourth most common cause of death with around
170 000 deaths per year (data from 2002) 36,37. That makes up approximately 2% of
all deaths worldwide, and also 2% of deaths in Europe 35,32,34. However, it seems
likely that there is a high number of unreported and / or undetected cases as the
initial stage of liver cirrhosis is asymptomatic or the disorder remains undiagnosed 38,39.
Liver cirrhosis can arise as a consequence of a range of chronic stimuli including
toxic, viral, autoimmune, vascular, cholestatic or metabolic diseases (Table 1) 22,40,41,31. Among those, alcoholic and, increasingly non-alcoholic liver diseases
(NAFLD), as well as hepatitis B or C infections, are the most common risk factors 42,43,38,40,36,44,45.
Table 1: Causes of liver cirrhosis
Stimuli Examples
toxic
infectious - viral - others
autoimmune
vascular
cholestatic
metabolic
alcohol-induced steatohepatitis, medications and chemicals
hepatitis B, C, D schistosomiasis and toxoplasmosis
autoimmune hepatitis, primary sclerosing cholangitis and primary biliary cholangitis
including apoptosis, senescence and reversion to quiescent HSCs, and are thus
essential for fibrosis regression 56,47,57.
Due to these dual roles, recruited macrophages are major regulators of liver fibrosis
progression and resolution 46,57. However, KCs as well as recruited macrophages can
adopt their phenotype, depending on signals from the hepatic microenvironment,
making the role of the immune system in reversibility of hepatic fibrosis even more
complex 56,57.
The perpetuation phase starts once the HSCs are activated and aim to maintain their
activated phenotype, which is characterized by various changes in cell behavior and
properties. Whereas quiescent HSCs primarily serve as vitamin A reservoirs, the
activated phenotype shows acquisition of ECM-generating, contractile, proliferative,
migratory, immunomodulatory and phagocytic properties and simultaneously a loss
of vitamin A storage capacity 47,50,54.
The activated phenotype of HCSs, the myofibroblasts, are the principle source of
ECM constituents, including collagen. Moreover, myofibroblasts synthesize tissue
inhibitors of matrix metalloproteinases (TIMPs), which are secreted into the
extracellular environment to inhibit matrix metalloproteinases (MMPs), a family of
ECM-degrading enzymes 42. Being released from infiltrating macrophages and KCs,
MMPs are present in the liver even during progressive fibrogenesis, demonstrating
that ECM accumulation by far exceeds its degradation by MMPs 26.
Initially the encapsulation of inflamed or damaged liver tissue by ECM indeed
represents a beneficial mechanism in the wound healing process and ensures liver
repair; however, when the stimulus for wound healing remains sustained persistently,
fibrogenesis escalates. At first fibrosis develops around either portal tracts or central
veins, ultimately forming bridging fibrosis with nodule formation surrounded by thick
bands of fibrous connective tissue 48,26,24. As a consequence of ongoing distortion of
liver architecture, the transition from liver fibrosis into cirrhosis takes place.
Introduction
13
2.5.4 Pathophysiology of Portal Hypertension (PH) Secondary to liver cirrhosis one of the earliest and most crucial complication is PH,
which is characterized by an abnormally increased PVP 58,24,59. Defined clinically, the
term “PH” describes an increase of the hepatic venous pressure gradient (HVPG)
between the portal vein and the inferior vena cava above normal values (≥ 5 mmHg) 60. It is accompanied by distinct alterations not only in the intra-, but also in the
extrahepatic circulation, and underlies most of the clinically significant complications
of liver cirrhosis 61. The intrahepatic, sinusoidal PH (see 2.5.6), the most common
form occurring secondary to liver cirrhosis, will be focused on in the following 62,63.
When considering the pathophysiology of PH the first concept that needs to be
readdressed is the hemodynamic application of Ohm's Law 60:
2.5.5 Symptoms of Liver Cirrhosis and PH Since it can take years or even decades until liver cirrhosis causes any obvious signs
or symptoms, it is not surprising that in many affected people the disease remains
undiagnosed. The interval of progression from liver damage to liver cirrhosis seems
to be highly individual despite the same etiology; in some cases, the process can
take 40 years (slow fibrosers), whereas in others it can take less than 15 years (rapid
fibrosers). Accordingly, incidental liver screening tests, i.e. laboratory tests or
examinations using imaging modalities often lead to the diagnosis of liver cirrhosis in
an early stage than the disease itself. In a fairly advanced stage, it is much more
likely that the disease is diagnosed as a consequence of the occurrence of PH and
other clinically significant complications. 78,38,31,39
2.5.6 Diagnosis and Classification of Liver Cirrhosis and PH Once there are symptoms or indications of liver cirrhosis, defining the underlying
etiology and the stage of the disease is essential for the choice of therapy and the
prediction of the prognosis. Liver fibrosis per se can occur as a consequence of any
chronic liver disease regardless of etiology 42,47. However, the predominant
profibrogenic mechanisms, as well as the patterns of parenchymal damage indeed
vary with the etiology of the underlying liver disease 79. The etiology can be identified
Introduction
18
by the patient’s history combined with laboratory tests and histological examinations 39. Histological examinations are furthermore considered to be the reference standard
for the assessment of the degree of liver fibrosis, although this involves the invasive
procedure of a biopsy 80,30. In addition, laboratory tests and imaging modalities, e.g.
ultrasound (US) or magnetic resonance (MR) imaging can be used for the
assessment of the degree of liver fibrosis. In the following, only the reference
standard will be focused on.
Liver biopsy can be carried out from a percutaneous or a transjugular route under
local anesthesia 80. After having taken the liver tissue samples, cross-sections are
performed before liver tissue sections are evaluated histologically. For the evaluation
of the grade, measuring of necro-inflammatory activity, and the stage, measuring
fibrosis and architectural changes, several histological scores exist 81. One example
is the Desmet score (DS) with DS=0: no fibrosis, DS=1: mild fibrosis, DS=2:
moderate fibrosis, DS=3: severe fibrosis, and DS=4: cirrhosis 82. But regardless of
the scoring system used, from a histopathological perspective the diagnosis of
cirrhosis is established once liver fibrosis has reached its terminal stage and the
process is considered “end-stage” 59. Moreover, for many years only liver fibrosis was
regarded as a dynamic, and potentially reversible process, whereas liver cirrhosis
was described as a static and irreversible terminal disease 31,83. However, nowadays
the concept of a dynamic, and at least partly reversible multi-stage process for liver
cirrhosis is being increasingly accepted 38,61,59.
The course of liver cirrhosis can initially be classified into two major stages: a
compensated, or asymptomatic phase, followed by a rapidly progressive
decompensated stage 23. The decompensated stage is defined by the presence of
clinical complication events secondary to PH, such as variceal bleeding, ascites or
hepatic encephalopathy 23,38,59. PH can be categorized according to anatomical
location into either pre-, intra- or posthepatic, with the intrahepatic, sinusoidal PH
being the most common form secondary to liver cirrhosis, regardless of etiology
(Table 3) 62,63. Since prognosis and predictors of death differ between these two
major stages of compensated and decompensated cirrhosis, each of them should be
regarded as separate entities 58. In fact, much effort has been put into the clarification
of the predominant pathogenic mechanisms of PH in each stage, which finally led to
the discovery of further substages of cirrhosis. Referring to recent publications of
Introduction
19
Abrades et al 58, D’Amico et al 84, and Garcia-Tsao et al 68, five prognostic stages
with a significant increase in the risk of death can be proposed (Table 4).
Table 3: Classification of PH according to anatomical location
Classification Subclassification
prehepatic
intrahepatic
posthepatic
- congential portal atresia
- intraluminal obstruction (thrombus, neoplasia)
- extraluminal vascular compression
- presinusoidal
- sinusoidal
- postsinusoidal
- luminal vascular obstruction
- extraluminal vascular compression
Table 4: Stages of liver cirrhosis
Stage Definition
1
2a
2b
3
4
5
compensated cirrhosis with mild PH
compensated cirrhosis with clinically significant PH, no varices
compensated cirrhosis with clinically significant PH, and varices
(no bleeding)
bleeding without other disease complications
first non-bleeding decompensating even
any second decompensating event
Introduction
20
As the majority of complications are caused by PH, the diagnosis of liver cirrhosis
often implicates the necessity to evaluate PVP. In a clinical setting the reference
standard to assess PVP is the hepatic venous pressure gradient (HVPG), which
represents an indirect measurement of PVP. Also non-invasive imaging modalities,
e.g. ultrasound (US) or magnetic resonance (MR) imaging can be used to assess
portal hemodynamics, but not PVP. However, the use of these imaging techniques is
still a matter of debate.
The assessment of the HVPG has essential prognostic relevance that even might
exceed that of histological examinations 80,61. The determination of the HVPG
requires the measurement of two pressure values: the wedged (or occluded) hepatic
venous pressure (WHVP) and the free hepatic venous pressure (FHVP). To measure
WHVP, a balloon catheter is inserted under local anesthesia through the jugular,
femoral or cubital vein into the hepatic vein. Through inflation of the balloon, the
hepatic venous outflow is blocked. After 1 to 2 minutes of blockade, the pressure at
the tip of the catheter finally reflects that of the hepatic sinusoidal pressure. On the
other hand, to measure FHVP, the balloon is deflated at 2 to 3 cm from the hepatic
vein ostium, so that the pressure at the tip of the catheter usually reflects the
pressure in the inferior vena cava.80
HVPG is finally calculated as the difference between WHVP and FHVP and hence
represents the pressure gradient between the portal vein and the intraabdominal
portion of inferior vena cava:
HVPG = WHVP – FHVP. 80,85,86
Since the WHVP, and accordingly the HVPG, is a measure of sinusoidal pressure, it
is important to mention that this measurement does not deliver reliable data with
respect of prehepatic or presinusoidal PH 68. In intrahepatic, sinusoidal PH however,
the HVPG is a reliable diagnostic tool which gives an accurate estimation of PVP. It
can be interpreted as follows: A HVPG < 5 mmHg is considered to be normal,
whereas PH is defined as an HVPG > 5 mmHg, a HVPG > 5 but < 10mmHg being
defined as mild PH, and a HVPG ≥ 10 mmHg as clinically significant PH. Above this
threshold of 10 mmHg, all complications induced by PH are more likely to occur. 24,62,68
Introduction
21
An accurate evaluation of hepatic, but also systemic hemodynamic status in chronic
liver diseases is thus essential for prevention or therapy of PH and its complications.
2.5.7 Therapy of Liver Cirrhosis and PH The main aim in PH therapy is to reduce HVPG to less than 12 mmHg or at least
20% of baseline to reduce the risk of variceal bleeding or rebleeding 38,87,27. To attain
this goal, PH management ideally involves addressing the underlying etiology,
inhibiting fibrosis development and regression, diminishing intrahepatic resistance
and / or splanchnic vasodilation, and treating complications 27,67. According to the
Baveno guidelines, PH management can involve pharmaceutical, endoscopic and
mechanical therapies 88.
As a first step in PH management, correct identification and extinguishing of the
origin of the evil is essential, since clearance or control of the underlying etiology of
liver damage is always the most effective therapy 31,42,89. However, in many affected
people the primary event or relevant mediators cannot be eliminated. In addition,
since affected people commonly only appear at an advanced stage of the disease,
reversal may not be rapid enough to prevent complications 89.
In suspected variceal bleeding, pharmaceutical therapy with vasoactive substances
should be started as soon as possible, before endoscopic therapies, such as band
ligation or sclerosing, are applied 88. In acute esophageal variceal bleeding events
however, a combined pharmaceutical and endoscopic therapy is recommended 88.
When endoscopic therapies are applied, it should be considered that they indeed
help to stop the bleeding, but simultaneously lead to an enhancement of PVP,
thereby worsening PH.
In a very advanced stage, when initial pharmaceutical and endoscopic therapy show
no effect or are likely to show no effect, transjugular intrahepatic portosystemic stent
shunting (TIPS) presents another therapy option 90. This mechanical, minimal
invasive therapy lowers PVP, but at the same time increases the risk of serious side
effects, such as the development of hepatic encephalopathy. In the event that all
these therapies fail, radical treatment by liver transplantation is the only remaining
option to increase survival odds 28,31. Since donors for liver transplantations are rare,
and the current therapy options are far from satisfying, new approaches are urgently
needed 91,92. In the following, the focus will be on pharmaceutical therapies.
Introduction
22
Current pharmaceutical therapy is mainly stratified according to the presence and
characterization of esophageal varices, meaning that a complication rather than the
disease itself is treated 88,90,91. Hence, research and also pharmaceutical companies
have been working intensively on the development of novel drugs to improve PH
therapy. Some aimed at developing antifibrotic drugs to reverse or at least inhibit
fibrogenesis, but no drug has yet been approved for use in humans 42,48,31.
Consequently, treating PH it is still challenging, since up to now the mechanical
component of increased intrahepatic resistance remains mostly irreversible. The
good news is that the functional component of PH can indeed be targeted
pharmaceutically and might potentially improve the future management of PH 42,93.
The functional component of increased intrahepatic resistance can be influenced
positively, either by a decrease in intrahepatic vascular tone, a decrease in
splanchnic vasodilation, or ideally both (Table 5) 67.
