Aus der Klinik für Gastroenterologie, Hepatologie und Endokrinologie der Medizinischen Hochschule Hannover Zentrum Innere Medizin (Direktor: Prof. Dr. med. Michael P. Manns)
Genetic deletion of the carbonic anhydrase XIV induces low
biliary bicarbonate output and enhances the toxic effect of bile acids to the cholangiocytes - an “in vivo” confirmation of
the “biliary bicarbonate umbrella” hypothesis
Dissertation Zur Erlangung des Doktorgrades der Medizin in der Medizinischen Hochschule Hannover vorgelegt von Jiajie Qian aus Hangzhou, China Hannover 2016
Angenommen vom Senat der Medizinischen Hochschule Hannover am Gedruckt mit Genehmigung der Medizinischen Hochschule Hannover
Präsident: Prof. Dr. med. Christopher Baum
Wissenschaftliche Betreuung: Prof. Dr. med. Ursula Seidler
1. Referent: Prof. Dr. med. Ruthild Weber 2. Referent: Prof. Dr. med. Claus Petersen
Tag der mündlichen Prüfung: 04.10.2017
Prüfungsausschuss:
Vorsitz: Prof. Dr. med. Tobias Welte 1. Prüfer: Prof. Dr. med. Carlos Guzman 2. Prüfer: PD Dr. med. Frank Gossé
TABLE OF CONTENTS
TABLE OF CONTENTS ...... I
ABBREVIATIONS ...... IV
LIST OF TABLES AND FIGURES ...... V
1.Introduction ...... 1 1.1. General Introduction ...... 1 1.1.1. The Bicarbonate Umbrella ...... 3 1.1.2. Major ion carriers that may regulate pH in BECs ...... 3 1.1.2.1. Bicarbonate extruder ...... 3 1.1.2.2. Acid extruder or bicarbonate loader ...... 6 1.2. Carbonic Anhydrase XIV ...... 7 1.3. Study of Liver function: Liver injury models ...... 9 1.3.1. Chemically-induced cholangitis:DDC-fed mouse model ...... 10 1.3.2. Knock-out mouse models: Mdr2/Abcb4-deficient mice ...... 11 1.4. Tauroursodeoxycholic acid ...... 13 1.5. Aim of the study ...... 14
2. Materials and Methods ...... 15 2.1. Animals ...... 15 2.2. Diet and feeding protocol ...... 15 2.3. Reagents ...... 16 2.4. In vivo biliary drainage experiments ...... 17 2.5. Measurement ...... 18 2.5.1. Measurement of bile flow ...... 18 2.5.2. Measurement of biliary bicarbonate secretion ...... 18 2.6. Handling of tissue and samples ...... 19 2.6.1. Blood Sample ...... 19 2.6.2. Liver tissue work-up ...... 19 2.7. Staining protocol ...... 20 2.7.1. Histology staining ...... 20 2.7.2. CK 19 Immunohistochemistry Staining ...... 21 2.8. Statistics ...... 21
3. Results ...... 23
I
3.1. Monitoring of vital signs ...... 23 3.2. Loss of carbonic anhydrase 14 results in decreased biliary bicarbonate output and a slowly progressive cholangiopathy ...... 25 3.2.1. Bile flow and biliary bicarbonate output rates in Car14-/- and WT mice at different ages ...... 25 3.2.2. Histological analysis ...... 29 3.3. Car14 expression cannot protect against the severe toxic cholangiopathy induced by DDC-feeding ...... 36 3.3.1. DDC cholangiopathy model development ...... 36 3.3.2. DDC-induced decrease in biliary secretion in Car14 WT and KO mice .. . 38 3.3.3. Morphological and histological evaluation of the DDC-fed liver of Car14 KO and WT mice ...... 40 3.4. Loss of carbonic anhydrase XIV aggravates the bile-acid induced cholangiopathy of the Mdr2-/- mouse model ...... 44 3.4.1. Result of the in vivo experiments ...... 44 3.4.2. Result of histological analysis...... 49
4. Discussion ...... 60 4.1. Improvement of experimental design ...... 60 4.1.1. Choosing the appropriate model to study the importance of biliary bicarbonate output on cholangiocytes health in vivo (the so-called “biliary bicarbonate umbrella” hypothesis) ...... 60 4.1.2. Hematocrit determination and continuous fluid resuscitation during the biliary drainage period ...... 62 4.1.3. Advantage of direct collection of bile from the hepatic duct ...... 62 4.1.4. The pH-stat titration technique is optimal for exact assessment of bicarbonate output ...... 63 4.1.5. Control of the systemic acid/base status during the in vivo biliary drainage ...... 63 4.1.6. Being able to get at the “liver function reserve” in order to assess early biliary functional defects...... 64 4.2. Role of the CAXIV in biliary physiology ...... 65 4.3. Using the DDC-fed mice as a pathological model to study the protective effects of a higher bicarbonate output into bile in the presence of CAXIV expression ...... 68 4.4. The Mdr2-/- mouse model ...... 69 4.5. The protective function of “bicarbonate umbrella” in liver ...... 74 4.6. Conclusion ...... 76
5. References ...... 77
II
6. Abstract ...... 87
7. Eigenständigkeiterklärung ...... 89
8. Curriculum Vitae ...... 90
9. Acknowledgement…………………………………………………………………………………………..92
III
ABBREVIATIONS
GCDC glycochenodeoxycholate
PBC primary biliary cirrhosis
PSC primary sclerosing cholangitis
IAC IgG4-associated cholangitis,
CFALD cystic fibrosis-associated liver disease
BECs biliary epithelial cells
PHi intracellular PH
AE2 anion exchanger protein 2
CFTR cystic fibrosis transmembrane conductance regulator
NBCs Na+-Bicarbonate Cotransporters
NHEs Na+/H+ exchangers
NDCBEs Na+-Driven chloride bicarbonate exchangers
Car14 c arbonic anhydrases XIV
DDC 3,5-diethoxycarbonyl-1,4-dihydro-collidine
Mdr2 multidrug resistance protein 2
TUDCA tauroursodeoxycholic acid
KO knock-out
WT wild-type
CK19 cytokeratin-19
MAP mean arterial pressure
BG blood gas
PFA paraformaldehyde
IV
LIST OF TABLES AND FIGURES
Tab.1 Mouse strains used in this study...... 15
Tab.2 List of reagents...... 16
Tab.3 Blood gas analysis of Car14 mice and DDC-fed Car14 mice...... 24
Tab.4 Blood gas analysis of Mdr2 mice...... 24
Fig.1 Portal triad...... 2
Fig.2 Bicarbonate extruder...... 4
− Fig.3 Acid extruders or HCO3 loaders...... 6
Fig.4 Structure of the murine CA XIV...... 8
Fig.5 The mechanism of DDC-induced sclerosing cholangitis...... 11
Fig.6 The possible mechanism of sclerosing cholangitis in Mdr2−/− mice ..... 12
Fig.7 Liver preparation and processing...... 20
Fig.8 Biliary bicarbonate output and bile flow in 11 weeks old Car14-/- mice..
...... 26
Fig.9 Biliary bicarbonate output and bile flow in 20 weeks old Car14-/- mice.
...... 27
Fig.10 Biliary bicarbonate output and bile flow in 1 year old Car14-/- mice..
...... 28
Fig.11 Liver Hematoxylin and eosin(H&E)staining in central part of Car14-/-
mice at different ages...... 30
Fig.12 Liver Hematoxylin and eosin(H&E)staining in peripheral part of
Car14-/- mice at different ages...... 31
Fig.13 Liver CK19 immunohistochemistry staining in central part of Car14-/-
mice at different ages...... 32
Fig.14 Liver CK19 immunohistochemistry staining in peripheral part of
Car14-/- mice at different ages...... 33
Fig.15 Liver Sirius red staining in central part of Car14-/- mice at different
V
ages...... 34
Fig.16 Liver Sirius red staining in central part of Car14-/- mice at different
ages...... 35
Fig.17 Biliary bicarbonate output and bile flow in 0.1% DDC-fed C57BL/6N
mice...... 37
Fig.18 Biliary bicarbonate output and bile flow in DDC-fed Car14 mice. .... 39
Fig.19 Appearance of normal liver and DDC fed liver...... 40
Fig.20 Liver H&E staining of DDC-fed mice...... 41
Fig.21 Liver CK19 immunohistochemistry staining of DDC-fed mice...... 42
Fig.22 Liver Sirius red staining of DDC-fed mice...... 43
Fig.23 Biliary bicarbonate output in Car14/Mdr2 mice at age of 6 weeks. .. 45
Fig.24 Bile flow in Car14/Mdr2 mice at age of 6 weeks...... 46
Fig.25 Biliary bicarbonate output in Car14/Mdr2 mice at age of 11 weeks.. 48
Fig.26 Bile flow in Car14/Mdr2 mice at age of 11 weeks...... 49
Fig.27 Liver H&E staining of Car14/Mdr2 hybrid mice at age of 3 weeks.. 50
Fig.28 Liver CK19 immunohistochemistry staining of Car14/Mdr2 hybrid
mice at age of 3 weeks...... 51
Fig.29 Liver Sirius red staining of Car14/Mdr2 hybrid mice at age of 3 weeks..
...... 52
Fig.30 Liver H&E staining of Car14/Mdr2 hybrid mice at age of 6 weeks.. 54
Fig.31 Liver CK19 immunohistochemistry staining of Car14/Mdr2 hybrid
mice at age of 6 weeks...... 55
Fig.32 Liver Sirius red staining of Car14/Mdr2 hybrid mice at age of 6 weeks..
...... 56
Fig.33 Liver H&E staining of Car14/Mdr2 hybrid mice at age of 11 weeks...
...... 57
Fig.34 Liver CK19 immunohistochemistry staining of Car14/Mdr2 hybrid
mice at age of 11 weeks...... 58
VI
Fig.35 Liver Sirius red staining of Car14/Mdr2 hybrid mice at age of 11
weeks...... 59
Fig.36 Mechanisms of bile duct injury in the Mdr2 (Abcb4)–/– mouse
cholangiopathy model...... 72
VII
1. Introduction
1.1. General Introduction
The liver has a vital function in the human body and plays important roles in
metabolism, digestion and excretion, including the regulation of glycogen storage, the
secretion of bile acids and phospholipids, the excretion of bilirubin and cholesterol,
plasma protein synthesis, hormone production and detoxification1. As an accessory digestive gland, it produces alkaline bile, which is composited by 97% water, 0.7% bile acids, 0.2% bilirubin, 0.51% fats (cholesterol, fatty acids and lecithin)2 and 200
meq/l inorganic salts3. Bile acids are conjugated with taurine or glycine in the liver,
forming the primary bile salts. They help in emulsification of fat and formation of fat micelles, which increases their contact with the intestinal brush border membrane4.
The secondary bile slats are those who have been altered by bacterial metabolism in the
intestine and reabsorbed first by the enterocytes, entering the enterohepatic circulation,
and then by the hepatocytes and secreted again into be bile.
The biliary ductal system transfers bile juice from hepatocytes to intestinal tract.
Any pathological changes of bile composition, biliary anatomy and function will
cause biliary disease, thus influence biliary flow. On the contrary, the bile juice itself
is a very important reason for bile duct disease. The acidic toxicity of high millimolar
levels of hydrophobic bile salt monomers such as glycochenodeoxycholate (GCDC)
can injure cholangiocytes which may ultimately develop to pathological changes5,6.
Many bile duct diseases finally lead to end stage of hepatic failure, such as primary
biliary cirrhosis (PBC), primary sclerosing cholangitis (PSC), IgG4-associated
cholangitis (IAC), cystic fibrosis-associated liver disease (CFALD), or non-anastomotic stricture (NAS) after liver transplantation.
Since the bile canaliculus is the space formed between rows of hepatocytes,
forming a continuous groove covered with the hepatocyte apical surface (Fig.1),
injury to hepatocytes and to cholangiocytes cannot be distinguished at the level of the
1
primary bile ducts. Cholangiocytes are the epithelial cells that line the lumen of the bile
ducts once the bile canaliculi have merged to larger structures. Hepatocytes and
cholangiocytes act as facultative stem cells for each other, which means when cholangiocytes are damaged that the hepatocytes surround them have the capability to be transferred into cholangiocytes to repair the damage and vice versa7,8.
Physiologically, they both play a role in secretion of bicarbonate ions into the biliary
lumen. Bicarbonate ions are important for the alkalinization of bile and thought to
protect the bile ducts from damage by bile acid toxicity as well as other organic acids,
a concept that is called the “biliary bicarbonate umbrella”, which has so far not been
unequivocally proven in vivo. Hepatocytes and cholangiocytes form a single
functional unit in the formation of bile and are involved collectively in a biliary
pathology and hence the cells covering the luminal surface of biliary tree are referred
together as “biliary epithelial cells (BECs)”.
Fig. 1 Portal triad. Consists of a biliary (represented here as “bile ductules”), hepatic arterial, and portal venous component9.
2
1.1.1. The Bicarbonate Umbrella
In recent years, the concept of ‘biliary bicarbonate umbrella’ is considered as a
critical protective mechanism of cholangiocytes. As we know, intracellular pH (pHi)
plays a key role in many cellular functions10,11. To maintain the intracellular pH in a
narrow physiological range, a variety of ion carriers including channels, pumps,
exchangers and transporters have collaborative work to import or export acid and base
equivalents12,13. But not only the intracellular pH is regulated, recent results suggest that epithelia also regulate a luminal pH-microclimate close to their apical surfaces.
This concept was first established for the gastric epithelium, but has been shown to
apply also to small and large intestine, the lung, and most recently, the pancreatic and
14,15 - biliary ductal system . According to this hypothesis, abundant apical biliary HCO3
secretion is considered an important protection against uncontrolled penetration of
protonated bile acid monomers and hydrophobic bile salt monomers which can lead to
biliary epithelial cell damage. A defect in this biliary umbrella function could
contribute to development and progression of various chronic cholangiopathies
including PBC, PSC, CFALD, or NAS.
1.1.2. Major ion carriers that may regulate pH in BECs
1.1.2.1. Bicarbonate extruder
3
Fig. 2 Bicarbonate extruder: AE2 transporter exchanges one Cl− ion for a HCO3− ion transported out of the cells. CFTR protein in turn extrudes Cl− ion to the extracellular side to create a higher Cl− gradient. NBC protein couples efflux of HCO3− ions and Na+ ion at a ratio of 3:1; and is regulated by the pHi, concentration of intracellular cAMP and Ca2+, the mechanism still remains unknown16
+ − − AE2 Na -independent Cl / HCO3 exchanger Both in human and rodent17 cholangiocytes, the output of bicarbonate to bile
+ − − takes place through a Na -independent Cl /HCO3 exchanger mediated by the acid
loader of the SLC4/SLC26 family which is generally involved in pHi regulation.
Many members of SLC4 and SLC26 families can exchange an influx of Cl− with a
− + − − 18 HCO3 efflux as a Na -independent Cl /HCO3 exchanger (Fig.2) , but only
SLC4A2(also known as AE2) has been identified in the liver19. AE2 imports one Cl−
− and extrudes one HCO3 and has been identified immunohistochemically on
membranes of hepatocytes with predilection in the periportal area19. AE2 gene
knock-down cholangiocytes in human and rodent demonstrated that AE2 plays a
− − 17,20,21 critical role in Cl / HCO3 exchanging . Interestingly, NBC1 was found in more
active function and higher mRNA expression in AE2-/- mice experiments22. These data
4
− suggest that NBC1 will take over the HCO3 excluding when AE2 is defective. On the
other hand, both AE2 and NBCe1 are expressed exclusively basolaterally in all other
epithelia, and it is unclear at this point if AE2 and NBCe1 are truly expressed apically
in hepatocytes, or if members from the Slc26a family, who have been identified as the
apical anion exchangers in most epithelia to date, are performing this function also in
the hepato- and cholangiocytes.
Cystic fibrosis transmembrane conductance regulator (CFTR)
Cystic fibrosis transmembrane conductance regulator (CFTR) is classically
described as a Cl− channel. It has important roles as a hub protein in the regulation and
in co-operative transport function of many transport proteins. It has been shown to play a role in transport of HCO3−, the mechanism of which is still unclear. In the
intestinal tract, CFTR was reported to transport Cl− ion in exchange of HCO3− ion with a stoichiometry of 1:123–25. CFTR protein is more likely to be co-localized with
AE2 transporter in the cholangiocytes. Efflux of Cl− by CFTR in these cells probably
creates the necessary gradient for the eventual HCO3− secretion into the lumen by
AE2, in exchange of apical Cl− ion (Fig.2)26,27.
Na+-Bicarbonate Cotransporters (NBCs)
+ − The function of Na -Bicarbonate cotransporters (NBCs) couple HCO3 influx
with a downhill influx of Na+, but the stoichiometry and the direction of transport flux
depend upon the location of the transporter and intracellular environment. NBC is
known to mediate the bicarbonate influx in mouse cholangiocytes28. The isoform
+ − Na -HCO3 cotransporter 1 (NBCe1) was found in basolateral membrane of
− pancreatic ducts, gastrointestinal and renal epithelial cells. It couples HCO3 with
+ − downgrade influx of Na from extracellular side to intracellular side, with the HCO3 :
Na+ coupling ratio of 2:1 for the gastrointestinal subtype pNBC or NBCe1B, and 3:1
for the renal isoform kNBC or NBCe1A. When the stoichiometry is 3:1, the direction
5
of bicarbonate transport is efflux rather than influx. Possibly a change in
stoichiometry can occur also in a regulated fashion, by changes in pHi, concentration
of intracellular cAMP and Ca2+, the mechanism of which is still unknown (Fig.2 and
Fig.3 NBC*)29,30.
1.1.2.2. Acid extruder or bicarbonate loader
− + + Fig. 3 Acid extruders or HCO3 loaders: NHEs extrude intracellular H and load Na from + − extracellular surface with a stoichiometry 1:1. In addition, Na /HCO3 cotransporters (NBC) and + - − − 16 Na -dependent Cl / HCO3 exchanger (NDCBE) may load cholangiocytes with HCO3 .
Na+/H+ exchangers (NHEs) Na+/H+ exchangers (NHEs) are a family of ATP-independent membrane
transporters which are involved in many cellular functions, such as regulation of pH,
cell volume and the cellular response to hormones and mitosis. These exchangers
mediate the simultaneous efflux of one H+ ion and the influx of one Na+ ion through the
membrane, the driving force being the inwardly directed electrochemical Na+ gradient. 6
There are 9 known mammalian NHE isoforms (NHE1-9), all of which differ in their
biological function and as well as their subcellular localization. The output of H+ from
intracellular side to ductular lumen is mainly through NHEs31. The major members of
NHEs family are NHE1-3 which present in most mammalian cells. They have
different locations, functions, regulatory mechanisms, inhibitors, but all of them play
− a crucial role in maintenance pHi homeostasis. In HCO3 loading cells, carbonic
anhydrase catalyzes CO2 hydration by the following reaction: CO2+ H2O ↔
− + + HCO3 +H ; the H produced in turn is extruded by NHEs. NHE1 is primarily located
in the basolateral membrane, where it plays important roles in regulation of
intracellular pH and cell volume32. It is believed that the NHE2 and NHE3 isoforms are responsible for the majority of Na+ and fluid absorption from the apical membrane
of bile duct (Fig.3)33,34.
+ − − Na -driven Cl / HCO3 exchangers (NDCBEs) NBCs are found staying an inactive status in human cholangiocytes under
physiological intracellular pH, and the influx of HCO3− is carried out by Na+-driven
Cl−/HCO3− exchangers (NDCBEs) which is located in the basolateral membrane35.
NDCBE transports two HCO3− ion and one Na+ ion into the cell in exchange of transporting one Cl− ion out of the cell. SLC4A8 is the only known isoform of
NDCBEs known to be expressed in human liver36.
1.2. Carbonic Anhydrase XIV
Carbonic anhydrases (CAs) are zinc-containing enzymes that catalyze the
reversible hydration of carbon dioxide, and accelerate the spontaneous reaction by up to 6,000 times:
- + CO2+ H2O ⇔ HCO3 +H
They are a large family of enzymes with great physiological importance and 7
participate in a wide range of physiological processes in the vertebrate organism, such
as pH homeostasis, carbon dioxide and ion transport and respiration. The different
isoforms have a distinct tissue expression, but overall carbonic anhydrases are present
in every tissue37.
CAXIV was first discovered by Fujikawa-Adachi et al. in 1999. When the group of scientists were searching an EST database for sequences homologous to CAs, they
identified a partial cDNA encoding CAXIV38. CAXIV is a membrane-associated
protein with a 275 amino acid long extracellular catalytic domain, a transmembrane
helix, and a short intracellular polypeptide at the C-terminus (Fig. 4).