The current reference standard in pharmaceutical therapy, i.e. nonselective beta
blockers (NSBBs, beta-adrenergic receptor antagonists), vasopressin derivatives or
intestinal hormones, mainly counteract splanchnic vasodilation. Oral administration of
NSBBs is recommended to prevent bleeding, whereas in acute variceal bleeding
events, vasopressin derivatives or intestinal hormones should be administered
intravenously to stop bleeding 91. However, since these vasoactive substances not
only affect intrahepatic, but also extrahepatic circulation and hence can cause
massive contrary effects, their use has always been a matter of debate 91,94. Looking
for better alternatives, modulating NO availability and / or NO downstream signaling
seem to be promising options, since NO plays a vital role in the pathophysiology of
PH.
Introduction
23
Table 5: Reference standards and potentially novel drugs for PH therapy
Mode of action Drug group and names
reduced splanchnic vasoconstriction
reduced intrahepatic resistance (intrahepatic vascular tone ↓)
- nonselective ß-blockers (NSBBs)
propaponol, nadolol and carvedilol only in acute variceal bleeding events: - vasopressin derivatives
benazepril and captopril / losartan and valsartan - statins (HMG-CoA-reductase-inhibitors)
simvastatin and atorvastatin
- PDE5 inhibitors
sildenafil, udenafil and vardenafil - endothelin-receptor-antagonists
ambrisentan, bosentan and macitentan
Introduction
24
2.5.7.1 NO – A Multifunctional Molecule NO is an unstable free radical with a short biological half-life 95. First of all, NO is
known to be a potent endothelium-derived vasodilator, but it is likewise involved in
various other physiological processes in the cardiovascular, immune, gastrointestinal,
genitourinary, respiratory and nervous systems 96–99.
After NO is generated, it quickly diffuses into surrounding cells, where it can interact
with different reactants, such as transition metals and free radicals, and affect
proteins, nucleic acids, as well as fatty acids 96,100,101. What kind of interactions are
finally favored depends on several factors like the cellular environment, the available
concentration of NO and reactants and the reaction rates 96,98,102,103. Its physiological
effects are caused either directly or indirectly by its reactive and radical nature 96,104,100. Its unique chemistry, specifically the unpaired electron, but also the fact of
nitrogen being able to reach various oxidation states to generate different reactive
nitrogen species (RNS), vastly raises the potential NO effects 101,98. Thus, it is still
challenging to specify its physiological effects in specific cell types or complex
neuronal assembles 105. However, a well-studied and recognized NO target is the
soluble guanylyl cyclase (sGC) 102,106,99, a key cytosolic enzyme in the NO-cGMP
signaling pathway. The activation of this pathway implicates vasodilation and is
therefore essential for vasoregulation, including vascular tone and resistance.
2.5.7.2 NO – Generation and Function The activation of the NO-cGMP pathway takes place once NO is generated and
diffuses into the cytoplasm of surrounding cells, where it binds to the enzyme soluble
guanylyl cyclase (sGC). The interaction of NO with sGC causes a conformational
change, which results in the catalytic conversation of guanosine-5’-triphosphate
(GTP) to cyclic guanosine-3’,5’-monophosphate (cGMP). cGMP, an intracellular
second messenger, triggers various downstream signaling effects, which induce
vasodilatation. (Figure 3)
Indeed, the NO-cGMP pathway is much more complex. The three main enzymatic
steps NO generation, cGMP generation and degradation will be described in more
detail.
Introduction
25
NO generation occurs in a broad number of different cell types; however, to regulate
vascular tone, its synthesis in ECs of the vascular endothelium, and in case of the
corpus cavernosum also in neurons, is particularly important. Both, biomechanical
and biochemical stimuli, such as shear stress, VEGF and bradykinin can precipitate
NO generation 100,107,108,63,109. The synthesis itself can occur in two different ways:
either non-enzymatically from the transformation or degradation of inorganic nitrogen
chemicals in the organism and diet, or enzymatically from the oxidation of L-Arginine
to NO and L-citrulline 95,110. In mammals, the enzymatic redox reaction can be
catalyzed by three different isoforms of the enzyme nitric oxide synthase (NOS),
which were named according to the cell type or condition first described: endothelial
NOS (eNOS), inducible or inflammatory NOS (iNOS) and neuronal NOS (nNOS) 98,102.
All NOSs differ slightly in expression profile and in physiological function: eNOS and
nNOS, are both expressed progressively and generate continuous, but moderate
amounts of NO. eNOS is primarily expressed in ECs and primarily regulates vascular
tone. In addition, it induces vasoprotective and anti-atherosclerotic effects 104. nNOS
is primarily expressed in neurons and skeletal muscle and is responsible for synaptic
plasticity in the central nervous system, central regulation of blood pressure, smooth
muscle relaxation and vasodilation via peripheral nitrergic nerves 98. These nerves
are involved in the relaxation of corpus cavernosum and penile erection 104. iNOS
expression was originally identified in macrophages. Later, however, it was
demonstrated in almost all cell types as a defense mechanism against infections
from invading bacteria, viruses and fungi or against inflammation 97,111,112. Since
iNOS up-regulation is usually a consequence of pathological conditions, induction of
iNOS expression generates huge amounts of NO. The cell-specific roles of iNOS-
derived NO, however, need further investigation. Under physiological conditions,
iNOS expression is minimal or even absent 113,114.
Regarding the liver, eNOS and iNOS are the major players, whereas only little is
known about the role of nNOS in this organ 100, eNOS being primarily expressed in
SECs and in ECs of the portal vein, hepatic artery, central vein, and lymphatic
vessels 100, whereas iNOS can potentially be expressed in almost all hepatic cells 100,115.
Introduction
26
Figure 3: Schematic diagram of the NO-cGMP pathway
NOSs are generated as inactive monomers. For activation monomers must dimerize
and bind different cofactors. Tetra-hydrobiopterin (BH4), haem, flavin adenine
dinucleotide (FAD) and flavin mononucleotide (FMN) are cofactors of all three
isoforms 104,98. On binding calmodulin, a calcium-binding protein, the active enzyme
catalyzes the oxidation of L-arginine to NO and L-citrulline. For eNOS and nNOS
calmodulin binding, and hence also enzyme activity, is highly calcium-dependent,
whereas in iNOS calmodulin is bound constitutively 98,101. Moreover, post-
translational modifications and protein-protein-interactions can also regulate NOS
activity 97,98. Finally, NO generation requires molecular oxygen and nicotinamide
adenine dinucleotide phosphate (NADPH) as co-substrates for the oxidation of L-
arginine.
cGMP generation requires direct interaction between NO and the enzyme soluble
guanylyl cyclase (sGC). sGC is a heterodimeric hemoprotein composed of an a- and
b-subunit, which are both required for enzyme activity 116. Two isoforms of each
Introduction
27
subunit exist: a1/a2 for the a-subunit, as well as b1/b2 for the b-subunit, but only
a1/b1 and a2/b1 are active heterodimers. The a1/b1 heterodimer is regarded as the
major sGC isoform, since it is expressed in most mammalian tissues, including liver
tissue 117–119. Essential for sCG activation is an interaction between the heme-binding
domain, located on the b-subunit, and a heme moiety. The heme moiety is a large
heterocyclic ring with a transition metal, building the metal center of sGC.
Once NO is generated and diffused into vascular smooth muscle cells, or in the liver
into HSCs, it induces a conformational change of the sGC heterodimer by binding
avidly to its transition metal (ferrous heme iron), thereby activating sCG, which, in
turn, catalyzes the conversion of guanosine triphosphate (GTP) to cyclic guanosine
monophosphate (cGMP), an intracellular second messenger 106. Increased cGMP
concentrations exert downstream signaling effects by directly modulating various
effector proteins, i.e. cGMP-dependent protein kinases (PKGs), cGMP-hydrolzying
phosphodiesterases (PDEs), as well as cGMP-gated ion channels 120,121.
Two PKG families (PKGI and PKGII) and PDE classes (class I and class II) exist,
whereby PDE class I includes all known mammalian PDEs, which comprise 11
families (PDE1-11) 122,120,123. Only the roles of the cGMP-dependent PKGI family and
the cGMP-selective PDE5 family will be focused on, since their members are key
players in the NO-cGMP pathway 120,124.
An increase of cGMP concentrations induces a decrease in intracellular free calcium
concentrations through multiple mechanisms. First, cGMP is capable of inhibiting
calcium release from intracellular stores; second, it triggers removal and
sequestration of intracellular calcium through calcium pumps; and third, it induces
direct, as well as indirect inhibition of the influx of extracellular calcium through
voltage-gated ion channels 125. As a result, the indirect inhibition of calcium influx is
mediated by PKGI. Upon cGMP binding to allosteric sites in the regulatory domain,
PKGI undergoes a conformational change. This conformational change leads to the
release of the N-terminus inhibition of the kinase domain, and hence to an increase
in phosphotransferase activity of the dimeric enzyme 126,120,125. Thus, PKGI
stimulation results in phosphorylation of several proteins, which results in two
primary effects: first, a decrease in intracellular calcium levels and second, calcium
desensitization of the actin-myosin contractile elements 127.
As a consequence of these cGMP-induced downstream signaling effects, vascular
dilation is initiated, eventually leading to a reduction in vascular tone.
Introduction
28
Moreover, cGMP binds to the high-affinity GAF-A domain of PDE5, thereby
increasing the hydrolytic activity of the dimeric holoenzyme. The hydrolytic activity
can be further promoted through stabilization of the cGMP binding by
phosphorylation at a separate N-terminal site by PKGI 128–130. Once PDE5 is
activated, it initiates the hydrolysis of cGMP into inactive guanosine-5′-
monophosphate (GMP). Hence, rising intracellular cGMP concentrations are
associated with activation of PDE5 as a negative feedback mechanism mediating
cGMP degradation 131. All these feedback mechanisms happen within seconds and
are pivotal in lowering cGMP concentrations to basal levels in the short term after NO
stimulation 120. Prolonged NO exposure and increased cGMP concentrations,
however, seem to induce more persistent modifications at several steps in the NO-
cGMP pathway, including down-regulation of PKGI and up-regulation of PDE5 120.
Knowing about the essential role of NO in terms of vasoregulation, the role of NO
and NOS in the pathophysiology of PH has been investigated extensively.
2.5.7.3 NO and NOS in the Pathophysiology of PH According to currently available data, NO is described as an important molecular
factor involved in the pathophysiology of PH secondary to liver cirrhosis. The
paradoxically controlled intra- and extrahepatic vascular tone is characterized by NO
deficiency in the intrahepatic vasculature and, on the other hand, NO excess in the
extrahepatic vasculature.
Considering the intrahepatic vasculature, a down-regulation of eNOS activity in SECs
is described to be primarily responsible for intrahepatic NO deficiency, whereas data
considering eNOS expression are inhomogeneous described 63,76,132–139. Underlying
causes of the down-regulated eNOS activity can involve different factors, such as
oxidative stress or decreased BH4 synthesis and activity 63,139,140. Furthermore, an
up-regulation of iNOS expression is associated with liver cirrhosis, which can be
stimulated by endotoxins, cytokines and bacterial infections 63,141–144. iNOS
expression can take place in all hepatic cell types, but little can be revealed about
iNOS activity in the different cell types. However, it is known, that iNOS activity is
dependent on several factors such as availability of its substrate arginine and BH4 115. Interestingly, eNOS-derived NO is described to maintain liver homeostasis and
Introduction
29
counteract pathological conditions within the liver 100, while iNOS-derived NO is in
general pathological, and does not seem to cause vasodilation 100. The paradox of
decreased eNOS-derived NO and increased iNOS-derived NO, finally resulting in
increased intrahepatic resistance, might point out that the source of NO and the
surrounding microenvironment strongly influence the NO-induced effects 100. Its
inducible nature as well as the observation that iNOS can act as regulator of other
effectors, e.g. eNOS, increases the potential impact of iNOS massively and needs
further investigation 63,111.
Regarding the extrahepatic vasculature, an up-regulation of primarily eNOS activity,
but also eNOS expression in ECs of the vascular endothelium is assumed to cause
NO excess 64,63,145. The up-regulation of eNOS can be triggered by several stimuli,
e.g. shear stress, VEGF, or increased BH4 synthesis and activity 64,145–147. For iNOS,
however, data are inconsitstent. Some studies have implicated that enhanced iNOS
expression and activity is involved in the vasodilation of the extraheptic vasulature 141,148, whereas others found evidence against the involvement of iNOS 149. It seems
like eNOS- rather than iNOS-derived NO contributes to NO excess in the
extrahepatic vasculature, but equivalent to the intrahepatic vasculature the role of
iNOS-derived NO needs to be clarified further 74,150–152 .
Moreover, NO contributes to angiogenesis and as a result to collateral formation,
which again promotes the progression of PH 63.
Bearing that paradox in mind, the ideal concept for PH therapy should specifically
target intrahepatic vasculature and on site enhance NO availability and / or NO-
cGMP downstream signaling to counteract intrahepatic NO deficiency 60.
2.5.7.4 Strategies to Increase NO Availability and NO-cGMP Signaling To increase NO availability and / or NO-cGMP downstream signaling, different
pharmaceutical strategies can be used (Table 6). However, it must always be kept in
mind that drugs acting within the liver may have extrahepatic effects as well 94.