Fig. 4 Structure of the murine CA XIV: 275 amino acid long extracellular catalytic domain is shown in ball-and-stick representation outside the cell membrane as a functional structure37.
Using in situ hybridization, CAXIV was localized in mouse renal proximal
convoluted tubule, which is the major segment for bicarbonate reabsorption, and also
in the outer border of the inner stripe of the outer medulla39–41, all these results
suggested that CAXIV probably contributes to the bicarbonate reabsorption in kidney.
The most interesting research of CAXIV outside liver is in the nervous system. It
8
has been reported to be a key element in the normal bicarbonate homeostasis in
excitable tissues, acting synergistically with Cl−/HCO3− exchanger protein AE342. In the central nervous system, the catalytic domain of CAXIV is anchored to the extracellular surface and associate with AE3 forming a bicarbonate transport
− metabolon. Acting together, they maximize the HCO3 influx across the plasma
membrane of neuronal and glial cells43.
High mRNA expression of CAXIV has been reported in liver tissue by northern blots analysis38,39. In histochemical staining, hepatocytes displayed a plasma
membrane-associated carbonic anhydrase activity44. The predominant canalicular localization of this protein probably plays a role in regulation of pH and ion transport between the hepatocytes, sinusoids and bile canaliculi45.
We can summarize a clue from previous descriptions. In nervous system, it has
been proved that CAXIV plays important roles in bicarbonate homeostasis in
cooperation with AE3. In the liver, high expressed CAXIV was located in the apical
membranes of hepatocytes, in which the major bicarbonate extruder is AE2, a
homolog of AE3. But since the catalytic site of the CAXIV is in the lumen, it would
- - not necessarily help HCO3 export into the lumen by increasing the HCO3 availability
at the inner transport site of AE2. Therefore, physiological role of the high expression
of CAXIV in the canalicular membrane requires further study.
1.3. Study of Liver function: Liver injury models
The liver has a rapid regenerating capacity and a high functional reserve. In order
to study the protective effects of biliary bicarbonate, and the alkaline microclimate
adjacent to the canalicular membrane of hepatocytes and cholangiocytes, the so called
“bicarbonate umbrella”, it is therefore important for us to establish a model of liver
injury, in order to find evidence for this protective mechanism to operate in an in vivo
pathophysiological context. Mouse models for sclerosing cholangitis have
9
demonstrated inflammation and fibrosis of the bile ducts in liver, and also an altered
composition of the bile with higher concentration of bile salts9,46,47. These models
may be useful models to study this protective function. Mdr2/Abcb4-deficient mouse
model and 3,5-diethoxycarbonyl-1,4-dihydro-collidine (DDC) fed mouse model are
the two of the most popular and well-established mice models of cholangitis
nowadays.
1.3.1. Chemically-induced cholangitis: DDC-fed mouse model
DDC is a dihydropyridine with 2,4,6-trimethyl-1,4-dihydropyridine substituted by ethoxycarbonyl groups at positions 3 and 5.
Chronic feeding of DDC in mice was first used for studying the formation of
Mallory bodies which are hepatocytes inclusion bodies characteristically associated with alcoholic cirrhosis, metabolic liver diseases, and chronic cholestatic liver diseases48,49. Further investigation also exhibited that DDC feeding leads to inflammation, sclerosing cholangitis, a reactive cholangitic phenotype and biliary type liver fibrosis50.
10
Fig. 5 The mechanism of DDC-induced sclerosing cholangitis50
The toxic damage of DDC leads to a reaction of biliary epithelial cells, in which
they express proinflammatory and profibrogenetic cytokines and adhesion molecules,
such as TNF-α,VCAM, PDGF, TGF-β. These cytokines and molecules cause
infiltrations of the portal fields by neutrophils51,52. Then the portal-to-portal bridge
fibrosis appears as a ductal reaction after the injury of biliary epithelial cells (BECs)
getting worse. Inflammation of periductal induction promotes fibrosis, induces
proliferation of fibroblasts and fibrocytes, and deposition of extracellular
matrix(ECM)53. At the same time, porphyrin emboli inside the bile ducts aggravate the cholangitis and fibrosis (Fig.5).
1.3.2. Knock-out mouse models: Mdr2/Abcb4-deficient mice
Mdr2(Abcb4) is a murine gene homologue of the human Mdr3(Abcb4) gene encoding for a canaliculi phospholipid transporter54. The function of this protein
involves the transport of phospholipids from the liver hepatocytes into the bile.
Mdr2−/− mice have a complete absence of phosphatidylcholine from bile55, and
develop hepatic inflammation, fibrosis, sclerosing cholangitis, and fibrous obliteration 11
of bile ducts.
Fig. 6 The possible mechanism of sclerosing cholangitis in Mdr2−/− mice 56
The lack of phospholipid in bile in Mdr2−/− mice results in a higher concentration of a change in the micellar composition, a loss of vesicular transport of cholesterol and bile acids, an increase in the free bile acids56. The toxic effects of the bile salts
cause inflammation of biliary epithelial cells, increase in tight junction permeability and subsequently develop into biliary ductal leakage57,58. In addition, the bile salts
leaked into the periductal area lead to a secondary portal inflammation with
infiltration of CD4-/CD8-rich lymphocytes and neutrophil granulocytes59,60. The
inflammation also activates periductal myofibroblasts to form a fibrosis around the
bile duct and subsequently affect the connection between BECs and their vascular
plexuses. Finally, BECs atrophy and death develops, which lead to fibrosa obliterans
of the bile duct (Fig.6). During the progression of the disease, a likely alteration in the
enterohepatic circulation will also take place, potentially with more toxic hydrophobic
secondary bile acids in the hepatic bile.
From previously description, we can summarize that both DDC-fed mice model
12
and Mdr2−/− mice model develop to inflammation, fibrosis, and sclerosing cholangitis
of bile duct, they may be suitable animal models to investigate the function of CAXIV
and the “bicarbonate umbrella theory”.
1.4. Tauroursodeoxycholic acid
Tauroursodeoxycholic acid (TUDCA) is a hydrophilic bile acid which is the
taurine conjugate form of ursodeoxycholic acid (UDCA).
UDCA/TUDCA was first identified by Chinese, who have used it to treat liver
and eye-related diseases for centuries, from the bile of the black bear. It was first
manufactured synthetically in 1954 in Japan61. Nowadays, UDCA/TUDCA is used
clinically to treat cholestatic liver disorders.
The major protective functions of TUDCA include following four aspects62:
TUDCA replace the hydrophobic acids in bile; the cytoprotection to the hepatocytes and BECs; immunomodulatory effects; and stimulation of bile secretion by hepatocytes and bile duct epithelial cells with higher bile flow and bicarbonate-rich
composition.
In rodents, the bile flow and bicarbonate output of the bile was significantly
increased, within a few minutes of TUDCA infusion63,64. TUDCA is described to
− stimulate the exocytosis from BECs and also to activate HCO3 anion transporters,
such as AE2 in the liver 16,21,62.
13
1.5. Aim of the study
The aim of this study includes:
1) To study the role of carbonic anhydrase 14 in biliary bicarbonate secretion.
2) To evaluate the development of sclerosing in different mouse models of liver injury, and to evaluate the changes in total bile flow secretion and biliary bicarbonate output in the same models.
3) To study the role of carbonic anhydrase 14 in development and protection against the biliary injury in the mouse models of cholestatic liver injury.
14
2. Materials and Methods
2.1. Animals
Mouse Strain
Car14 B6.129S1-Car14tm1Sly
tmBor Mdr2 FVB/N-Abcb4
tmBor Car14/Mdr2 B6.129S1-Car14tm1SlyxFVB/N-Abcb4
B6N C57BL/6NCrl
Tab. 1 Mouse strains used in this study.
All mice were bred in in the animal facility of Hannover Medical School. All of these animal strains were maintained with controlled light/dark cycles and free access to water and food. All experiments were approved by the Local Institutional Animal Care and Research Advisory Committee at the Hannover Medical School and authorized by the Niedersächsisches Landesamt für Verbraucherschutz und Lebensmittelsicherhetit (LAVES). At the end of the experiments, mice were under anesthesia by cervical dislocation. The experimental procedures performed and the type of anesthesia used were according to university and national guidelines and are explained below.
2.2. Diet and feeding protocol
For the establishment of the 3,5-diethoxycarbonyl-1,4-dihydroxychollidine (DDC)
induced mouse model of sclerosing cholangitis and biliary fibrosis, the mice were fed with 0.1% DDC supplemented diet for a period of 3 weeks65. DDC-supplemented diet
was purchased from Altromin Spezialfutter GmbH, Germany. Age-matched mice fed
with isocaloric diet were used as a control group of all experiments. After beginning
with the DDC diet, the mice were weighed every three days. Initial weight loss was
due to avoidance of the chow, then stabilized or even was followed by a weight gain,
15 but from 10-14 days there was progressive weight loss. The mice, which lost more than 30% from their initial body weights, were sacrificed and excluded from this study. Feeds and bedding were weighed and changed twice per week to provide an estimate of food intake. All the mice were fasted 4 hours before experiments.
2.3. Reagents
Reagent Company
TUDCA(Tauroursodeoxycholic Calbiochem/MerckBiosciences(Germany)
acid)
Mayer's hematoxylin Sigma-Aldrich GmbH (USA)
Eosin Y solution Sigma - Aldrich GmbH (USA)
Xylene J.T. Baker GmbH (Netherlands)
Permount Thermo Fisher Science GmbH (Germany)
Picric acid Sigma-Aldrich GmbH (USA)
Direct Red 80 Sigma -Aldrich GmbH (USA)
Goat Serum Vector Laboratories GmbH (USA)
CK19 antibody DSBH of Iowa university(USA)
EDTA Sigma-Aldrich GmbH (USA)
AEC (red) substrate Kit Life Technologies GmbH(Germany)
Fast Green FCF Thermo Fisher Science GmbH (Germany)
30% H2O2 Merck GmbH (Germany)
PBS Life Technologies GmbH(Germany)
Goat anti-mouse IgG Life Technologies GmbH(Germany)
Tab. 2 List of reagents. This table lists reagents applied during this study with the respective companies from which they were bought. All other chemicals were obtained from Applichem GmbH, Germany, unless mentioned otherwise.
16
2.4. In vivo biliary drainage experiments
Mice were anaesthetized by isoflurane (Forene; Abbott Germany, Wiesbaden,
Germany) under spontaneously breathing. After induction of anesthesia, the mice received tracheal intubation and mechanical ventilation was initiated with an anesthesia unit which was constituted of an isoflurane pump (Univentor 1250
Anaesthesia Unit; AgnTho, Lidingö,Sweden) and a ventilator (MiniVent Type 845;
Hugo Sachs Electronik, March-Hugstetten, Germany). The isoflurane pump supplied mixed narcotic gas (mixture of 10-15% oxygen, 85-90% air, and 2.0 ± 0.2% isoflurane) to the ventilator. Mice∼ were ventilated∼ mechanically at a rate of 120-160/min with a tidal volume of 6-8 ml/kg body weight and kept on a rectal thermistor-controlled heating pad, to maintain the core body temperature between
37-38°C, for the duration of the surgery and the experiment. A catheter was then placed into the left carotid artery and was connected with a blood pressure transducer operating with PowerLab system (AD Instruments, Hastings, UK), for a continuous monitoring of blood pressure and continuous injection. An alkaline solution was infused into carotid artery at a rate of 0.3ml/h to correct the systemic acid–base
+ 2- + balance as following composition: 200mM Na , 100mM CO3 , 0.005mM K and
0.005mM Cl−. The left carotid vein was also intubated for infusion of tauroursode-oxycholate(TUDCA) dissolved in PBS at the rate 600 nmol/min63,64. The
abdomen was opened with a short (2-3 cm) mid-ventral celiotomy and the neck of the
gallbladder was ligated with a suture. The common bile duct was then cannulated with
a polyethylene tubing, for continuous collection of bile. Finally, the abdominal cavity
was closed with a continuous suturing and the animal was allowed to rest for ~20 min,
before start of the experiment.
Mice were killed by cervical dislocation at the end of the collection period. Livers
and blood samples were preserved for histology and laboratory examinations.
During the experiment, blood pressure was continuously monitored and if too low,
either the isoflurane concentration was adjusted, or the infusion speed was increased, 17 or both. Sequential blood samples were taken for blood gas analysis and adjustments were made by increasing the infusion speed, or by giving a solution with less NaCO2
(rarely necessary). The sample for blood gas was taken as described in Chapter 2.5.2, but it also could be done via cannulation of the left carotid artery by a heparinized glass capillary (Clinitubes, Radiometer, Copenhagen, Denmark) during the experiment if necessary. The blood sample was analyzed in a Radiometer blood gas analyzer
(ABL-5 Blood Gas Analyzer, Radiometer, Copenhagen, Denmark) immediately before coagulation.
2.5. Measurement
2.5.1. Measurement of bile flow
After the surgery, the abdominal cavity was closed and the mice were left to stabilize for 20 min before the collection of bile flow. The common bile duct was cannulated and bile samples were collected every 20 min. Baseline values were collected for 40 min, after which the mice were infused with TUDCA via carotid artery for another 60 min.
Bile flow was measured by gravimetric method of fluid measurement. Bile juice was collected in a pre-weighed plastic tube and was weighed again, immediately after the end of the collection period. The difference in measured weight was considered as the bile flow for that collection period. All volume measurements were calculated from weight, with the density of normal saline set arbitrarily to 1.0. The rates of bile flow are expressed as microliter of the base secreted per hour (µl/h).
2.5.2. Measurement of biliary bicarbonate secretion66,67
The rate of luminal alkalinization was determined via back titration of the bile
18
sample to pH 4.5 with 5mM HCl under continuous N2 gassing using pH-stat equipment (PHM82 Standard pH meter, Radiometer, Copenhagen, Denmark). The pH electrode was routinely calibrated with standard buffers before the initiation of the titration. The amount of titrated HCl was considered equivalent to the biliary bicarbonate secretion. The rates of bile alkalinization are expressed as micromoles of the base secreted per hour (µmol/h).
2.6. Handling of tissue and samples
2.6.1. Blood Sample
After in vivo experiments or anesthesia of mice, carotid artery was found and cut
off. Blood samples were collected by a syringe. The fresh blood was centrifuged at a
speed of 100000 times/min and a temperature of 4℃ immediately (Sigma 2-16KL
refrigerated centrifuge, Germany). Serum was pipetted into ice-cold Eppendorf tubes
and immediately frozen, and kept frozen at -80℃ separately after centrifugation for
further analyses.
2.6.2. Liver tissue work-up
We followed the guidelines for standardized work-up for mouse models of the International PSC Study Group (IPSCSG)68.
19
Fig.7 Liver preparation and processing. Liver lobes are labeled continuously according to their size from 1 to 7. A central part should be excised from lobes 1 and 2 that is further fixed in 4% neutral-buffered formaldehyde solution (parts 1a and 2a), embedded in paraffin and processed for histological work-up (i.e., H&E, immunohistochemical staining) and a peripheral part was excised for cryopreservation. Lobe 3 is used for hydroxyproline measurement and lobe 5 for RNA isolation 68.
Liver tissue was collected immediately after the carotid dissection, at the
conclusion of experiment. It was flushed in 1× PBS solution and was separated into
different sections, for different analyses (Fig.7). The whole separating process was
done in 1× PBS solution.
For routine histological stains, tissue was drop-fixed in 5% paraformaldehyde
(PFA) for 24 hours. The excess PFA was removed with PBS rinse; and the sample
was put in 1× PBS at least 12 hours. After paraffin infiltration, the tissue was
embedded into wax blocks which were cut into 2 µm sections.
2.7. Staining protocol
2.7.1. Histology staining
H&E staining and Sirius Red staining were performed according to the suggestions
of the manufacturer’s instructions (Sigma-Aldrich, USA). For H&E staining, sections
were put in xylene for deparaffinization with a slide holder for 2 changes, 10 minutes 20
each. The sections then were dehydrated in 2 changes of absolute alcohol, 5 minutes
each, 95% alcohol for 2 minutes, 70% alcohol for 2 minutes, and were washed briefly
in distilled water at the end. After that, sections were stained in hematoxylin solution for 30 seconds and washed in running tap water for 5 minutes. Then sections were counterstained in eosin solution for 8 seconds, dehydrated with 95% alcohol for 2
minutes, 2 changes of absolute alcohol, 5 minutes each and clear in 2 changes of xylene,
5 minutes each. Finally, slides were mounted with xylene based mounting medium.
For Sirius Red staining, sections were deparaffinized in warming bath at 60˚C for 45
min and incubated in picro-sirius red solution for 60 min, the rest of the staining
procedure was the same as H&E staining.
2.7.2. CK 19 Immunohistochemistry Staining
The CK 19 immunohistochemistry staining was performed based on Ki67
immunohistochemistry staining protocol, described by Vector Labs Company
(Germany) with the following changes. Sections were placed in small metal containers,
containing hot EDTA antigen retrieval buffer. The container was heated in a water bath
for 30 min at 96-98°C, blocked with 5% goat serum in PBST for 60 min, and incubated with anti-cytokeratin 19 (CK19) 1:200 overnight at 4°C. The sections were then
thawed at room temperature for 30 min, incubated with goat anti-rat IgG 1:500 for 60
min, and stained by AEC substrate Kit. Finally, the slides were mounted with medium.
2.8. Statistics
Statistical analyses were performed using the Prism analysis program (Graphpad
6.0, San Diego, CA, USA). The statistical significance of data was tested via repeated
measures analysis of variance. To test differences within a group, one-way ANOVA
was used followed by a Tukey post-hoc test. Between groups, two-way ANOVA was
21
used followed by a Bonferroni post-hoc test. A p-value of less than 0.05 was considered significant. All data are presented as means ± standard error of the means (SEM).
22
3. Results
3.1. Monitoring of vital signs
Blood pressure and temperature of all the animals were continuously monitored
and kept in a physiological range during the course of all experiments. The mean
arterial pressure was maintained between 80-120 mmHg. Apart, blood-gas analysis was done in each experiment. In our surgical setup, the mice were maintained with very good vital values, including the blood-gas analysis (Tab.3). Any deviation from normal was corrected by changing the dose of isoflurane, rate of breathing, amount of
O2 and infusion speed. Animals with abnormal blood-gas values and/or blood pressure
control were excluded from the study.
As previously described, CAXIV probably accounts for a large proportion of the
bicarbonate reabsorption in the kidney 39–41, the deficiency of CAXIV in Car14 KO
mice may theoretically affect the acid-base balance in vivo. All the blood gas data we collected in the in vivo experiments indicated that there was no difference between
Car14 WT mice and KO mice (Tab.3), as well as Mdr2 mice (Tab.4) during the course of the experiment.
23
Car14 DDC-fed
WT (n=18) KO (n=18) Car14 WT(n=7) Car14 KO(n=7)
PH 7.39±0.041 7.38±0.038 7.399±0.015 7.39±0.039
PCO2 35±5.54 31±4.58 36.72±1.68 42.875±5.35
PO2 (mmHg) 177±32.15 169.±22.23 187.7±33.05 155.5±41.90
SO2 (%) 99.2±1.05 99.0±1.03 99.5±0.19 99.67±0.51
- HCO3 (mM) 24.6±2.42 22.7±3.19 25.33±0.65 25.25±2.12
ABE (mM) 0.18±2.40 0.06±2.97 0.83±1.72 0.52±2.26
Tab. 3 Blood gas analysis of Car14 mice and DDC-fed Car14 mice. All the mice were infused with the same carotid perfusion solution for correcting the systemic acidosis caused by anaesthesia. There was no difference between Car14 WT mice and KO mice. Data are presented as means±SD.
WT Mdr2-/- Car14-/-Mdr2-/-
(n=13) (n=13) (n=13)
PH 7.38±0.042 7.39±0.053 7.39±0.0432
PCO 2 41.14±6.91 40.15±6.96 42.42±5.48
PO (mmHg) 2 186.14±56.94 183.5±45.81 184.8±50.25
SO (%) 2 99.4±0.86 99.3±0.75 99.5±0.65
- HCO3 (mM) 24.07±3.262 24.0±4.16 25.21±2.96
ABE (mM) -0.28± 2.78 -0.46±3.97 0.64±3.12
Tab. 4 Blood gas analysis of Mdr2 mice. All the mice were infused with the same carotid perfusion solution for correcting the systemic acidosis caused by anaesthesia. There was no difference between Mdr2 WT mice and KO mice. Data are presented as means±SD.