Hence, testing those, intra- and extrahepatic effects should be subject of scrutiny 91.
In this experimental study, we investigated the effect of PDE5 inhibitors, which act as
indirect stimulators of NO downstream signaling; thus, they will be focused on in the
following.
Introduction
30
Table 6: Reference standards and potentially novel strategies to increase NO
downstream signaling 96
Mode of action Drug group and name
1. NO delivery / generation a) NO-delivering compounds b) NO-releasing compounds c) enhancing endogenous NO generation by promoting eNOS activity d) other drugs that enhance endogenous NO generation by promoting eNOS activity and expression 2. Prevention of NO scavenging by other radicals or transition metals 3. Stimulating NO downstream signaling
Significant differences (p<0.05) between Sil 0.1mg/kg and CON or Sil 1.0mg/kg and CON are marked by an *. No significant differences between Sil 0.1mg/kg
and Sil 1.0mg/kg were observed.
Results
55
3.3.3 Effect of MAP on PVP 1 Regarding the course of MAP and PVP of the individual rats during the measurement
interval (Figure 10), a change in MAP led to a slightly delayed change in PVP in the
same direction (decrease / increase). Hence, hemodynamic data were also used to
evaluate the effect of MAP on PVP over the first 30 minutes.
For the statistical analysis of the correlation between MAP and PVP, the 108 rats
(Table 15) were classified into nine subgroups as before (see 3.3.2).
All parameters of interest were normalized (PVPrel, MAPrel) to compensate
differences in absolute values (Table 16) between healthy and diseased rats. Since
the correlation of the two parameters was visible, particularly in the first few minutes
in which administration of 600 µl liquid volume into the right atrium caused parameter
variations, time point “0 min” was taken as the baseline and set to 100%.
Figure 10: Course of MAP (black) and PVP (blue) of an exemplary rat after sodium
chloride (NaCl) administration
In CON a significant effect of MAPrel on PVPrel (p=0.001) was found in all subgroups
(Table 18). For every 1% change in MAPrel in the NaCl subgroup, PVPrel varies by
0.35%. 40% of the variation in PVPrel within one rat can be explained by MAP. For
every 1% change in MAPrel in the subgroups Sil 0.1 mg/kg and Sil 1.0 mg/kg, PVPrel
varies by 0.37% and 0.53%, respectively. 46% and 43% of the variation in PVPrel
within one rat can be explained by MAP.
1 Data of this analysis were previously published in the medical dissertation by Adhara Lazaro: “Correlation between mean arterial pressure (MAP) and portal venous pressure (PVP) in rats” (2018).
Results
56
In FIB a significant effect of MAPrel on PVPrel (p=0.001) was found in all subgroups
(Table 18). For every 1% change in MAPrel in the NaCl subgroup, PVPrel varies by
0.52%. 42% of the variation in PVPrel within one rat can be explained by MAP. For
every 1% change in MAPrel in the subgroups Sil 0.1 mg/kg and Sil 1.0 mg/kg, PVPrel
varies by 0.48% and 0.61%, respectively. 46% and 43% of the variation in PVPrel
within one rat can be explained by MAP.
In CIR as well a significant effect of MAPrel on PVPrel (p=0.001) was found in all
subgroups (Table 18). For every 1% change in MAPrel in the NaCl subgroup, PVPrel
varies by 0.32%. 61% of the variation in PVPrel within one rat can be explained by
MAP. For every 1% change in MAPrel in the subgroups Sil 0.1 mg/kg and Sil 1.0
mg/kg, PVPrel varies by 0.32% and 0.39%, respectively. 40% and 23% of the
variation in PVPrel within one rat can be explained by MAP.
Results
57
Table 18: Regression analysis between MAPrel and PVPrel in the subgroups. Parameters were normalized with time point “0 min” being
set to 100%.
The change in PVPrel for every 1% change in MAPrel is described by the regression coefficient (s). Significant effects of MAP on PVP (p<0.05) are marked
by an *. The explained variation (%) within one rat is described by r-squared (r²).
rate were determined. Moreover, MR data were used to test whether the degree of
liver fibrosis can be assessed using a self-established MR score.
Discussion
75
The idea for this study was born, since at present, the reference standard for
hemodynamic evaluation in clinical practice is the assessment of portal blood
pressure by measuring the HVPG 191,192. However, this technique is an invasive
procedure and involves some degree of inconvenience and risk for the patients.
Moreover chronic liver diseases, as well as the induced hemodynamic alterations
(mainly PH) are heterogeneous and dynamic conditions 38,81. Consequently, a
noninvasive and repeatable assessment of hemodynamics is warranted.
To date, there are few noninvasive modalities that quantify blood flow in clinical
practice. Most of them are based on ultrasound (US) and MR. Both of these
modalities have been investigated whether measurements of portal parameters (e.g.
flow velocity or volume rate) may correlate with the degree of PH with variable results 193–195. Even if technical modalities have substantially advanced over the last
decades, low reproducibility and lack of standardized examination protocols are still
mentioned as limitations of noninvasive techniques 196,197. Thus, preclinical studies
are warranted in order to evaluate and further optimize recently developed clinical
imaging techniques, and to aid in biomedical and pharmaceutical research.
Multi-dimensional phase-contrast MR (PC-MR) imaging is a preferred technique to
determine blood flow in preclinical and clinical studies. Some preclinical
hemodynamic studies in small laboratory animals using multi-dimensional PC-MR
imaging have already been performed, mostly investigating cardiovascular
hemodynamics 198–201. Only a few focus on portal hemodynamic changes induced by
liver cirrhosis 202–204. To address this lack of portal hemodynamic data, two-
dimensional PC-MR (2D PC-MR) imaging, a technique being well-established,
validated and in use for clinical practice and research 205, was chosen to determine
different hemodynamic parameters in this study. Thereby, hemodynamic changes in
the portal vein and the abdominal aorta, as well as morphological changes of the liver
tissue were focused on.
For the MR scoring of the rat livers, an MR approach and a self-established MR
score were used. Therefore, histological criteria 82 were adapted for MR imaging to
determine morphological alterations induced by liver inflammation, fibrosis, or
cirrhosis including liver tissue density, nodules and liver surface 206. From the
diseased rats 11% (4/36) were scored false negative, but all of the healthy rats were
identified as such. The accurate assessment of the degree of liver fibrosis in
Discussion
76
diseased rats, as well as the detection of CCCs was not possible by MR rat liver
image evaluation. Histological evaluation is required instead. Thus, even if histology
remains to be the reference standard, the evaluation of MR rat liver images can be
helpful to noninvasively discriminate between a healthy and a diseased liver,
consequently reducing the unnecessary use of laboratory animals.
Looking at the detected flow velocity patterns for the portal vein and the abdominal
aorta during a cardiac cycle, these reflect the physiological conditions as one would
expect in a venous or arterial vessel: the flow velocity pattern in the portal vein
appears constant, whereas the observed flow velocity pattern in the abdominal aorta
reveals a pulsatile structure. These flow velocity patterns are consistent with those
previously demonstrated by Wang et al 207 for a phantom and preclinical in vivo study
in rats using PC-MR. However, compared to Wang et al, substantially higher field
strength was used in this study (9.4 T vs. 1.5 T); thus, it can be assumed that at least
the same or, even more likely, a better signal-to-noise ratio for the detection of PC-
MR data, and therefore more robust hemodynamic parameters were achieved.
Considering the hemodynamic alterations induced by the chronic treatment of the
hepatotoxic agent TAA, the most distinct alteration in diseased rats in comparison
with healthy rats was the marked reduction of portal flow velocity and volume rate.
Results indicate that in the model of TAA-induced liver disease, the development of
fibrosis is sufficient to cause a significant decrease in portal flow velocity and volume
rate. In contrast, the development of cirrhosis caused no further significant decrease
in portal flow velocity and volume rate.
However, from these results it cannot reliably be concluded that the total liver
perfusion in the diseased rats is diminished. A reduction of portal perfusion can at
least partially be compensated (25%-60%) by an increase of arterial hepatic
perfusion – the so-called hepatic buffer response 7. Parameters of the hepatic artery
were not obtained in this study. Since aortic flow volume rate in diseased rats
remained constant, the reduction of portal flow is not a consequence of a reduced
aortic flow, but most likely an indicator of increased intrahepatic resistance.
The findings of a reduced portal flow velocity and volume rate in diseased rats are
contrary to the results given by Wang et al 207. In their preclinical study, differences in
portal flow volume rate (portal flow velocity data were not shown) between CON and
Discussion
77
FIB, as well as between CON and CIR were nonsignificant. They also investigated
Wistar rats, but used the model of CCl4-induced liver disease.
According to clinical studies, portal perfusion in humans is altered depending on the
degree of liver fibrosis and with advanced PH the portal blood flow may even become
reversed 107,208,209. Furthermore, the portal cross-sectional area may increase with
rising PVP, but the determination of the portal cross-sectional area does not seem to
be a reliable diagnostic indicator for PH 210–212.
In the present study, the portal cross-sectional area of the rats remained constant
even in severely diseased rats. It could be likely that the increase in PVP in the
tested animal model is not pronounced enough to cause a dilation of the portal vein.
In further experiments, it should be clarified to what extent PVP in diseased rats is
indeed enhanced. Moreover, an investigation should be made in which morphological
or biochemical modifications are responsible for the unexpected finding of the strong
portal hemodynamic effect even in rats with fibrotic livers.
In terms of limitations in the study design, the anesthesia, which is unavoidable,
could have added more variability and uncertainty to the measurements 213. The
physical states of the rats during the MR measurement were not perfectly equal for
all of them even if every possible precaution was taken to keep the conditions and
their physical state stable. The procedure itself - the preparation of the rats and the
MR measurement - lasted about 1 h per rat and was performed at different times of
the day.
The main technical causes of errors in the 2D PC-MR technique were possibly the
2D plane application and positioning in the complex liver vasculature, as well as the
Venc setting 205,214.
In conclusion, in rats with fibrotic or cirrhotic livers, markedly reduced portal flow
velocity and volume rate were found compared with rats with healthy livers.
Moreover, the evaluation of the MR rat liver images of the livers enables
differentiation between healthy and diseased livers.
Discussion
78
4.3 Invasive Hemodynamic Measurements 4.3.1 Portal Flow Volume Rate The aim of this part of the study was to invasively evaluate hepatic and systemic
hemodynamic changes induced by liver fibrosis / cirrhosis in rats with healthy, fibrotic
or cirrhotic livers. Alterations in the parameters portal flow volume rate and mean
arterial pressure (MAP) were determined. Portal flow volume rate was measured with
a flow probe, whereas MAP was measured by a pressure transducer. Moreover,
results for the portal flow volume rate between noninvasive and invasive
measurements were compared.
Considering the hemodynamic alterations induced by the chronic treatment of the
hepatotoxic agent TAA, the most distinct alteration in diseased rats in comparison
with healthy rats was the marked reduction of portal flow volume rate and MAP.
Equivalent to the results of the noninvasive measurements, the current findings also
indicate that in the model of TAA-induced liver disease, the development of fibrosis is
sufficient to cause a significant decrease in portal flow volume rate.
However, in contrast to the noninvasive measurements, in which no significant
differences between diseased rats in FIB and CIR were observed, a significant
reduced portal flow volume rate in CIR was detected by invasive measurements
when compared to FIB. Moreover, absolute values for the portal flow volume rate
measured in the groups were around 20 to 47% higher in the noninvasive
measurements. A comparison of methods was not performed due to several reasons:
First, at least partially different rats were used for the noninvasive and invasive
measurements, as well as a different kind of anesthesia. Second, even if rats were
investigated twice, noninvasive and invasive measurements were not performed on
the same day, but with a 2- or 3-day recovery period in between. Third, it is beneficial
to have an arterial reference parameter since changes in the arterial circulatory
system could lead to changes in the venous circulatory system but, whereas for the
noninvasive measurement the abdominal aorta flow volume rate was referred to,
MAP was used as an arterial reference parameter for the invasive measurement
since the anatomical closeness to the vena cava does not allow the positioning of a
flow probe at the abdominal aorta. Hence, the arterial reference parameters differed
between noninvasive and invasive measurements and were not comparable.
Discussion
79
In terms of limitations in the study design, the anesthesia as well as the operative
conditions, which are unavoidable, could have added more variability and uncertainty
to the measurements 213. The physical states of the rats during the MR measurement
were not perfectly equal for all of them even if every possible precaution was taken to
keep the conditions and their physical state stable. The procedure itself – the
preparation of the rats and the portal flow volume rate measurement - lasted about 1
to 1.5 h per rat and was performed at the same time of the day.
The main technical causes of errors in the measurement technique were the size and
positioning of the ultrasonic transit time flow probe, and the loss of the applied
ultrasound gel due to body fluids.
In conclusion, in rats with fibrotic or cirrhotic livers, markedly reduced portal flow
volume rate and MAP were found compared with rats with a healthy liver.
Comparing the results for the portal flow volume rate between noninvasive and
invasive hemodynamic measurements, the same pattern of a liver disease-induced
decrease was found, but absolute values were not equivalent.