24
3.2. Loss of carbonic anhydrase 14 results in decreased biliary bicarbonate
output and a slowly progressive cholangiopathy
3.2.1. Bile flow and biliary bicarbonate output rates in Car14-/- and WT mice at
different ages
Basal and TUDCA-stimulated bile flow and biliary bicarbonate output was
measured in 11 weeks-, 20 weeks- and 1 year-old mice. There was a small and
non-significant decrease in basal bile flow and in biliary bicarbonate secretion in
Car14 KO mice as compared to WT mice in baseline at all age groups (Fig.8, 9, 10).
Both biliary bicarbonate output and bile flow increased rapidly after TUDCA infusion
in all different age groups. While basal bile flow did not change with age in WT mice as well as Car14 KO mice, basal biliary bicarbonate output decreased progressively in both genotypes. Car14-/- mice displayed significantly lower bicarbonate output rates
after TUDCA stimulation compared with WT mice in all age groups (Fig.8A, 9A and
10A). In contrast, TUDCA-stimulated bile flow rates decreased with age, and a
significantly lower bile flow rate was observed in Car14-/- compared to WT mice at 1
year of age (Fig.8B, 9B and 10B).
25
Fig. 8 Biliary bicarbonate output and bile flow in 11 weeks old Car14-/- mice. Biliary bicarbonate output increased after TUDCA infusion in both KO and WT mice, but the increase was significantly lower in KO mice than in WT mice (A). Bile flow also increased after being stimulated by TUDCA. However there is no difference in bile flow between WT mice and KO mice (B). # Significant increase as compared to basal within the same group; * Significant difference as compared to the Car14 WT mice. Data are presented as means±SEM, n=8 in each group (* ,# p<0.05; **, ## p<0.001; ***, ### p< 0.0001).
26
Fig. 9 Biliary bicarbonate output and bile flow in 20 weeks old Car14-/- mice. In 20 weeks old mice, the bicarbonate output of KO mice was significantly lower than WT mice after TUDCA stimulation (A), but lower bicarbonate output did not affect the bile flow (B). # Significant increase as compared to basal within the same group; * Significant difference as compared to the Car14 WT mice. Data are presented as means±SEM, n=5 in each group(* ,# p<0.05; **, ## p<0.001; ***, ### p< 0.0001).
27
Fig. 10 Biliary bicarbonate output and bile flow in 1 year old Car14-/- mice. In 1 year old mice, the difference in bicarbonate output was seen only after 40 mins of TUDCA perfusion between WT mice and KO mice. But bile flow of KO mice is significantly lower than WT mice. # Significant increase as compared to basal within the same group; * Significant difference as compared to the Car14 WT mice. Data are presented as means±SEM, n=5 in each group(* ,# p<0.05; **, ## p<0.001; ***, ### p< 0.0001).
28
3.2.2. Histological analysis
No apparent difference between WT and Car14 KO mice was observed in the histological and immunohistological evaluation at 11 weeks and 20 weeks of age. In the 1 year old Car14 KO mice, pathological changes suggesting fibrosis and small bile duct cholangiopathy were present in the peripheral parts of the liver. Proliferation of small bile ducts (Fig.14F), with some mild fibrosis (Fig.16F), was evident in those segments, without any apparent changes in the central liver regions.
29
Fig. 11 Liver Hematoxylin and eosin(H&E)staining in central part of Car14-/- mice at different ages. There was no apparent difference in portal area between control mice[11weeks(A), 20 weeks(B), 1 year(C)], 11 weeks old Car14-/- mice(D), 20 weeks old Car14-/- mice(E) and 1 year old Car14-/- mice(F). PV, portal vein; PA, portal artery; BD, bile duct
30
Fig. 12 Liver Hematoxylin and eosin(H&E)staining in peripheral part of Car14-/- mice at different ages. There was no apparent difference in portal area between control mice[11weeks(A), 20 weeks(B), 1 year(C)], 11 weeks old Car14-/- mice(D), 20 weeks old Car14-/- mice(E) and 1 year old Car14-/- mice(F). PV, portal vein; PA, portal artery; BD, bile duct
31
Fig.13 Liver CK19 immunohistochemistry staining in central part of Car14-/- mice at different ages. There was no apparent difference between control (WT) mice [11 weeks (A), 20 weeks (B) and 1 year (C)], 11 weeks old Car14-/- mice (D) and 20 weeks old Car14-/- mice (E) and 1 year old Car14-/- mice (F).
32
Fig. 14 Liver CK19 immunohistochemistry staining in peripheral part of Car14-/- mice at different ages. There was no apparent difference between control (WT) mice [11 weeks (A), 20 weeks (B) and 1 year (C)], 11 weeks old Car14-/- mice (D) and 20 weeks old Car14-/- mice (E). In 1 year old Car14-/- mice (F) proliferation of small primary bile ducts could be seen in the peripheral parts of liver, with no apparent proliferative changes in the central regions.
33
Fig.15 Liver Sirius red staining in central part of Car14-/- mice at different ages. There was no apparent pathological change in control (WT) mice [11weeks (A), 20 weeks (B) and 1 year (C)], 11 weeks old Car14-/- mice (D) and 20 weeks old Car14-/- mice (E) and 1 year old Car14-/- mice (F).
34
Fig. 16 Liver Sirius red staining in central part of Car14-/- mice at different ages. There was no apparent pathological change in control (WT) mice [11weeks (A), 20 weeks (B) and 1 year (C)], 11 weeks old Car14-/- mice (D) and 20 weeks old Car14-/- mice (E). However, in 1 year old Car14-/- mice (F), some fibrosis could be seen in the peripheral parts of liver, with no apparent differences in the central regions (yellow arrows).
35
3.3. Car14 expression cannot protect against the severe toxic cholangiopathy
induced by DDC-feeding
3.3.1. DDC cholangiopathy model development
In order to investigate whether the absence of CAXIV expression aggravates the bile duct injury induced by DDC (3,5-diethoxycarbonyl-1,4-dihydroxychollidine), we first established the DDC-induced cholangiopathy model in our laboratory.
C57BL/6N mice were fed with 0.1% DDC supplemented diet for 3 weeks. In comparison to the mice fed with normal mouse chow, the DDC-fed mice gradually lost weight, became lethargic, the skin in fur-free areas and the abdominal fat pad appeared icteric. Both the bile flow rate and the biliary bicarbonate output rate was significantly lower in the DDC-fed than the control group (Fig.17).
36
Fig. 17 Biliary bicarbonate output and bile flow in 0.1% DDC-fed C57BL/6N mice. After TUDCA infusion, increase of biliary bicarbonate output and bile flow was significantly lower in the DDC-fed mice compared with chow-fed control mice. # Significant increase as compared to basal within the same group; * Significant difference as compared to the control mice. Data are presented as means±SEM, n=5 in control group, n=8 in DDC-fed group (* ,# p<0.05; **, ## p<0.001; ***, ### p<0.0001).
37
3.3.2. DDC-induced decrease in biliary secretion in Car14 WT and KO mice
In order to investigate whether the loss of Car14 and the resulting decrease in
biliary bicarbonate output aggravates DDC-induced cholangiopathy, Car14 WT and
KO mice, aged 20 weeks, were fed with 0.1% DDC for 3 weeks, after which bile flow and biliary output rates were determined. The results showed that biliary bicarbonate output in the DDC-fed WT mice was, as expected, significantly lower than the chow-fed control mice, and DDC-fed Car14 KO mice had a reduced biliary bicarbonate output compared to DDC-fed WT mice (Fig.18A). However, the difference of bicarbonate output between DDC-fed WT mice and KO mice was not accompanied by a difference in bile flow rates between DDC-fed Car14 KO and WT mice, which were equally reduced compared to control chow-fed mice (Fig.18B).
38
Fig. 18 Biliary bicarbonate output and bile flow in DDC-fed Car14 mice. After three weeks of DDC feeding, we tested biliary bicarbonate output and bile flow in vivo. (A): DDC caused significant − − reduction of HCO3 secretion both in CAXIV WT mice and KO mice even in basal condition. But HCO3 secretion after TUDCA stimulation in CAXIV KO mice is still lower than WT mice. (B): DDC-fed CAXIV WT and KO mice also demonstrate lower bile flow in comparison with the control mice. However there is no significant difference between DDC-fed CAXIV WT and KO mice. # Significant increase as compared to basal within the same group; * Significant difference as compared to the control mice;
▲Significant difference as compared to the DDC-fed CAXIV WT VS. the DDC-fed CAXIV KO. Data are presented as means±SEM, n=5 in control group, n=7 in DDC-fed groups (* , #, ▲p<0.05; **, ##, ▲▲ p<0.001; ***, ###, ▲▲▲ p<0.0001).
39
3.3.3. Morphological and histological evaluation of the DDC-fed liver of Car14
KO and WT mice
Morphologically, the liver in mice fed with 0.1% DDC (for 3 weeks) appeared appreciably darker in color, harder in consistency and larger, as compared to the control mice receiving normal diet (Fig.19).
Fig. 19 Appearance of normal liver and DDC fed liver. DDC fed liver (A) appeared darker color and harder texture than normal liver (B), and some of the DDC fed liver had larger volume than normal.
After 3 weeks of DDC feeding, livers showed ductular reaction, ductules and
small bile ducts proliferation and wide range of fibrosis (Fig.20, 21 and 22). In addition, these ductules frequently contained porphyrin plugs (Fig.17B and C). No
obvious histological difference was observed between the DDC-fed Car14 WT and KO
mice.
40
Fig. 20 Liver H&E staining of DDC-fed mice. In DDC-fed Car14 WT mice (B) and Car14 KO mice(C) reveals increasing ductular proliferation(black arrow) and periductal fibrosis as compared to control mice (A). Porphyrin plugs (yellow arrow) occluding the luminal side of the small bile ducts is observed (B, C). PV, portal vein; PA, portal artery; BD, bile duct.
41
Fig. 21 Liver CK19 immunohistochemistry staining of DDC-fed mice. DDC caused severe periductal proliferation in DDC-fed mice. There was no obvious difference between DDC-fed Car14 WT mice(B) and Car KO mice(C). PV, portal vein.
42
Fig. 22 Liver Sirius red staining of DDC-fed mice. Compare to the control mice (A), strong fibrotic changes were observed in both DDC-fed Car14 WT (B) mice and KO mice (C) (red staining). These changes appeared somewhat more pronounced to the eye in the Car14 KO mice in this liver, but such differences were not consistently seen between the two genotypes.