4.3.2 Effect of Sildenafil on Hemodynamics The aim of this part of the study was to evaluate hepatic and systemic hemodynamic
changes induced by the administration of the PDE5 inhibitor sildenafil in rats with
healthy, fibrotic or cirrhotic livers. Therefore, additional invasive hemodynamic
measurements were performed to determine the acute effects of administration of
either sodium chloride (NaCl) or sildenafil (Sil 0.1 mg/kg or Sil 1.0 mg/kg) on the
parameters portal venous pressure (PVP), mean arterial pressure (MAP),
microvascular flow (MF), and heart rate (HR) over 50 minutes. PVP, MAP and HR
were measured using pressure transducers, whereas MF was determined with a
microvascular flow probe.
Considering the hemodynamic alterations induced by acute sildenafil administration,
a dose-dependent effect was observed. The most distinct alteration was observed
after high-dosage administration (1mg/kg) in rats with cirrhotic livers, which led to a
trend towards a decreased PVP and was associated with a significant reduction of
HR and a nonsignificant lowering of MAP.
Discussion
80
Looking at PVP, a decrease among all subgroups was determined. In rats with
healthy and fibrotic livers, the decrease was nonsignificant regardless of intervention.
However, in rats with cirrhotic livers sildenafil administration led to a trend towards a
decreased PVP, PVP decreasing more prominently after high-dosage administration.
But while healthy rats were considered to have a physiological PVP, it is hard to
make any statement about how pronounced the elevation in PVP in diseased rats
really was, particularly in association with their markedly lower baseline MAP values
in comparision with healthy rats. On the other hand, the latter could indicate that
those rats have been in a hyperdynamic circulatory state, which is typically a
consequence of PH 74.
For MAP and MF, a nonsignificant decrease was also observed among all subgroups
regardless of intervention. Changes in MAP were measured as administration of
PDE5 inhibitors has been reported to be associated with a decrease in MAP, which
could influence PVP. The decrease in MF, however, occured unexpectedly since
administration of PDE5 inhibitors should lead to sinusoidal vasodilation and hence
increased MF. It might be speculated that the effect of sildenafil on MF has been
covered by the decrease in the residual hemodynamic parameters. Moreover, a
nonsignificant decrease in HR was found among almost all subgroups regardless of
intervention, exclusively in rats with a cirrhotic liver, which received the high-dosage
In a preclinical study, investigating the effects of PDE5 inhibitors in healthy rats, it
was shown that after acute administration of either sildenafil or vardenafil (1-100
µg/kg, intravenous) PVP remained unchanged or showed a trend towards a decrease 215. A dosage of 10 µg/kg, which was most effective, led to a significant increase in
MF, but at the same time to a significant reduction in MAP. HR remained unaltered
regardless of the dosage applied.
In contrast in the model of BDL-induced liver disease, acute administration of
sildenafil (0.01-10 mg/kg, intravenous or intramesenteric) led to a dose-dependent
increase in PVP and a significant decrease in MAP 216. Diseased rats also tended to
have lower baseline MAP values compared to sham-operated rats, which is
consistent with our observations. The same model was used in another study
considering the effect of a chronic one-week administration of sildenafil (0.25 mg/kg,
2 x daily, oral) 217. Whereas in sham-operated rats no effect was found, in diseased
Discussion
81
rats a nonsignificant decrease in PVP and portal perfusion pressure, and a significant
increase in MF were determined. These findings coincide with the results of a further
study, which also used the model of BDL-induced liver disease, showing that after
chronic administration of the PDE5 inhibitor udenafil (1, 5 or 25 mg/kg; 1 x daily,
oral) for 3 weeks PVP decreased by approximately 30% 218.
In a clinical trial on patients with liver cirrhosis and a significantly elevated HVPG,
acute administration of sildenafil (50 mg, oral) caused no changes in HVPG and HR,
whereas MAP decreased 219. These findings were confirmed by a subsequent study
with a similar study design in patients with compensated cirrhosis (Child A) 220. In a
further study investigating patients with compensated and decompensated liver
cirrhosis (Child A-C), acute administration of sildenafil (50 mg, oral) showed no effect
on PVP, MAP and HR, but induced a significant reduction in intrahepatic resistance 221. Moreover, the effect of an acute and chronic one-week administration of udenafil
(12.5 -100 mg, 1 x daily, oral) was tested in patients with decompensated liver
cirrhosis (Child A-B), a dosage of 75 mg or 100 mg being found to be most effective 222. After one hour HVPG was lowered by 25% (75 mg) or 17% (100 mg)
respectively, whereas after one week HVPG was reduced by 14% (75 mg) or 17%
(100 mg) respectively. By combining the results of these two dosages a significant
reduction in HVPG of 19% in the acute setting and of 16% in the chronic setting was
found, while HR remained unchanged. However, reduced HVPG was associated with
a significant lowering of MAP of 4% in the acute setting and of 6% in the chronic
setting which, according to the authors, was well tolerated by the patients.
In a further pilot study on patients with compensated liver cirrhosis (Child A), acute
administration of vardenafil (10 mg, oral) caused a decrease in HVPG and
intrahepatic resistance in four out of five patients, whereas HR remained constant 223.
Moreover, a recent case-report about a female patient with compensated liver
cirrhosis (Child A) revealed promising results for the chronic use of PDE5 inhibitors 224. In the acute setting, administration of vardenafil (5 mg, 1 x daily, oral) led to a
reduction of HVPG by 13%. For the maintenance medication over the following eight
years with tadalafil (5 mg, 1 x daily, oral), similar effects on HVPG were described.
MAP also slightly decreased in the acute as well as in the chronic treatment phase,
but changes were reported to be clinically irrelevant.
Discussion
82
Taken together, data from preclinical and clinical studies provided promising results
regarding the effect of PDE5 inhibitor administration on PVP (or HVPG), even though
results were partly variable. The associated decrease in MAP seemed to be
tolerable. Only in the current study there was a significant decrease in HR
determined after acute high-dosage sildenafil administration (1mg/kg), which
contradicts all other existing data and needs to be clarified.
Believing the hypothesis that the presence of high PDE5 expression in a particular
tissue should predict the effect of a PDE5 inhibitor 225, the current finding of hepatic
PDE5 overexpression in diseased rats (see 3.4) reveals the need for further
investigations in order to better evaluate drug- and dose-dependent effects of PDE5
inhibitors in the acute and chronic setting. It might be assumed that in particular a
prolonged chronic administration of PDE5 inhibitors could be beneficial to counteract
PDE5 overexpression. However, potential therapy effect heterogeneity should always
be considered since the degree of liver fibrosis / cirrhosis and PH could also
influence the effectiveness of a therapy 58.
In terms of limitations in the study design, administration of 600 µl liquid volume into
the right atrium most likely caused transient cardiac decompensation probably due to
volume overload, which was accompanied by variations in PVP, MAP, MF and HR
during the first minutes of the hemodynamic measurements. After approximately 10
minutes a new steady state was reached. Therefore, timepoint “10 min” was taken as
the baseline value for further calculations. However, that measure did not seem to be
ideal for the NaCl subgroup in FIB since relative median of differences for the
parameters seemed to be markedly lower compared to those in CON and CIR. This
might have influenced the results of intragroup comparisons in FIB. Moreover, the
decrease in parameter values during the measurement interval, which occurred
consistently among all subgroups even in the absence of a vasoactive drug (NaCl
subgroups), indicates that the anesthesia as well as the operative conditions, which
are unavoidable, could have added more variability and uncertainty to the
measurements 213,226. This disruptive factor occurred although pentobarbital was
used for anesthesia, which is described to be one of the best forms of anesthesia for
the performance of invasive blood pressure measurements in rats 227.
The physical states of the rats during the hemodynamic measurement were not
perfectly equal for all of them regardless of every possible precaution taken to keep
the conditions and their physical state stable. The procedure itself – the operative
Discussion
83
procedure of the rats and the hemodynamic measurement - lasted about 2.5 to 3 h
per rat and was performed at the same time of the day. The use of different rat
strains as well as the fact that some of the diseased rats had CCCs could have
influenced the results of the hemodynamic measurements.
The main technical causes of errors in the hemodynamic measurement technique
were the location and a potential clogging of the catheters.
In conclusion, acute high-dosage sildenafil administration (1mg/kg) led to a trend
towards decreased PVP in rats with cirrhotic livers, which were characterized by
hepatic PDE5 overexpression, and furthermore to a significant lowering of HR and a
nonsignificant reduction of MAP. Hence, PDE5 inhibitors might be a promising
adjunct in PH therapy and should be investigated further.
4.3.3 Effect of MAP on PVP The aim of this part of the study was to determine the influence of systemic blood
pressure on portal blood pressure. Therefore, the effect of MAP on PVP over the first
30 minutes was evaluated based on the available hemodynamic data (see 3.3.2).
Considering the course of MAP and PVP of the individual rats, a change in MAP led
to a slightly delayed change in PVP in the same direction (decrease / increase). This
was best visible within the first minutes of measurements, in which hemodynamic
parameters were “manipulated” unintentionally by the bolus injection of 600 µl liquid
volume into the right atrium.
The results showed that the effect of MAP on PVP was significant in all subgroups
regardless of intervention. For every 1% change in MAPrel ,PVPrel varies by 0.32% to
0.61%, which implies a distinct relationship between the change in MAP and the
change in PVP.
These findings are of particular importance regarding the current pharmaceutical
options in PH therapy. As mentioned before (see 2.5.7), the functional component of
increased intrahepatic resistance can be influenced positively, either by a decrease
in intrahepatic vascular tone, a decrease in splanchnic vasodilation, or ideally both 67.
In general, NSBBs, the current cornerstone in pharmaceutical PH therapy, block the
binding of catecholamines, such as norepinephrine and epinephrine, to beta1 and
Discussion
84
beta2 adrenergic receptors 33,228. To lower PVP (HVPG), NSBBs act in two different
ways: whereas a beta1 blockade reduces portal inflow by decreasing the heart rate
and cardiac output, a beta2 blockade leads to unopposed alpha1 activity resulting in
splanchnic vasoconstriction 90. The latter however, seems to be the essential mode
of action 68. In comparison, beta1-selective beta blockers, which only lower cardiac
output, show a less pronounced effect on PVP (HVPG) than NSBBs 229–233.
Unfortunately, not only NSBBs, but most of the drugs used in PH therapy are
vasoactive and do not only affect hepatic, but to some extent systemic
hemodynamics as well. Statins exclusively, best known for their cholesterol lowering
effects, are able to decrease intrahepatic resistance without affecting systemic
hemodynamics simultaneously 66. However, since PH per se is already associated
with splanchnic vasodilation and the development of a hyperdynamic circulatory
state, a further reduction of systemic blood pressure induced by vasoactive drugs is a
matter of concern, especially in advanced stages of the disease 234.
Even the use of NSBBs has been a matter of ongoing controversy due to potential
hemodynamic, but also nonhemodynamic adverse effects in patients with liver
cirrhosis 68,228,235–240. Those side effects led to treatment termination in approximately
15% of patients, whereas another 15% a priori had contraindications to the use of
NSBBs 241. Moreover, 30 to 40% of patients did not show a portal hemodynamic
response to NSBB administration 242,243. In addition, the therapeutic window
hypothesis resulting from a meta-analysis by Krag et al 33 limits their use as well. The
therapeutic window hypothesis states that in patients with liver cirrhosis NSBBs
improve survival only during a certain time window in the disease. This window
opens when medium to large esophageal varices occur 244–246, and closes when a
very advanced stage of liver cirrhosis is reached 87,229,247–249. Since outside this
therapeutic window NSBBs have been described to be potentially ineffective or even
detrimental, it is obvious that novel therapeutic strategies are needed.
Nevertheless, NSBBs might be an excellent example to investigate the correlation
between systemic and hepatic hemodynamics. In a recent review by Tripathi 250, the
changes in HVPG and the associated changes in MAP after acute and chronic
administration of propaponol and carvediol in patients with PH were presented.
These data also suggest a correlation between systemic and hepatic hemodynamics
since after acute or chronic administration of propranolol, a conventional NSBB,
Discussion
85
reduction of HVPG, but also MAP was less pronounced than after acute or chronic
administration of carvediol, a NSBB with an additional alpha1 blocking capacity.
However, whereas those results are based on measurements at selected time points
only, results of the current experimental study were based on continuous invasive
hemodynamic measurements over 30 min, and therefore more robust results were
achieved. Another advantage of this study was the unintentional “manipulation” of
hemodynamics within the first minutes, which best illustrated the correlation between
MAP and PVP. Moreover, even in the absence of a vasoactive drug (NaCl
subgroups) the course of hemodynamic parameters revealed a decrease not only in
MAP but also in PVP, suggesting a correlation between these two parameters, but on
the other hand, indicating that the sildenafil-induced decrease in PVP was also at
least partly a consequence of the lowering of MAP. However, although no intragroup
comparisons were performed, results showed that for a 1% change in MAP the
change in PVP was highest in the subgroups, in which the high dosage of sildenafil
(Sil 1 mg/kg) was applied, which might imply a partly liver-specific effect. Hence,
future studies evaluating the effect of vasoactive drugs on PVP (HVPG) should
distinguish between the portion induced by the decrease in MAP, and the portion that
indeed reflects a liver-specific mode of action. Ideally, the effect on PVP should be
markedly higher than the effect on MAP.
In terms of limitations, reference can be made to those listed above (see 4.3.2).
In conclusion, there is a distinct correlation between MAP and PVP, i.e. between
systemic blood pressure and portal blood pressure. This should be considered in all
other studies evaluating the effect of vasoactive drugs on PVP (HVPG).
4.4 Biochemical Investigations The aim of this part of the study was to evaluate changes biochemically in the key
parameters of the nitric oxide-cyclic guanosine monophosphate (NO-cGMP) pathway
induced by liver fibrosis / cirrhosis in rats with healthy, fibrotic or cirrhotic livers. The
NO-cGMP pathway is a key regulator of vascular tone and thus plays an important
role in sinusoidal vasoreactivity, which is impaired in PH. Alterations in hepatic gene
expression of the enzymes endothelial and inducible NO synthase (eNOS, iNOS),
soluble guanylyl cyclase subunit a1 and b1 (sGCa1, sGCb1) and phospho-
Discussion
86
diesterase 5 (PDE5) were analyzed by qRT-PCR, whereas changes in serum cGMP
concentrations from carotid arterial blood samples were determined using ELISA.
Additionally, blood samples were used to determine serum parameters (clinical
chemistry) for the sake of completeness, but the results will not be discussed further.
In this context, it was also evaluated whether the hemodynamic measurement and in
particular the associated operative procedure affected gene expression or serum
cGMP concentrations. Moreover, the effect of sildenafil administration (1.0 mg/kg) on
serum cGMP concentrations was determined. The main finding of gene expression
analyses was finally confirmed by immunohistochemical staining.
Considering the alterations in the NO-cGMP pathway induced by the chronic
treatment of the hepatotoxic agent TAA, distinct alterations in diseased rats in
comparison with healthy rats were observed. In terms of hepatic gene expression, an
iNOS up-regulation and a marked PDE5 overexpression were detected. The less
pronounced increase in the expression of the residual genes, i.e. eNOS, sGCa1 and
sGCb1, might represent a compensatory mechanism to balance PDE5 over-
expression. The latter was confirmed by immunohistochemical investigation of PDE5
protein expression, which furthermore revealed a loss of hepatic zoning. Serum
cGMP concentrations were slightly decreased in diseased rats, but high-dosage
sildenafil administration (1mg/kg) nearly led to renormalization. Finally, a significant
decrease in eNOS gene expression was detected due to the hemodynamic
measurement and the associated operative procedure.
Regarding hepatic eNOS gene expression, a significant elevation in rats with cirrhotic
livers was found in the current study, whereas eNOS elevation in rats with fibrotic
livers was significant in nonadjusted pairwise comparisons only. As described above
(see 2.5.7.3), other preclinical and clinical studies investigating eNOS gene and / or
protein expression showed inhomogeneous results, including enhanced, unchanged,
or diminished eNOS expression, whereas a down-regulation of eNOS activity, in
particular in SECs, has been consistently described in the context liver cirrhosis 63,76,132–139. Moreover, results revealed that iNOS gene expression was absent in
healthy rats, but up-regulated in diseased rats, which is in good agreement with
findings from other preclinical and clinical studies 251–253. Since iNOS can be
expressed potentially by all hepatic cell types, its activity might vary in dependency
with its localization. The cell-specific role of iNOS-derived NO and its potential impact
Discussion
87
on vascular regulation needs to be clarified further (see 2.5.7.3). So far, up-regulated
iNOS protein expression has been described as not causing vasodilation 100, but
instead has been associated with intrahepatic microvascular dysfunction in some
other animal studies investigating endotoxemia and steatosis 112,254. These
observations make sense in that eNOS and iNOS compete for their common cofactor
BH4, meaning iNOS up-regulation could lead to reduced eNOS activity and hence
HSC activation 255,256.
Looking at sGC, the results showed a slightly increased gene expression in the
current model of TAA-induced liver cirrhosis, which coincides with former preclinical
studies reporting elevated sGC protein expression in the model of CCl4-induced 257
and BDL-induced liver cirrhosis 217. Furthermore, in both models a markedly elevated
expression of PDE5 protein was shown by western blot analyses 257,217, which was
validated by the current immunohistochemical PDE5 staining in the TAA model. In
contrast, sGC activity was found to be significantly decreased in the model of BDL-
induced liver cirrhosis 258, whereas data for PDE5 activity are lacking. Increased sGC
activity, however, occurred regardless of the amount of NO available in the liver,
indicating that with increasing intrahepatic NO deficiency the effect on vascular
regulation might be magnified by reduced cGMP generation 258.
In terms of protein distribution in healthy livers, eNOS has been described to show a
nonzonal distribution in human livers, eNOS being expressed predominantly by
hepatocytes, but also by ECs of hepatic arteries, terminal venules, sinusoids and
biliary epithelium 137, whereas iNOS is normally absent in healthy livers. In rat livers,
sGC showed a zonal distribution, whereby in the parenchyma almost all HSCs
expressed sGC, while the number of sGC-expressing HSCs decreased towards the
central vein 106. No sGC has been expressed in the innermost region around the
central vein, in which in the current study an accumulation of PDE5, expressed by
perivenular hepatocytes, was found.
The opposing enzyme zoning of sGC and PDE5 within a healthy hepatic lobule could
represent a regulatory mechanism to control sinusoidal cGMP concentrations.
Whereas cGMP generation in the parenchyma is maintained by high sGC expression
to ensure appropriate sinusoidal vasoreactivity, it is attenuated towards the central
vein due to reduced sGC expression. The latter, together with the increased PDE5
Discussion
88
expression around the central vein, might serve to inactivate excess cGMP in the
blood before it finally passes from the intrahepatic into the extrahepatic vasculature.
In the context of liver cirrhosis, loss of hepatic zoning and formation of bands
composed of fibrous connective tissue (septa) was observed, leading to altered
enzyme distribution. In diseased human livers, not only eNOS-, but also iNOS protein
showed nonzonal distribution within the hepatic lobule 137,252. Furthermore, Mc
Naughton et al 137 found a translocation of eNOS from hepatocytes to hepatocyte
nuclei. This phenomenon of translocation to cell nuclei has also been described
under other pathological conditions for enzymes and molecules, such as NOS (all
isoforms), sGC and cGMP 259. Data for sGC distribution are lacking, but for PDE5
results of the current study revealed that its expression in diseased rats is not only
markedly up-regulated, but postponed towards the parenchyma and fibrous
connective tissue (septa) (CIR 1 and CIR 2). In the parenchyma, PDE5 was
expressed by different perisinusoidal cells, predominantly HSCs (and / or
myofibroblasts), but also likely by macrophages and SECs. Which cell types in the
fibrous connective tissue express PDE5 has yet to be clarified.
In general, it seems like hepatic zoning is an underrated topic regarding the
pathophysiology of PH, although the concept of hepatic zoning itself and functional
differences between zones is nothing new as exemplified by hepatic enzymes
involved in different metabolic pathways 260,137. For hepatocytes it has also been
described that their functions and gene profiles depend on their location within the
hepatic lobule 261. Consequently, it is obvious that correct hepatic zoning is required
to ensure physiological liver functions, including adequate NO generation and / or
subsequent downstream signaling. Liver cirrhosis, however, is associated with loss of
enzymatic and metabolic zoning due to structural modifications of the liver
architecture. Resulting alterations in expression, activity and distribution of key
enzymes involved in the NO-cGMP pathway and their specific role in the
pathophysiology of PH have never been investigated systemically and should be
clarified further.
Referring to intrahepatic cGMP concentrations, it might be assumed that these
should be decreased in liver cirrhosis due to hepatic PDE5 overexpression, which
leads to an enhanced hydrolysis of cGMP into inactive GMP. In the current study, a
trend towards decreased cGMP concentrations in carotid arterial serum was indeed
Discussion
89
found, but most clinical studies detected increased cGMP concentrations in
association with liver cirrhosis in arterial as well as portal venous blood 262–264. These
opposing results need further clarification, but it might be that cGMP concentrations
depend on the location blood samples are taken from. In the current study, blood
samples were taken from the carotid artery since the blood volume from the portal
vein would not have been sufficient for all serum analyses. One could also speculate
that the rats developed pulmonary hypertension secondary to liver cirrhosis. In turn,
pulmonary hypertension leads to increased PDE5 expression in lung vascular
smooth muscle cells 265,225, which could have contributed to the reduction of cGMP
concentrations in the blood before it eventually reached the carotid artery. After acute
high dosage sildenafil administration (1mg/kg, intravenous) a significant increase in
serum cGMP concentrations in rats with healthy and cirrhotic livers was revealed by
current results. This is in good agreement with findings from a clinical study by Lee et
al 221, showing a significant increase in intraheptic NO and cGMP concentrations
after acute administration of sildenafil (50 mg/d, oral) in patients with liver cirrhosis.
These data were furthermore supplemented by a preclinical study by Lee et al 217 in
the model of BDL-induced liver cirrhosis, in which a one-week administration of
sildenafil (0.25 mg/kg, 2 x daily, oral) led to increased sGC and simultaneously
decreased PDE5 expression, accompanied by a reduction of PVP and portal
perfusion pressure, and a significant increase in MF.
Moreover, further experiments were performed in the current study to evaluate
whether the hemodynamic measurement and the associated operative procedure
affected gene expression or serum cGMP concentrations. A significant decrease in
eNOS gene expression was detected that might have influenced hepatic
hemodynamic parameters. This finding should be considered in all other invasive
hemodynamic studies.
In terms of limitations in the study design, qRT-PCR data does not allow any
conclusion about the enzymes’ activity. Moreover, cGMP concentration determined in
carotid arterial serum does not necessarily reflect cGMP concentrations in portal
venous serum or intrahepatic cellular cGMP concentrations. Carotid arterial serum
was taken to ensure an adequate amount of serum for serum parameter analyses
and ELISA.
Discussion
90
For the microscopic quantification of the PDE5, stained liver tissue samples only
stained cells in the parenchyma were counted (see 3.4.3).
In conclusion, this part of the study contributes to the understanding of the
pathophysiology of PH and particularly its functional component. Not only marked
alterations in the key parameters of the NO-cGMP pathway, a key regulator of
vascular tone, but also loss of hepatic zoning were found in association with liver
cirrhosis. These changes support the hypothesis that sinusoids remain in a
contractile state, thereby contributing to PH.
4.5 Concluding Remarks The ongoing interest in PDE research since their discovery coincides with the
development of their inhibitors 122. The fact that PDEs exist ubiquitously in every cell
in the body, but with distinct cellular and subcellular distribution of the 11 PDE
families, provided new opportunities for selective therapeutic targets 154. For diseases
with an underlying vascular impairment, notably PDE5 has been described to
represent a promising target due to its presence in vascular smooth muscle cells and
in platelets, and its specific hydrolysis of cGMP 266.
Regarding the historical development of PDE5 inhibitors, sildenafil synthesized by
Pfizer was the first potent und selective PDE5 inhibitor that was finally marketed for
the therapy of erectile dysfunction 156, but zaprinast was the first compound ever
described for selective inhibition of PDE5 267. In vitro studies investigating human
corpus cavernosum tissue showed a sildenafil effect, which was around 240-fold
more potent at inhibiting PDE5 than zaprinast 268. Later on, two further PDE5
inhibitors, vardenafil and tadalafil, were developed and also approved for treatment of
erectile dysfunction and pulmonary hypertension 117,269,270,154. But whereas PDE5
inhibitors had been successfully launched in the therapy of erectile dysfunction and
pulmonary hypertension, their use in the management of PH still needs approval.
For this reason, in the current study the potential of PDE5 inhibitors in PH therapy
was further elucidated in the animal model of TAA-induced liver fibrosis/ cirrhosis. As
a basis, liver disease-induced hemodynamic changes in this model were determined,
before additional hemodynamic measurements were conducted to evaluate the
changes induced by the administration of the PDE5 inhibitor sildenafil. Moreover,
Discussion
91
biochemical analyses of alterations in the key parameters of the NO-cGMP pathway
were performed to extend the general understanding of the pathophysiology of PH,
emphasizing on the functional component. In summary, current findings suggest that
administration of PDE5 inhibitors might at least partly correct the intrahepatic
dysregulation of the NO-cGMP pathway and the associated changes in hepatic
hemodynamics. Hence, PDE5 inhibitors could present a promising adjunct in PH
therapy.
As a future perspective, the use of PDE5 inhibitors as an antifibrotic drug should also
be taken into account since a preliminary preclinical study showed that chronic
administration of the PDE5 inhibitor udenafil over 3 weeks exhibited antifibrotic
effects, probably due to HSC deactivation 218. Equivalent to the effects induced by
chronic administration of PDE5 inhibitors, chronic administration of sGC activators
also led to a decrease in PVP and antfibrotic effects in some initial preclinical studies 91,271,272. Thus, a combined therapy of PDE5 inhibitors and sGC activators could also
represent a favorable adjunct in PH therapy. Another interesting approach for a
future use of PDE5 inhibitors is the sinusoidal pressure hypothesis, which states that
an elevation of sinusoidal pressure is the major upstream event that initiates fibrosis 21. This hypothesis contradicts the commonly accepted opinion that pressure
changes are exclusively a consequence of liver cirrhosis, but assuming it turns out to
be true, the potential of PDE5 inhibitors would be further extended.