43
3.4. Loss of carbonic anhydrase XIV aggravates the bile-acid induced
cholangiopathy of the Mdr2-/- mouse model
The loss of the lecithin exporter Mdr2 results in an alteration of the transport of
bile acid, cholesterol and phospholipids in so called mixed micelles and “liquid
crystals” and overall decreased bile flow. The resulting cholangiopathy that develops
in Mdr2-/- mice is believed to be due to the toxic action of bile acid to the
cholangiocytes. Thus, this model seems optimally suited to test the “bicarbonate
umbrella” hypothesis by generating CAXIV/Mdr2 double knockout mice.
We chose to cross B6.129S1-Car14 mice with FVB/N-Abcb4 mice, because the loss of MDR2 causes a more rapidly progressing fibrosing cholangiopathy in the
C57B/6 background than the FVB/N background, with early cholangiocarcinoma development, which was not our goal. From the same progenitors, we bred the four different genotypes: Car14+/+Mdr2+/+, Car14+/+Mdr2-/-, Car14-/-Mdr2+/+ and
Car14-/-Mdr2-/-, and studied bicarbonate output and bile flow rates in these four
genotypes at the age of 6 weeks and 11 weeks.
3.4.1. Result of the in vivo experiments
Bicarbonate output was measured in Car14+/+Mdr2+/+, Car14+/+Mdr2-/-,
Car14-/-Mdr2+/+ and Car14-/-Mdr2-/- mice at age of 6 weeks. Compare to basal rates,
− after TUDCA stimulation, HCO3 output rate increased significantly slower and to lesser values in the absence of CAXIV, both in Car14-/- mice, regardless of the defectiveness of Mdr2 gene. However in Car14+/+Mdr2-/- mice, bicarbonate output was
surprisingly higher than the other genotypes (Fig.23). Only Car14-/-Mdr2-/- mice has
significantly lower of bile flow increase after the stimulation of TUDCA, the other
genotypes had almost the similar volume of bile flow secretion (Fig.24).
44
Fig. 23 Biliary bicarbonate output in Car14/Mdr2 mice at age of 6 weeks. Bicarbonate output was measured in four different genotype combinations of Car14 and Mdr2. Under TUDCA -/- -/- − stimulation, only Car14 Mdr2 mice had significantly less increase of HCO3 output compared to − -/- -/- other genotypes (A). HCO3 output was significantly reduced in CAXIV defective mice, Car14 Mdr2
45 mice were even significantly lower than Car14-/-Mdr2+/+ mice (B), but surprisingly higher in Car14+/+Mdr2-/- mice than other genotypes (C). # Significant increase as compared to basal within the same group; * Significant difference as compared to the control mice; ▼Significant difference as compared to the Car14-/-Mdr2-/- mice VS. the Car14-/-Mdr2+/+ mice; ▲Significant difference as compared to the Car14-/-Mdr2-/- mice VS. the Car14+/+Mdr2-/- mice. Data are presented as means±SEM, n=6-7 in each group (* , #, ▼, ▲p<0.05; **, ##, ▼▼, ▲▲ p< 0.001; ***, ###, ▼▼▼, ▲▲▲ p<0.0001).
Fig.24 Bile flow in Car14/Mdr2 mice at age of 6 weeks. Only Car14-/-Mdr2-/- mice has significantly lower of bile flow increase after the stimulation of TUDCA, other genotypes had almost the same volume of bile flow secretion. # Significant increase as compared to basal within the same group; * Significant difference as compared to the double KO mice VS. the other groups. Data are presented as means±SEM, n=6-7 in each group (* ,# p<0.05; **, ## p<0.001; ***, ### p <0.0001).
46
At the age of 11 weeks, bicarbonate output increase was significantly less in
Car14+/+Mdr2-/-, Car14-/-Mdr2+/+ and Car14-/-Mdr2-/- mice compared to the control
mice after TUDCA stimulation. Car14-/-Mdr2-/- group had even less increase than
Car14-/-Mdr2+/+ group (Fig.25). Increase of bile flow in Car14-/-Mdr2-/- group was
significantly less in comparison with the other groups. Car14+/+Mdr2-/- group also had
lower bile flow secretion than Car14+/+Mdr2+/+ group (Fig.26).
If we put the data of 6 weeks-old mice and 11 weeks-old mice together, we can find some interesting changes. Bicarbonate output after TUDCA stimulation in
Car14+/+Mdr2-/- mice decreased from significantly higher than control group at 6
weeks of age to significantly lower than control group at 11 weeks of age, so was the
drop of bile flow.
47
Fig. 25 Biliary bicarbonate output in Car14/Mdr2 mice at age of 11 weeks. The bicarbonate output of Car14-/-Mdr2-/- group has no response to the TUDCA stimulation (A). The stimulated bicarbonate output were significantly lower in Car14+/+Mdr2-/- and Car14-/-Mdr2+/+ groups than in WT group. It was also lower in Car14-/-Mdr2-/- group than in Car14-/-Mdr2+/+ group (B-C).
48
# Significant increase as compared to basal within the same group; * Significant difference as compared to the control mice; ▼Significant difference as compared to the Car14-/-Mdr2-/- mice VS. the Car14-/-Mdr2+/+ mice. Data are presented as means±SEM, n=7 in each group (* , #, ▼ p<0.05; **, ##, ▼▼ p<0.001; ***, ###, ▼▼▼ p<0.0001).
Fig. 26 Bile flow in Car14/Mdr2 mice at age of 11 weeks. Bile flow of Car14-/-Mdr2-/- group was significantly reduced compared with the other groups. Car14+/+Mdr2-/-group also had lower bile flow secretion than Car14+/+Mdr2+/+ group. # Significant increase as compared to basal within the same group; * Significant difference as compared to the control mice; ▲Significant difference as compared to the Car14-/-Mdr2-/- mice VS. the Car14+/+Mdr2-/- mice. Data are presented as means±SEM, n=7 in each group (* , #, ▲p<0.05; **, ##, ▲▲ p<0.001; ***, ###, ▲▲▲ p<0.0001).
3.4.2. Result of histological analysis
In all age groups, Car14-/-Mdr2+/+ mice had exactly the same histological phenotype function in comparison to Car14+/+Mdr2+/+ which were consistent with our previous data. Therefore those histological images are not shown here.
In 3 weeks-old Car14/Mdr2 mice, all the groups maintained normal in histology.
Only Car14-/-Mdr2-/- mice showed mild incrassation of periductal tissue and infiltration of inflammatory cells in the central part of the liver (Fig. 27C).
49
Fig. 27 Liver H&E staining of Car14/Mdr2 hybrid mice at age of 3 weeks. Histology of Car14+/+Mdr2+/+ mice (A) and Car14+/+Mdr2-/- mice(B) maintained normal. Car14-/-Mdr2-/- mice (C) revealed incrassated periductal tissue and infiltration of inflammatory cells. In peripheral area, there was no difference between Car14+/+Mdr2+/+ mice (D) and Car14+/+Mdr2-/- mice (E) and Car14-/-Mdr2-/- mice (F). PV, portal vein; PA, portal artery; BD, bile duct.
50
Fig. 28 Liver CK19 immunohistochemistry staining of Car14/Mdr2 hybrid mice at age of 3 weeks. Increasing CK19 positive cholangiocytes or portal fields were not observed in any of the genotypes Car14+/+Mdr2+/+ mice [central part (A), peripheral part (D)] and Car14+/+Mdr2-/- mice [central part (B), peripheral part (E)] and Car14-/-Mdr2-/- mice [central part (A), peripheral part (D)] at this age. PV, portal vein; PA, portal artery; BD, bile duct.
51
Fig. 29 Liver Sirius red staining of Car14/Mdr2 hybrid mice at age of 3 weeks. Fibrosis did not develop in 3 weeks-old mice. All groups: Car14+/+Mdr2+/+ mice [central part (A), peripheral part (D)] and Car14+/+Mdr2-/- mice [central part (B), peripheral part (E)] and Car14-/-Mdr2-/- mice[central part (C), peripheral part (F)] maintained normal.
52
Age of 6 weeks is a very crucial time point to Car14/Mdr2 mice according to our
results. In contrast to the 3 weeks of age Mdr2-/- mice, Mdr2-/- mice at 6 weeks of age
showed significant differences in histological examination. Interestingly,
Car14+/+Mdr2-/- mice and Car14-/-Mdr2-/- mice also showed different pathological
degree. The inflammation, ductular proliferation and fibrosis of bile duct in
Car14+/+Mdr2-/- mice were limited in periductal area (Fig.30B, 31B and 32B).
Car14-/-Mdr2-/- mice not only had severe pathological changes in periductal area, but
also extended to portal-to-portal bridge (Fig.30C-D, 32C).
53
Fig. 30 liver H&E staining of Car14/Mdr2 hybrid mice at age of 6 weeks. The Mdr2-/- mice developed ductular proliferation and cholangitis with onion skin type-like periductal fibrosis in the liver as compare to Car14+/+Mdr2+/+ mice (A). Car14-/-Mdr2-/- mice (C) exhibited more severe pathological changes in comparison with Car14+/+Mdr2-/- mice (B). The inflammation already spread to portal-to-portal bridge in Car14-/-Mdr2-/- mice(D). PV, portal vein; PA, portal artery; BD, bile duct.
54
Fig. 31 Liver CK19 immunohistochemistry staining of Car14/Mdr2 hybrid mice at age of 6 weeks. Increasing CK19 positive cholangiocytes and portal fields were observed in Mdr2-/- mice at this age. But the difference between Car14+/+Mdr2-/- mice (B) and Car14-/-Mdr2-/- mice (C) was apparent. PV, portal vein; PA, portal artery; BD, bile duct.
55
Fig. 32 Liver Sirius red staining of Car14/Mdr2 hybrid mice at age of 6 weeks. In contrast to Car14+/+Mdr2+/+ mice (A), fibrosis in the liver of Car14+/+Mdr2-/- mice (B) was limited in periductal tissue, but Car14-/-Mdr2-/- mice (C) already developed mild fibrosis at portal-to-portal bridge (yellow line).
56
In 11 weeks-old groups, characteristic pathological changes of cholangitis became much more obvious in Car14+/+Mdr2-/- mice and Car14-/-Mdr2-/- mice. Severe periductal fibrosis and inflammation, wide range of ductular proliferation and fibrosis of portal-to-portal bridge can be found in both of these two genotypes. The progression of pathology in Car14-/-Mdr2-/- mice is still faster than Car14+/+Mdr2-/- mice, but the gap is narrowing.