When it comes to clinical use of PDE5 inhibitors however, differences in clinical
pharmacology should be considered. For sildenafil, vardenafil and tadalafil
differences regarding pharmacokinetics and pharmacodynamics has been well
described in a review by Mehrotra et al 273. In short, sildenafil and vardenafil are very
similar in terms of their chemical structure, while tadalafil has a markedly different
structure 274. These chemical similarities and differences are reflected in the clinical
pharmacokinetics and pharmacodynamics of these compounds, which lead to
substance-specific properties 273,274. Appreciation of the latter is needed to ensure a
rational dosage and compound selection based on the individual needs of the patient 275. Moreover, treatment with PDE5 inhibitors can implicate adverse events, such as
headache, flushing, dyspepsia, rhinitis, and visual disturbances 155,276,154,277. The
latter were reported in particular in association with sildenafil and vardenafil
administration, and are most likely caused by their nonselectivity towards PDE6, an
Discussion
92
enzyme located in the retina 278,279,154. Tadalafil, on the other hand, shows a clearly
higher selectivity towards PDE5 relative to PDE6, which might explain the lower
frequency of visual disturbances associated with tadalafil administration 273,277. In
general, however, the use of PDE5 inhibitors for the treatment of erectile dysfunction
or pulmonary hypertension has been reported to be safe, effective and well-tolerated 280,270,277,281. Should they become approved prospectively for the treatment of PH in
liver cirrhosis patients, it should be considered that drug safety in this particular
setting is a more delicate matter. Since marked changes in terms of drug disposition,
metabolism, excretion and elimination might occur as a consequence of liver
cirrhosis, it can be challenging to determine how best to prescribe drugs, including
PDE5 inhibitors, or to predict drug-drug interactions in these patients 282–284.
Materials and Methods
93
5. Materials and Methods 5.1 Materials 5.1.1 Chemicals, Reagents and Other Matters
Chemicals / Reagents Manufacturer
b-mercaptoethanol (98+%)
EDTA (ethylenediamine tetraacetic
acid disodium) (99+%)
Entellan® mounting medium
ethanol (100%)
formalin solution
(neutral buffered, 10%)
hematoxylin
Histoacyrl®
InvitrogenTM SYBR® Green
phosphate buffered saline (PBS)
stevia liquid sweetener
thioacetamide
tris (trishydroxymethylamio-
methane) (99.8+%)
Tween® 20 solution
Sigma-Aldrich, Schnelldorf, Germany
Serva Electrophoresis, Heidelberg, Germany
Merck Chemicals, Darmstadt, Germany
Honeywell, Morris Plains, New Jersey
Sigma-Aldrich, Schnelldorf, Germany
Sigma-Aldrich, Schnelldorf, Germany
B. Braun Melsungen, Melsungen, Germany
Thermo Fisher Scientific, Waltham,
Massachusetts
Oxoide, Hampshire, England
Borchers fine food, Oyten, Germany
Sigma-Aldrich, Schnelldorf, Germany
Sigma-Aldrich, Schnelldorf, Germany
PanReac AppliChem, Darmstadt, Germany
Materials and Methods
94
5.1.2 Anaesthetics and Drugs
Anaesthetics / Drugs Manufacturer
Forene®
Heparin sodium (25000 I.E./5 ml)
Jonosteril®
Pentobarbital sodium
Pancuronium bromide (2mg/ml)
Revatio® (0.8mg/ml)
sodium chloride (0,9%)
AbbVie, Wiesbaden, Germany
ratiopharm, Ulm, Germany
Fresenius Kabi, Bad Homburg, Germany
Fagron, Barsbüttel, Germany
Inresa Arzneimittel, Freiburg, Germany
Pfizer, Berlin, Germany
B. Braun Melsungen, Melsungen, Germany
5.1.3 Antibodies, Kits, Primer, and Probes
Antibodies / Kits / Primer / Probes Manufacturer
anti-PDE5a-antibody (ab64179)
cDNA synthesis kit
cGMP ELISA kit (ab133052)
Dako EnVision® +, System-HRP
(DAB) kit
dNTP mix (10mM each)
primer
RNeasy® Plus Mini kit
Taq DNA polymerase kit
(Taq DNA polymerase: 500 units)
Abcam, Cambridge, UK
Thermo Fisher Scientific, Waltham,
Massachusetts
Abcam, Cambridge, UK
Dako, Glostrup, Denmark
Thermo Fisher Scientific, Waltham,
Massachusetts
Microsynth, Balgach, Switzerland
Qiagen, Hilden, Germany
InvitrogenTM, Thermo Fisher Scientific,
Waltham, Massachusetts
Materials and Methods
95
5.1.4 Consumables
Consumables Manufacturer
Bepanthen® eye and nose cream
catheter
(Tygon® R3607, ID 1.14 mm)
catheter PE-10
(PE, ID 0.28mm)
catheter PE-50
(Portex®, ID 0.58mm)
Cellstar® serological pipette
(5ml, 10ml, 50ml)
Cellstar® centrifuge tube
(15ml, 50ml)
cotton swab
culture dish (sterile, ID 60mm)
Discofix® C three-way tap
electrode gel
Eppendorf® reaction vessel
(1.5ml)
face mask
Feather® standard scalpel (sterile)
Foliodress® head cover
gauze compress (sterile)
Graseby® respiration sensor
gigasept® FF(new) disinfection
Bayer, Leverkusen, Germany
IDEX Health & Science, Wertheim, Germany
Becton Dickinson Primary Care Diagnostics,
Sparks, Maryland
Smiths medical International, Kent, UK
Greiner Bio-One, Frickenhausen, Germany
Greiner Bio-One, Frickenhausen, Germany
neoLab Migge, Heidelberg, Germany
Carl Roth, Karlsruhe, Germany
B. Braun Melsungen, Melsungen, Germany
Gello Geltechnik, Ahaus, Germany
Eppendorf, Hamburg, Germany
3M Health Care, St. Paul, Minneapolis
pfmmedical, Osaka, Japan
Hartmann, Heidenheim, Germany
Fuhrmann, Much, Germany
Medicare Health & Living, Kilmacanogue,
Ireland
Schülke & Mayr, Norderstedt, Germany
Materials and Methods
96
Infuvalve® non-return valve
KendallTM neonatal ECG
electrodes H207PG
Kodan® tincture disinfection
(forte, colorless)
LightCycler® 480, 96 well plate
LightCycler® 480, cover sheeting
Omnifix® F Solo syringes (1ml)
Omnifix® Solo syringes
(5ml, 10ml)
Original-Perfusor® syringes (50ml)
Original-Perfusor® line (2m)
Mini-Spike® filter (green)
mirco tube (1.5ml)
MoliNea® operation pad
Microtouch® latex cloves
(powder-free)
Microtouch® Nitratex® nitrile
cloves (powder-free)
Parafilm M®
pipette tips (10µl)
pipette tips (diverse)
reaction vessel (1.5ml)
sample tube
(2ml, DNA-, DNase-, RNA-free)
silicone hose (ID 5mm)
B. Braun Melsungen, Melsungen, Germany
Covidien-Medtronic, Minneapolis, Minnesota
Schülke & Mayr, Norderstedt, Germany
Roche, Basel, Switzerland
Roche, Basel, Switzerland
B. Braun Melsungen, Melsungen, Germany
B. Braun Melsungen, Melsungen, Germany
B. Braun Melsungen, Melsungen, Germany
B. Braun Melsungen, Melsungen, Germany
B. Braun Melsungen, Melsungen, Germany
Sarstedt, Nümbrecht, Germany
Paul Hartmann, Heidenheim, Germany
Ansell, Brussels, Belgium
Ansell, Brussels, Belgium
Bemis, Neenah, Wisconsin
Biozym Scientific, Oldendorf, Germany
Mettler-Toledo Rainin, Oakland, California
Greiner Bio-One, Frickenhausen, Germany
Biozym Scientific, Oldendorf, Germany
Ketterer & Liebherr, Freiburg, Germany
Materials and Methods
97
Seraflex® (EP 1.5 / USP 4/0)
screw caps with sealing ring
(yellow, DNA-, DNase-, RNA-free)
screw caps (white)
Sterican® cannula (22G, 30G)
surgical instruments
ultrasound gel Caleo
Versatus® peripheral venous
catheter (26G)
QIAshredder
QIAxpert slide
Serag-Wiessner, Naila, Germany
Biozym Scientific, Oldendorf, Germany
Sarstedt, Nümbrecht, Germany
B. Braun Melsungen, Melsungen, Germany
Aesculap, Tuttlingen, Germany
Caesar & Loretz, Hilden, Germany
Terumo, Eschborn, Germany
Qiagen, Hilden, Germany
Qiagen, Hilden, Germany
5.1.5 Apparatus
Apparatus Manufacturer
Data acquisition system
(HSE-USB-HAEMODYN)
DPC MicroMix 5 shaker
ECG and respiration monitoring
and gating system Model 1030
Eppendorf® table centrifuge
5417C
Eppendorf® table centrifuge
5424R
Eppendorf® Thermomixer
Compact
Eppendorf® Multipipette® plus
Hugo Sachs Elektronik - Havard Apparatus,
March-Hugstetten, Germany
DPC systems, Benbrook, Texas
SA instruments, Stony Brook, New York
Eppendorf, Hamburg, Germany
Eppendorf, Hamburg, Germany
Eppendorf, Hamburg, Germany
Eppendorf, Hamburg, Germany
Materials and Methods
98
Eppendorf® pipettes (diverse)
freezer (-20°C)
freezer (-80°C)
hair clipper / trimmer QC5115
Heraeus® Megafuge® 1.0
universal centrifuge
Ice-maker
IKA® magnetic mixer
(COMBIMAG RET)
isoflurane vapor 19.3
LightCycler® 480
laser doppler blood flow monitor
DRT4 with a Titanium tipped low
profile disc probe type DP8C
liquid nitrogen container (TR11)
microwave
MouseOx®Plus pulse oximeter
system
MR scanner BioSpec 94/21 URS
(preclinical, 9.4T)
operating light KL 1500 LCD
operation table rat type 872H with
homeothermic controller type 874
(230 vac)
oven 400 HY-E
Perfusor® fm syringe pump
Eppendorf, Hamburg, Germany
Liebherr, Ochsenhausen, Germany
Heraeus, Hanau, Germany
Philips, Singapore, Singapore
Thermo Fisher Scientific, Waltham,
Massachusetts
Hoshizaki Europe, Amsterdam, Netherlands
IKA-Werke, Staufen, Germany
Drägerwerk, Lübeck, Germany
Roche, Basel, Switzerland
Moor Instruments, Devon, UK
KGW-Isotherm, Karlsruhe, Germany
Siemens, Munich, Germany
Starr Life Sciences, Oakmont, Pennsylvania
Bruker, Ettlingen, Germany
Schott, Mainz, Germany
Hugo Sachs Elektronik - Havard Apparatus,
March-Hugstetten, Germany
Bachofer, Reutlingen, Germany
B. Braun Melsungen, Melsungen, Germany
Materials and Methods
99
Pipetboy pro (pipetting aid)
Pipetman® pipettes (diverse)
pressure infusion cuff (500ml)
pressure transducer ATP300
(arterial)
pressure transducer P75 type 379
(venous)
QIAxpert spectrophotometer
quadrature volume rat coil
BioSpin MRI (Item: RF RES 400 1
H 112/072 QUAD TR AD)
rodent ventilator type 7025
Rotiolabo® Economy magnetic
bars
TAM-A plugsys transducer
amplifier module type 705/1
Transonic® animal research
flowmeter T206 series with
perivascular flow probe type 2.5S
Vortex-Genie2
water recirculator
µQuantTM spectrophotometer
Zeiss Axioplan microscope
Integra Biosciences, Biebertal, Germany
Gilson, Middelton, Wisconsin
Droh, Mainz, Germany
Hugo Sachs Elektronik - Havard Apparatus,
March-Hugstetten, Germany
Hugo Sachs Elektronik - Havard Apparatus,
March-Hugstetten, Germany
Qiagen, Hilden, Germany
Bruker, Ettlingen, Germany
Ugo Basile, Gemonio, Italy
Carl Roth, Karlsruhe, Germany
Hugo Sachs Elektronik - Havard Apparatus,
March-Hugstetten, Germany
Transonic Systems, Ithaka, New York
Scientific Industries, Bohemia, New York
supplied from Bruker, Ettlingen, Germany
Bio Tek Instruments, Bad Friedrichshall,
Germany
Carl Zeiss Microscopy, Göttingen, Germany
Materials and Methods
100
5.1.6 Software
Software Manufacturer
HSE-Basic Data Acquisition
Software (BDAS) 1.5
KC4
LightCycler® 480 Software 1.5
MatLab® 14b
ParaVision 5.1
PC-sam 32
SPSS® software 23.0 / 24.0
STATA® software 14
QIAxpert Software 2.2.0.21
Hugo Sachs Elektronik - Havard Apparatus,
March-Hugstetten, Germany
Bio Tek Instruments, Bad Friedrichshall,
Germany
Roche, Basel, Switzerland
MathWorks, Natick, Massachusetts
Bruker, Ettlingen, Germany
SA instruments, Stony Brook, New York
IBM, Armonk, New York
StataCorp LLC, Lakeway Drive, Texas
Qiagen, Hilden, Germany
5.1.7 Animals
Animals Manufacturer
Sprague Dawley rats
Wistar rats
Charles River, Sulzfeld, Germany
Charles River, Sulzfeld, Germany
Materials and Methods
101
5.2 Methods
5.2.1 Laboratory Animals The laboratory animal research protocol was approved by the local institutional
animal care and use committee (Regierungspräsidium Freiburg, ref. no.: G-13/89)
Animal care was performed in accordance to the rules and regulations of the German
animal protection law and the animal care guidelines of the European community
(2010/63/EU). A total of 275 male rats, specifically 147 Sprague Dawley and 128
Wistar rats (Charles River) were studied (Table 26). All of them were clearly
recognizable from their permanent and unique identifiers using an ear punch code.
Rats were housed in individually ventilated cages in a laboratory animal facility and
received daily human care. Their body condition was documented at least three
times a week according to a self-established score sheet (see 7.1). All rats had free
access to food and water and were exposed to a 12:12-h light–dark cycle at an
ambient temperature of 22-25 °C.
Before starting any experiments, the rats were allowed to acclimatize to the ambient
conditions for at least one week.
5.2.2 Induction of Liver Disease with TAA 133 rats were left untreated, whereas 142 rats received thioacetamide (TAA) (Sigma-
Aldrich) to induce liver disease (Table 26). The protocol of liver fibrosis / cirrhosis
induction described previously by Li et al 163 was used. TAA was administered orally
via drinking water for 12 to 24 weeks. 2.5 ml liquid sweetener (stevia liquid
sweetener) was added per 750 ml drinking water to mask the bitter taste of TAA.
Starting with an initial dosage of 0.03% TAA (30 mg TAA / 100 ml) in the first week,
TAA administration was continued with an individual TAA dosage adjusted weekly
according to each rat’s body weight change. If a rat gained or lost more than 20 g
body weight per week, the dosage was increased or decreased by 0.015%
accordingly. An increase of the TAA dosage by 0.015% was done in rats with an
overall body weight increase of more than 60 g.
Materials and Methods
102
Table 26: Number of untreated and TAA-treated rats sorted by strain
Strain Untreated TAA-treated n total
Sprague Dawley 106 41 147
Wistar 27 101 128
n total 133 142 275
5.2.3 Noninvasive Hemodynamic Measurements 5.2.3.1 MR Scanning Before MR scanning was started rats were fasted for 1.5 h to avoid prandial effects
on portal flow parameters. Anesthesia was initiated in an animal induction chamber
using a mixture of 3% isoflurane (Forene®) and 97% oxygen. It was maintained with
an animal nose mask applied with a mixture of 1.5% isoflurane (Forene®) and 98.5%
oxygen at a flow rate of 0.6 l/min. Eye cream (Bepanthen® eye and nose cream) was
applied on the eyes of the rats to prevent desiccation. ECG electrode pads
(KendallTM neonatal ECG electrodes H207PGT) with applied conductive gel
(electrode gel) were fixed on the forepaws, and a respiration sensor (Graseby®) on
the abdomen. Both were connected with a monitoring and gating system (Model
1030). ECG and respiration rate (spontaneous breathing) were continuously
monitored by the corresponding software (PC-sam 32). The body temperature was
not determined, but a warm water recirculator (supplied from Bruker) was used to
keep it stable at 37 ± 0.5 °C during the measurement. Then rats were scanned using
a 9.4 T preclinical scanner (BioSpec 94/21 URS), a dedicated quadrature volume rat
coil with an inner diameter of 68 mm (BioSpin MRI, Item: RF RES 400 1 H 112/072
QUAD TR AD) and the corresponding software (ParaVision 5.1). (Figure 13)
Materials and Methods
103
Figure 13: Preparation of the rat for the MR measurements and insertion into
MR scanner
The following scanning protocol was tested for the evaluation of MR rat liver images
to assess the degree of liver fibrosis and to determine the cross-sectional areas and
mean flow velocities of the rats’ portal vein and abdominal aorta:
To get a morphological overview for the planning of the measurements, several
localizers were used in multiple orientations (Figure 1). Then an ECG- and
respiratory-gated T1-RARE axial sequence was acquired to determine the
morphological alterations induced by liver inflammation, fibrosis, or cirrhosis. These
included liver tissue density, nodules and liver surface. Parameters for the T1-RARE
were FoV: 4.5 x 6 cm, MTX: 256 x 336, TE/TR: 8.87 ms / 1555 ms, slice thickness: 1
mm and spatial resolution: 0.0176 x 0.0179 cm/pixel.
The T1-RARE was also used for the planning of the two flow-sensitive 2D PC-MR
sequences measuring the hemodynamic parameters in the portal vein and the
abdominal aorta in a single slice perpendicular to the respective vessels (Figure 14).
Materials and Methods
104
Figure 14: T1-RARE images displaying the positioning of the PC-MR slice
orthogonally to the portal vein (a and b) and the abdominal aorta (c and d) of a rat on
the coronal (a and d) and sagittal (b and c) reference scans
The 2D PC-MR technique uses the fact that spins dephase in the presence of a field
gradient to produce a contrast between stationary tissues and flowing blood (Figure
15). Initially a reference scan is performed, in which all spin phases are in the same
position (flow-compensated image). Subsequently, the spin phase is manipulated by
a bipolar gradient pulse, such that the phase shifts of stationary spins are
compensated and the phase shifts of moving spins are proportional to the flow
velocity (flow-encoded image) 285. The faster the spins are moving the greater is their
phase shift and thus their phase angle (ϕ). The sensitivity to slow or fast flows is
determined by the velocity encoding (VENC), a user-defined parameter which
describes the amplitude, duration, and spacing of the bipolar gradient. The phase-
contrast image is generated by subtraction of these two sets of phase information
(flow-encoded data - flow-compensated data). The remaining phase difference (Δϕ)
can then be used for voxel wise calculation of flow velocities 286. Hyperintense voxels
(bright, white) represent a high flow velocity in the positive direction, whereas
hypointense voxels (dark, black) represent a high flow velocity in the opposite
(negative) direction. Stationary spins in stationary tissue with no net spin phase and a
flow velocity of zero are illustrated as mean gray areas. Scattered areas represent
irregular flow or noise.
Materials and Methods
105
Figure 15: Principle of 2D PC-MR Figure reprinted with permission of the Korean Society of Radiology.
Original source: H .Ha et al 2016: “Hemodynamic Measurement Using Four-Dimensional
Phase-Contrast MRI: Quantification of Hemodynamic Parameters and Clinical Applications”
Parameters of the axial flow-sensitive 2D PC-MR sequences were FoV: 4.5 x 6 cm,
MTX: 388 x 512, TE/TR/FA: 5 ms / 16.5 ms / 70°, slice thickness: 2.5 mm, spatial
resolution: 0.0116 x 0.0156 cm/pixel for the portal vein and FoV: 4.5 x 6 cm, MTX:
388 x 512, TE/TR/FA: 5 ms / 16.5 ms / 70°, slice thickness: 7.5 mm and spatial
resolution: 0.0116 x 0.0117 cm/pixel for the abdominal aorta.
Velocity encoding (Venc) settings were preset following literature values 82,202,
optimized in test measurements, and fixed at 18 m/min for the portal vein and 72
m/min for the abdominal aorta for the final experiment.
5.2.3.2 Data Acquisition / Postprocessing Data were taken directly from the software (ParaVision 5.1). The flow-sensitive 2D
PC-MR data were postprocessed by a blinded preclinical imaging expert (10 years of
experience) with a homebuilt analysis tool (MatLab® 14b), including noise filtering,
flow-compensated flow-encoded
Materials and Methods
106
correction for eddy currents and Maxwell terms, as well as velocity antialiasing 287.
2D PC-MR slices which were positioned orthogonally to the portal vein and the
abdominal aorta enabled the selection of the region of interest (ROI) of the two
vessels. Since ROI is equivalent to the vessel’s cross-sectional area, it was multiplied
by the corresponding flow velocity to calculate the flow volume rate (‘Volume rate =
Area * Velocity’) of the portal vein and the abdominal aorta for a cardiac cycle.
5.2.3.3 MR Assessment of the Degree of Liver Fibrosis Rat livers were scored via T1-RARE cross-sectional images based on morphological
hallmarks such as increased nodularity and irregular tissue appearance. The self-
established MR score is derived from the semiquantitative histological five-level
Desmet score (see 5.2.7.1). Only a four-level scoring system was used for the MR
score, as the difference between the first two levels described in the Desmet score is
not detectable with an MR approach. A blinded preclinical imaging expert performed
the evaluation of the MR rat liver images with the MR score (Figure 16): MR score=0:
no visible irregularities, MR score=1: minor irregularities and small nodular structures,
MR score=2: more prominent irregularities, solitary medium sized to large nodular
structures, and MR score=3: severe irregularities and prominent nodular structures
throughout the liver.
An independent and blinded radiologist repeated the MR scoring of the rat livers to
assess interobserver variability.
Materials and Methods
107
Figure 16: T1-RARE images showing morphological hallmarks of the MR score, such
as increased nodularity and irregular tissue appearance (arrows) with MR score=0:
no visible irregularities (a), MR score=1: minor irregularities and small nodules (b),
MR score=2: more prominent irregularities and medium-sized nodules (c), and MR
score=3: severe irregularities and prominent nodules (d)
5.2.4 Invasive Hemodynamic Measurements 5.2.4.1 Operative Procedure Before the invasive hemodynamic measurements were started rats were fasted for
1.5 h to avoid prandial effects on portal flow parameters. If rats have already passed
the MR measurements, they were again invasively measured two or three days after
scanning. Anesthesia was initiated in an animal induction chamber using a mixture of
3% isoflurane (Forene®) and 97% oxygen. It was maintained by an intraperitoneally
injected bolus of 0.3 - 0.4 ml pentobarbital [125 mg/ml] (pentobarbital sodium). After
having verified the depth of anesthesia, rats were shaved (hair clipper / trimmer
QC5115) and fixed on a homeothermic controlled operating table (Typ 872H), which
kept body temperature stable at 37 ± 0.5 °C. Vital parameters (i.e. heart and
respiration rates (HR), oxygen saturation) were determined by a pulse oximeter
(MouseOx®Plus) which was fixed on a hind paw.
Materials and Methods
108
After thorough disinfection (Kodan®) of the front neck region, a tracheotomy was
performed and a tracheal cannula was inserted. Since rats were mechanically
ventilated [50 breaths/min] (Rodent Ventilator Typ 7025), a muscle relaxation was
induced by intraperitoneal injection of 0.5 ml pancuronium [0.4 mg/ml] (pancuronium
bromide) to prevent spontaneous breathing.
To monitor central venous pressure (CVP) the right external jugular vein was
exposed and cannulated with PE-10 tubing (Becton Dickinson), which was positioned
near the right atrium. A second PE-10 tubing was inserted and was used to
compensate evaporative losses during the surgical procedure by a continuous
infusion of isotone electrolyte solution [1 ml/h] (Jonosteril®). The electrolyte solution
was enriched with pentobarbital [15 mg/ml] to ensure continuous anesthesia. Both
tubings were fixed with a ligature. To monitor mean arterial pressure (MAP) the left
carotid artery was exposed and cannulated with PE-50 tubing (Portex®). This tubing
was fixed with a ligature and was also used for the blood withdrawal at the end of
measurements (Figure 17). The surgical site around the front neck region was then
covered with wet gauze compress to avoid drying.
After thorough disinfection (Kodan®) of the abdomen, a median laparotomy (Figure
17) was performed and the portal vein was exposed. To measure the portal flow
volume rate an ultrasonic transit time flow probe (Transonic Animal Research
Flowmeter T206 with perivascular flow probe type 2.5S) was placed at the portal
vein, loosely encircling the vessel, before ultrasound gel (Caelo) was applied. After a
stabilization period of 10 to 15 minutes portal flow volume rate was measured over
five minutes without any intervention given, but presupposing a stable MAP.
The flow probe consists of a probe body which houses two ultrasonic transducers
and a probe reflector. The latter is positioned on one side of the vessel of interest,
whereas the two transducers are positioned on the opposite side. The first transducer
emits an ultrasonic beam that traverse the full width of the vessel. This beam is then
reflected by the probe reflector and captured by the second transducer. The time the
beam needs to travel from the first to the second transducer is termed “transit time”.
Basically, the beams traveling back and forth alternately cross the flowing blood in
upstream and downstream direction. During the upstream measurement the beam
travels against flow, and the resulting transit time is increased by a flow-dependent
factor. In contrast, during the downstream measurement, the beam travels with flow,
Materials and Methods
109
and the resulting transit time is reduced by the same flow dependent factor. The
subtraction of the upstream from the downstream integrated transit times provides an
accurate measure of the flow volume rate. Since the transit time is determined across
the cross section of the vessel, the flow volume rate is determined independently of
the vessel diameter 288.