Fig. 33 Liver H&E staining of Car14/Mdr2 hybrid mice at age of 11 weeks: Car14+/+Mdr2-/- mice(B) and Car14-/-Mdr2-/- mice (C) exhibited more severe ductular proliferation(yellow arrow) and cholangitis with onion skin type-like periductal fibrosis in liver as compare to Car14+/+Mdr2+/+ mice (A). Atrophy and death of BECs (red arrow) and more severe inflammation area (red dotted line) can be found in Car14-/-Mdr2-/- mice (D). PV, portal vein; PA, portal artery; BD, bile duct.
57
Fig. 34 Liver CK19 immunohistochemistry staining of Car14/Mdr2 hybrid mice at age of 11 weeks. In contrast to control mice (A), there is a significantly increased number of CK19-postive cholangiocytes and portal fields in Car14+/+Mdr2-/- mice (B) and Car14-/-Mdr2-/- mice (C) were apparent. PV, portal vein; PA, portal artery; BD, bile duct.
58
Fig.35 Liver Sirius red staining of Car14/Mdr2 hybrid mice at age of 11 weeks. Both of Car14+/+Mdr2-/- mice (B) and Car14-/-Mdr2-/- mice(C) developed fibrosis of portal-to-portal bridge as compare to control liver(A). However Car14-/-Mdr2-/- mice(C) had more extensive deeper and the thicker fibrosis staining than Car14+/+Mdr2-/- mice (B).
59
4. Discussion
4.1. Improvement of experimental design
4.1.1. Choosing the appropriate model to study the importance of biliary
bicarbonate output on cholangiocytes health in vivo (the so-called “biliary
bicarbonate umbrella” hypothesis)
It is known for a long time that a subgroup of patients with cystic fibrosis
develops liver fibrosis, cholangiopathy and in the worst case end-stage liver
cirrhosis69. The etiology is not completely understood. When the importance of the
CFTR channel for bicarbonate secretion became known, it was speculated that the
cholangiopathy that develops in cystic fibrosis patients and, under certain
circumstances, in CF mouse models, is due to the lack of bicarbonate in the bile ducts.
However, the absence of the CFTR channel results in a severe compromise of epithelial
Cl- and fluid secretion as well, and it is not possible to distinguish these effects from
each other when trying to determine the cause of pathological changes in the bile
ducts70. In vitro work in cholangiocytes cell culture substantiated the concept that this
epithelium is protected against the potential toxicity of bile acids by an alkaline
pH-microclimate at its apical membrane13,71. However, an in vivo unequivocal confirmation of the hypothesis is lacking. One publication has provided circumstantial evidence by stimulating bile salt synthesis through stimulation of the transcription factor FXR-α, which results in an increase in bile flow as well as bicarbonate output72.
We chose the Car14-/- mouse model, because we had observed in the stomach and
the duodenum that these enzymes conveyed a protective effect against the damaging
effects of luminal acid. However, the molecular mechanism of how this protection was
mediated was dependent on the location of the carbonic anhydrase. Expression analysis
of CAXIV in the different organs revealed a high expression in the liver40.
Immunohistochemical investigations demonstrated that the predominant expression of
60
the CAXIV was in the canalicular membrane of hepatocytes and cholangiocytes45. As
explained in the introduction, the carbonic anhydrase XIV is believed to be an
exoenzyme, with the catalytic part of the molecule exposed to the lumen. Any carbonic
- + anhydrase will enhance the conversion of HCO3 and H to CO2 and water, and vice
versa. The intracellular CAII, for example enhances the dissipation of pH gradients in
cardiomyocytes and may be important in energizing cardiomyocytes and possibly
preventing arrhythmogenic effects of the low pH during hypoxia73. In the duodenocyte,
it is essential for the signal of “acid in the lumen” to traverse the duodenocyte and
stimulate an increase in bicarbonate secretory rate into the lumen74. A luminally
situated carbonic anhydrase in the hepatic canaliculi and in the bile ducts could enhance
the “buffering” of bile acid monomers, which are present at much higher concentration
at low pH, and shift the balance to the micellar transport form of the bile acids. The
presence of the luminal carbonic anhydrase would not be expected to increase luminal
alkalinity directly. However, it may or may not enhance biliary bicarbonate secretion, if
the generated CO2 diffuses back into the hepatocyte and enhances bicarbonate-transporting ion transport processes. This has been shown for the
75 duodenum, for example . But it is also possible that the back-diffusing CO2 stimulates fluid absorption, via lowering the intracellular pH and thus stimulating the apical
Na+/H+ exchanger NHE3. This process has been demonstrated in the jejunum or the
colon76,77, and NHE3 has been shown to be expressed in biliary epithelial cells34. Thus,
it was not clear to us what effect the absence of CAXIV has on biliary fluid transport,
and we decided to study this mouse model in the hope to learn more about the
physiological function of CAXIV in the liver.
In order to do so, we applied our experience gathered over many years in studying
fluid, electrolyte and bicarbonate transport in murine intestine in vivo to investigate
biliary secretion. We studied the literature, but decided to optimize currently used in
vivo techniques to assess murine biliary secretion.
61
4.1.2. Hematocrit determination and continuous fluid resuscitation during the
biliary drainage period
Abdominal operations in mice are accompanied by evaporation of fluid, even if the wound is closed after the surgery. In addition, the mice loose fluid via the airways. In order to prevent hemoconcentration or hemodilution, if only the arterial blood pressure is taken as a parameter to guide fluid replacement, we performed sequential hematocrite measurements with very small glass capillaries, and kept the hematocrit stable during the experiment by placing both an arterial line into the carotis, for fluid and base replacement, and a venous line into the femoral vein for application of
TUDCA.
4.1.3. Advantage of direct collection of bile from the hepatic duct
Collection of bile is mostly performed by cannulation of the gallbladder64,78.
However, the major function of gallbladder is to concentrate and store bile, and its epithelium expresses, among other transporters, the Na+/H+ exchanger isoform 3 at very high levels. This transporter will not only rapidly absorb electrolytes and fluid and thus modify measured bile flow rates, but also transport hydrogen ions into the lumen and thus alter biliary bicarbonate concentrations in a major fashion. Therefore the measurement of external drainage from gallbladder could be imprecise.
We resolved this issue by improving our surgical process. We made a ligation at neck of gallbladder which is completely blocked the way of bile going ahead to gallbladder. Then we intubated the common bile duct under microscopy for external drainage of bile that the bile flow and bicarbonate secretion of live had a faithful reflection. The difficulty of this way lies in the fact that the common bile duct is very thin and it’s hiding in the hepatoduodenal ligament, thus the operator need to be familiar with the anatomy of mice and experienced at microsurgery. However, the technique allowed us to precisely assess basal and stimulated biliary fluid and
62
bicarbonate output rates.
4.1.4. The pH-stat titration technique is optimal for exact assessment of bicarbonate
output
Bicarbonate is commonly assessed by a CO2 electrode and by calculating the bicarbonate output with the Henderson Hasselbalch equation. With the pH-stat titration technique, the gaseous CO2 (which does not act as an alkaline buffer) is eliminated at a
pH set to of 7.4, and only the bicarbonate in the bile is assessed as bicarbonate.
4.1.5. Control of the systemic acid/base status during the in vivo biliary drainage
Another important experience we learned from human clinic practices is the
control of blood gas during the experiment. The only demand of most animal
experiment under anesthesia is to keep the animal alive, or to keep the blood pressure
stable during the experiment at most. We found that mice under anesthesia regularly
develop metabolic acidosis related to some metabolic effect of the isoflurane (which,
for example, decreases insulin secretion79), and they also develop either respiratory
alkalosis because of hyperventilation, if anesthetized lightly, or acidosis, if anesthetized deeply, which causes CO2 retention. This is not enough for our aim. It has been proved
+ decades ago that CO2 in plasma can rapidly cross cell membranes, dissociate into H
− 80 − and HCO3 with H2O . This is the major source of the intracellular HCO3 of
hepatocytes and cholangiocytes which can be transported into bile ductal lumen. The
disorder of acid-base in blood will certainly affect the bicarbonate secretion in all
organs. The crucial part of our experiment is to measure and analyze the bicarbonate
secretion in the bile, therefore control the blood gas in a physiological range seems
essential. All the mice in our vivo experiments were anaesthetized by isoflurane under
mechanical ventilation. A fixed setting of the anesthesia machine and the ventilator
cannot cover all the individual difference of these mice if we want a physiological
blood gas. We controlled blood gas by adjusting the depth of anesthesia, the tidal 63
volume and stock speed of ventilator, also the infusion speed of alkaline solution. These
changes were made based on the analysis of arterial blood gas during and after the
experiment. If mice did not have blood gases within a physiological “near normal”
range, the results were abandoned.
4.1.6. Being able to get at the “liver function reserve” in order to assess early biliary
functional defects
In our vivo experiments, we measured both biliary bicarbonate output and bile flow.
The reduction of bile flow is per definition the symptom of “cholestasis”, which can be
caused by a primary disturbance in bile formation81, or by biliary obstruction caused by
blockage of small or large bile ducts, which can be secondary to defects in the bile flow .
If we try to find out the relation between the changes of bicarbonate output in bile and
the injury of the bile ducts, bile flow rate is a very important parameter which can only be tested in vivo. The use of different gene-deleted mouse models demonstrate that the reduction of bicarbonate output is not necessarily linked to a decrease in bile flow, and we can thus study a reduction of bicarbonate output as an independent parameter, which has never been done before. TUDCA has been shown to stimulate bile flow and biliary bicarbonate secretion, although the mechanisms may be multifactorial21,82,83. By studying both spontaneous as well as bile-salt stimulated biliary secretion, we may assess even early changes biliary secretory function that is associated with no obvious alteration of the mouse behavior or weight development.
We know that the liver has powerful functional reserve which based on two of its properties. First of all, liver is an organ with high regenerative capacity84–86. In
mammals, animals can survive from a resection up to 75% of the total liver weight.
Within 1 week after surgery, the total volume of liver cells is restored87,88. And also, the
liver is capable of full function to support the whole body with less than half of its
normal hepatocyte and cholangiocytes content8. It suggests that cholangiocytes and
64 hepatocytes don’t devote all their efforts or/and just part of them are involved into daily task in a physiological situation. That also explained why some patients with a partial hepatectomy can have a completely normal liver function even from the first day after operation. The regeneration of liver is very fast but not that fast. For the same reason many patients still have functional compensation even in advanced liver disease.
When investigating the liver damage of Car14-/- mice, DDC-fed mice or Mdr2-/- mice, alterations in the expression of pro- and anti-inflammatory genes and profibrogenic genes are observed early (Zhenzhen Zhou, unpublished observations), while the mice are developing normally without any obvious symptoms of disease.
They gain weight and are fertile. Even the basal bile flow and bicarbonate output rates are normal. However, maximally stimulated flow rates and bicarbonate output rates are compromised. TUDCA is a good option to stimulate the liver to release the rest of reserve that we can take a panoramic view to its full capacity of bile flow and bicarbonate secretion in vivo. That is why in our results, normally there were no differences between control and experiment group in baseline but significant differences after TUDCA stimulation. Some previous studies, in which only the spontaneous bile flow and bicarbonate output rates were considered, concluded that
DDC-fed and Mdr2-deficient mice do not exhibit a reduction of bicarbonate output and bile flow50,56. Now we realize that this is true only for spontaneous secretory rate, which is not the physiologically important postprandial rate during which the bile salts taken up via the enterohepatic circulation stimulate bile flow to replenish the gallbladder bile content.
4.2. Role of the CAXIV in biliary physiology
We studied the Car14-/- mice and WT littermates at different ages: 11 weeks (young adulthood), 20 weeks (full maturity) and 1 year (late adulthood for laboratory mice, which usually die before 2 years of age). We observed a difference in basal bicarbonate
65
output in young mice, which had the highest bicarbonate output rates, but this did not
obtain significance when a two way ANOVA analysis was used for statistical
significance evaluation between all groups (both basal and stimulated). However, a
significant difference of biliary bicarbonate output after TUDCA stimulation was
observed in all age groups between CAXIV WT mice and KO mice, which indicated
that the loss of CAXIV will cause a decrease in the bicarbonate output into bile (Fig.8A,
9A, 10A). At week 11 and 20, the decrease in biliary bicarbonate output was not associated with a decrease of basal or stimulated bile flow. How can this be explained?
One possibility is that bile flow in mice is largely independent of
bicarbonate-transporting channels like CFTR, which would be linked to Cl- and fluid
secretion. Biliary bicarbonate is assumed to be regulated by cholangiocytes rather than
hepatocytes in normal rat liver. It appears that different anion transporters are involved,
since Ach-induced bicarbonate excretion depends on both chloride channels and
bicarbonate exchange, whereas secretin-induced bicarbonate excretion is independent
of bicarbonate exchange17, but a role of the CFTR channel has been postulated by
others89. It has been shown that ursodeoxycholic acid-stimulated choleresis is
independent of CFTR expression90, while secretin-stimulated choleresis is claimed to
be CFTR-dependent by some but not all researchers. It has also been shown in cultured
cholangiocytes that ursodeoxycholic acid stimulates cholangiocytes fluid secretion in
mice via CFTR-dependent ATP secretion. Thus, while both bile flow and biliary
bicarbonate output may depend on anion channels (for example CFTR, but not
exclusively so), the fact that CAXIV deficiency results in a significant decrease in
bicarbonate output but not bile flow rate argues against a channel-dependent
mechanism whose transport rate is augmented by CAXIV activity.
The recycling of CO2, which is bound to occur when bile acids are neutralized by
- biliary HCO3 , may result in an increase in the membrane association and/or functional
activity of another carbonic anhydrase, such as CA II, at the inner leaflet of the canalicular membrane, resulting in enhanced intracellular CO2 hydration, with
66
basolateral H+ extrusion and apical exchange of bicarbonate for chloride, either by AE2
(if that is truly apically expressed in the liver), by an as yet not identified member of the
Slc26 family, or by one or several other anion exchanger(s). This would result in an exchange of bicarbonate for chloride ions with no obligatory change in water flow.
Another mechanism how the presents of an extracellular carbonic anhydrase may augment the activity of an anion exchanger is by facilitation of the dissipation of nanoscale pH-gradients by increasing the conversion rate of the slowly diffusible protons and base ions to the rapidly diffusion CO2.
Another possibility is that CAXIV is involved in hepatocyte intracellular pHi
regulation, as suggested for the nervous system43,91 and the heart92 by facilitating the
operation of AE3, another member of the Slc4 anion transporter gene family. An
effect on intracellular pHi is more likely when there is a basolateral (as well as apical)
expression of the CAXIV, as described by Parkkila45. In that case, the carbonic
anhydrase activity would enhance the removal of protons from the outer leaflet of the
hepatocyte membrane by buffering with blood borne bicarbonate. This function has
been shown for the basolateral CAXIV in the stomach. It would also result in an
increase in intracellular pHi and more bicarbonate for output via the canalicular
membrane.
Although the bicarbonate output was reduced in 11 and 20 week old Car14-/- mice,
suggesting that their biliary protection by luminal bicarbonate is reduced, the mice did
not show any signs of histological damage to hepatocytes or bile ducts. At 1 year of age,
the mice displayed a reduction of basal bicarbonate output and less response to TUDCA
even in the WT mice, although basal bile flow was not different from young mice.
However, the fluid secretory response to TUDCA was more sluggish. Interestingly, in
that age group we observed a difference not only in bicarbonate output but in bile flow
rate between Car14-/- mice and WT littermates. Since biliary fluid secretion is normal in
Car14-/- mice at young and middle age, this reduction is not due to the lack of this gene
in the molecular mechanisms of biliary fluid transport. The surprise came with the
67
histochemical staining for Cytokeratin 19 (CK19), which demonstrated enhanced
CK19 expression in the walls of the small bile ducts in the peripheral parts of the liver, while the central parts of the lob did not appear significantly altered. Likewise, fibrotic changes were observed in the peripheral parts of the Car14-/- liver. This is indicative of
a slowly progressing small bile duct cholangiopathy in the absence of CAXIV
expression. We therefore assume that the reduction in TUDCA-stimulated biliary fluid
secretion is secondary to the development of biliary duct wall thickening and sclerosis,
and peripheral liver fibrosis leading to an overall reduction in active biliary units.
4.3. Using the DDC-fed mice as a pathological model to study the protective
effects of a higher bicarbonate output into bile in the presence of CAXIV
expression
We next wondered whether hepatocyte/cholangiocyte CAXIV expression is more
important during injury. We searched for mouse models of sclerosing cholangiopathy
and first tested the frequently utilized model in which the mice are fed with DDC. As
stated in the discussion, DDC is a cytotoxic molecule that leads to a massive
inflammatory reaction of the biliary epithelial cells which release proinflammatory and
profibrogenetic cytokines and adhesion molecules, such as TNF-α, VCAM, PDGF,
TGF-β. These cytokines and molecules cause periductal inflammation51,52,93,94. Then
the portal-to-portal bridge fibrosis appears as a ductal reaction after the injury of Bile
epithelial cells (BECs) getting worse. Inflammation of periductal induces promotes
fibrosis, increases proliferation of fibroblast and fibrocyte, and deposition of
extracellular matrix (ECM). The mechanism indicates that DDC-induced sclerosing cholangitis is an inflammation-induced disease. Using the standard feeding protocol, the mice developed liver disease within three weeks. Functionally, the DDC-fed
C57BL/6N mice displayed both lower bicarbonate output, as well as bile flow than chow-fed mice. We also observed onion skin type-like periductal fibrosis and 68
proliferation and cholangitis in DDC-fed liver. It’s reasonable to conclude that DDC induced severe toxic injury to the biliary system injury which results in including inflammation of periductal area, ductular proliferation and porphyrin plugs, as
demonstrated in the histological assessment.
We next subjected both Car14-/- and WT littermates to the DDC-feeding experiment, while carefully monitoring them for weight loss and signs of severe disease, upon which the experiment would have to be terminated prematurely. The mice developed rapid weight loss in the first few days, which was due to food avoidance. The weight then stabilized or even increased, but then the mice lost weight and became listless at about three weeks of DDC feeding, which is when the experiment was terminated. Chow-fed control mice had significant higher bicarbonate secretion and bile flow than DDC-fed CAXIV WT mice and KO mice. But we noticed that although
DDC-fed CAXIV WT mice has significant higher bicarbonate output in bile than KO mice, their bile flow rates were not significantly different. As shown in the results, the damage was severe and not significantly different between DDC-fed Car14-/- and WT
littermates. Obviously, the function of the CAXIV is not able to offer significant
protection against the toxic actions of DDC, which causes a very rapid release
proinflammatory and profibrogenetic cytokines and adhesion molecules, such as
TNF-α, VCAM, PDGF, TGF-β. Mechanism of action is very rapid and very severe,
with spontaneous deaths from end-state liver disease occurring within a few weeks. We
therefore conclude that this model does not reflect the “idiopathic sclerosing
cholangitis” in humans well, and that the lack of apparent protection by CAXIV
expression does not refute the “biliary bicarbonate umbrella hypothesis”. We needed to
search for a better model.
4.4. The Mdr2-/- mouse model
As described above, we found that the feeding of DDC results in very rapid and
69
very severe damage of the biliary ductular system, in which the role of biliary
bicarbonate output is not likely to play a major difference (since the exclusive loss of
CAXIV results in a slowly developing small bile duct cholangiopathy over many months). Therefore, we had to look for another model of injury to the biliary system.
Mdr2 (Abcb4) gene knockout in mice results in a deficiency in excretion of phosphatidylcholine into bile. Low biliary phospholipid levels trigger non-purulent
inflammatory cholangitis with portal inflammation and ductular proliferation
beginning shortly after birth and progressing to end-stage disease in the course of 3 to 6
months. The animals develop a phenotype resembling sclerosing cholangitis with
biliary fibrosis and develop hepatocellular carcinoma in the late stages of the disease95.
The interesting aspect of this model is that the pathophysiology of the Mdr2-/-
cholangiopathy is believed to that the lack of phospholipid transport into bile, and thus
an alteration of the mixed micelles, which under physiological situation carry the
cholesterol with phospholipids and bile acids into the gut. The lack of phospholipids
results in an alteration of those mixed micelles, with an excess of bile acid monomers,
and via an alteration of the enterohepatic circulating bile acid pool an increase in the
composition of the bile acids, with more hydrophobic bile acids in