To monitor portal venous pressure (PVP) the ultrasonic transit time flow probe was
removed and a peripheral venous catheter (Versatus®) was inserted into the portal
vein and fixed with a tissue adhesive (Histoacyrl®). In addition, a microvascular flow
probe (Laser Doppler blood flow monitor DRT4 with a Titanium tipped low profile disc
probe type DP8C) was set on the surface of the left liver lobe to determine
microvascular flow (MF). Thereby a laser beam (785 nm) emitted from the optic fiber
penetrates the liver tissue starting from the liver surface. The laser light emitted
interacts randomly with both, moving objects (primarily erythrocytes) and stationary
tissue. If the light hits a moving erythrocyte it is reflected in a different frequency and
magnitude of the wavelength (scatter) than it is emitted, i.e. a Doppler shift occurs. In
contrast, if the light is reflected from stationary tissue it remains unchanged. The light
changes induced by moving erythrocytes are detected by a photosensitive optic fiber
on the liver surface and returned to a photodetector and the signal processing
electronics. The Laser-Doppler system analyses the Doppler shift and calculates flux
values, a quantity proportional to the product of the average velocity of the
erythrocytes and their number concentration. Hence, erythrocyte motion in the
outmost layer of liver tissue was measured continuously. To provide consistent
measurements for all tissue types, the probes were calibrated with a motility standard
supplied with the monitoring system. The motility standard consists of a low
concentration of polystyrene microspheres in water undergoing thermal motion
(Brownian motion).
Materials and Methods
110
Figure 17: Insertion of catheter for MAP measurements and laparotomy.
A more detailed photo series of the operative procedure has been attached (see 7.2).
As a second measure, in addition to the homeothermic controlled operating table, the
torso of the rats was covered with aluminum foil to prevent it from becoming
hypothermic. After a stabilization period of 10 to 15 minutes, basal values of all
parameters were obtained and the intervention was administered through the second
CVP-tubing. The intervention was either sodium chloride (B. Braun Melsungen,
Melsungen, Germany) or sildenafil (Revatio®, Pfizer, Berlin, Germany). Rats were
randomly allocated in one of three intervention groups: sodium chloride (NaCl),
sildenafil 0.1 mg/kg (Sil 0.1 mg/kg) or sildenafil 1 mg/kg (Sil 1mg/kg). To minimize
hemodynamic alterations due to plasma volume changes, the intervention was
applied in a standardized volume of 600 µl.
To monitor arterial and venous pressures invasively, a solid column of liquid
connecting blood to the pressure transducer (ATP300 (arterial) or P75 type 379
(venous)) is required. Therefore tubings are be pre-filled with a heparinized isotone
electrolyte solution (Jonosteril®). The heparin sodium prevents occlusion of the
tubing due to thrombosis. The liquid within the tubing is in contact with a flexible
diaphragm which is located within the pressure transducer. The diaphragm moves in
response to the transmitted pressure waveform. The pressure transducer then
converts this movement into a proportional electrical signal (voltage, e.g.
5mV/V/mmHg), being enhanced by an amplifier module and send to the data
acquisition system (HSE-USB-HAEMODYN).
Materials and Methods
111
The high calibration point of the pressure transducers was calibrated with a pressure
manometer at 100 mmHg. Zero point calibration was performed 1 cm above the
operating table (heart height) by opening the pressure transducer to atmospheric
pressure and electronically zeroing the system.
Using the data acquisition system (HSE-USB-HAEMODYN) and the corresponding
software (HSE-Basic Data Acquisition Software (BDAS) 1.5) all measured
parameters were monitored continuously. Portal flow volume rate was recorded over
5 minutes before the intervention was applied. All other parameters (i.e. heart and
recorded over 60 minutes starting from time point “0min”. Right after determining
baseline values the intervention took place. Data were taken directly from the
software (BDAS).
5.2.5 Serum Analyses
5.2.5.1 Serum Parameters At the end of the invasive hemodynamic measurements blood samples were taken
via the left carotid artery. Serum was used for the analysis of the following serum
parameters by semi-automated clinical routine methods: glucose (Glc), sodium (Na),
potassium (K), total bilirubin (Bil), creatinine (Crea), albumin (Alb), aspartate amino-
transferase (AST), alanine aminotransferase (ALT), and alkaline phosphatase (AP).
5.2.5.2 Competitive cGMP Enzyme-linked Immunosorbent Assay (ELISA) A second serum sample was stored at −80 °C until used for the quantification of
cGMP concentrations. After defrosting, an in vitro competitive ELISA was performed
according to the manufacturers’ instructions of the cGMP ELISA kit (Abcam
ab133052). In total, two kits, including one 96 well plate each, were used. Each
sample was assayed in duplicates. In each set of experiments a standard curve was
assayed for calibration using standard serial dilution. In addition to the serum
samples, a negative control (blank) and two positive controls (two different standard
samples) were included. One out of these two positive controls was applied in six
replicates to determined inter- and intra-assay variability.
Materials and Methods
112
A goat anti-rabbit IgG capture antibody (secondary antibody) has been precoated
onto the wells. 100 µl of standard or serum samples (undiluted, non-acetylated) were
incubated togheter with 50 µl of an alkaline phosphatase (AP)-conjugated cGMP
antigen and 50 µl of a polyclonal rabbit cGMP antibody (primary antibody) per well at
room temperature for two hours on a plate shaker at 500 rpm. Thereby the AP-
labeled antigens from the conjugate and the unlabeled antigens from the serum
sample compete for binding to the target specific antibody (primary antibody) which
in turn binds to the capture antibody (secondary antibody).
After incubation, the excess reagents (unbound antibodies and other biological
materials) were removed by three wash steps using each 400 µl wash buffer per well.
Then 200 µl of substrate solution were applied and incubated at room temperature
for one hour without shaking. As a phosphatase substrate p-Nitrophenylphosphate
(pNpp) is used, which turns yellow, when dephosphorylated by ALP. The intensity of
the color change is inversely proportional to the amount of cGMP in the well. To
quench the enzyme reaction, 50 µl stop solution were pipetted into each well.
Immediately after, optical density absorbance at 405 nm was read using a scanning
microplate spectrophotometer (µQuantTM) and the corresponding software (KC4).
The fluorescence data were taken directly from the software (KC4). Since each
sample was assayed in duplicates, mean values of cGMP concentration were
determined and used for further statistical calculations.
5.2.6 Two-step Quantitative Real-time Polymerase Chain Reaction (qRT-PCR) Sample preparation, mRNA extraction, and cDNA synthesis At the end of the invasive hemodynamic measurements, the left lateral lobe of each
rat’s liver was excised, cut into pieces, snap frozen in liquid nitrogen, and stored at
−80 °C until used for qRT-PCR. qRT-PCR was used to quantified hepatic gene
expression of endothelial and inducible NO synthase (eNOS and iNOS),
were rinsed gently with wash buffer (1x PBS (Oxoide) + 0.1% Tween® 20 (PancReac
AppliChem)) before 1 to 2 drops of a peroxidase labelled polymer, which was
conjugated to a goat anti-rabbit IgG capture antibody (secondary antibody), were
applied on each section. Sections were covered with a sealing film (Parafilm M®) and
incubated for 60 min at 24 °C. During incubation the capture antibody reacts with the
anti-PDE5A-antibody, which has already bound to its target antigen. The labelled
polymer does not contain avidin or biotin, thus any nonspecific staining as a
consequence of endogenous avidin-biotin activity in the liver is eliminated or
significantly reduced. Afterwards sections were rinsed gently with wash buffer (1x
PBS (Oxoide) + 0.1% Tween® 20 (PancReac AppliChem)). In the next step
substrate-chromogen solution (DAB) was applied on each section and incubated for
10 min at 24 °C. During incubation antibody binding is visualized on the basis of the
occurring brown color reaction. Once more sections were rinsed gently with wash
buffer (1x PBS + 0.1% Tween® 20). (Figure 18)
Figure 18: Principle of the polymer two-step indirect technique for immunhisto-
chemical staining.
Figure reprinted with permission of Sudhanshu Goyal, employee of BioGenex.
Original source: http://www.biogenex.com/us/detection-systems (July 24, 2017)
To provide contrast that helps the primary stain stand out, a hematoxylin nuclear
counterstaining step was performed in which sections were incubated in filtered
hematoxylin (Sigma-Aldrich) for 5 min at 24 °C.
Materials and Methods
120
Dehydration
After the hematoxylin counterstaining sections were dehydrated. Therefore the
following wash steps were performed by immersing the sections in water and a
series of alcohols:
• running cold water (demineralized) for 5 min
• ethanol (80%) for 5 min
• ethanol (95%) for 5 min
• ethanol (100%) for 2 x 5 min
• xylene 2 for 10 min
• xylene 1 for 10 min
Finally, sections were mounted and coverslipped using a mounting medium
(Entellan®).
Microscopic analysis
For microscopic quantification (Zeiss Axioplan microscope) the number of stained
cells was counted in 20 random high power-fields (HPF) (400x magnification) for
each sample. Exclusively stained cells in the parenchyma were included in cell
counts, whereas PDE5 staining around the central vein (CON 1) and in fibrous
connective tissue (CIR 1 and CIR 2) was not considered.
5.2.8 Statistics Results were expressed as median ± interquartile range (IQR). Only results of the
qRT-PCR experiments were expressed as mean ± standard deviation (SD) to enable
the quantification of gene expression with the comparative Ct method.
To determine differences among groups or subgroups for the measured parameters,
the non-parametric Kruskal-Wallis test was used. Post-hoc pairwise comparisons
between groups or subgroups 179 were corrected for multiple comparisons according
to Bonferroni. For reasons of consistency the non-parametric Kruskal-Wallis test and
post-hoc pairwise comparisons with Bonferroni correction were also used for the
qRT-PCR experiments. A two-tailed p-value of < 0.05 was considered as statistically
significant.
Materials and Methods
121
For statistical analyses SPSS® software 23.0 (IBM Corp., Armonk, New York) was
used, only the regression analysis was calculated with STATA® software 14
(StataCorp LLC, Lakeway Drive, Texas).
Specific statistic information for each part of the study will be described in the
following.
Non-invasive hemodynamic measurements
To assess interobserver variability (2 readers) of the self-established MR score, a
weighted kappa analysis was performed. The kappa coefficient (ƙw) indicates the
strength of agreement and was categorized as follows: 0–0.20: slight; 0.21–0.40: fair;
0.41–0.60: moderate; 0.61–0.80: substantial; and 0.81–1.00: almost perfect 291.
Correlations were calculated using Spearman’s rank correlation coefficient (rs).
Invasive hemodynamic measurement
Invasive portal flow volume rate measurement
Correlations were calculated using Spearman’s rank correlation coefficient (rs).
Effect of sildenafil on hemodynamics
All parameters were normalized (PVPrel, MAPrel, MFrel and HRrel, CVPrel, respiration
raterel and oxygen saturationrel) to compensate differences in absolute values
between healthy and diseased rats. Thereby, time point “10 min” was taken as
baseline and set to 100% since the administration of 600 µl liquid volume into the
right atrium caused parameter variations for the next few minutes.
To evaluate the effect of sildenafil the change in parameter values at time point “60
min” compared to baseline (“10 min”) was determined by calculating relative median
of differences (RMD).
To illustrate the course of parameters during the measurement interval data of time
points 0, 1, 3, 5, 10, 20, 30, 45 and 60 min were chosen.
Materials and Methods
122
Effect of MAP on PVP
All parameters were normalized (PVPrel, MAPrel) to compensate differences in
absolute values between healthy and diseased rats. Thereby, time point “0 min” was
taken as baseline and set to 100%. The effect of MAP on PVP was evaluated by
linear regression analysis, including changes from baseline (“0 min”) for every
recorded time stamp (1 time stamp = 2 sec) over the first 30 minutes.
The change in PVPrel for every 1% change in MAPrel is described by the regression
coefficient (s), whereas the explained variation (%) within one rat is described by r-
squared (r²).
Biochemical investigations
Levels of gene expression were quantified relatively with the comparative Ct method 289. The individual mRNA levels in the samples were normalized for two reference
genes. Analyses of both, gene expressions and cGMP concentrations, were
performed in duplicates. Mean values of duplicates were used for further statistical
analyses.
For microscopic quantification of the immunohistochemical stainings, the number of
PDE5 stained cells was counted in 20 random high power fields (HPF) for each
tissue sample. Mean values of stained cells per HPF were used for further statistical
analyses.
123
6. References 1. Gao, B., Xu, M.-J. & Zhou, Z. Hepatocytes: a key cell type for innate immunity. Cellular &
Molecular Immunology 13, 301 (2015).
2. Si-Tayeb, K., Lemaigre, F. P. & Duncan, S. A. Organogenesis and Development of the Liver.
Developmental Cell 18, 175–189 (2010).
3. Muriel, P. Liver Pathophysiology: Therapies and Antioxidants. (Academic Press, 2017).
4. Brunt, E. M. et al. Pathology of the liver sinusoids. Histopathology 64, 907–920 (2014).
5. Lautt, W. W. Regulatory processes interacting to maintain hepatic blood flow constancy: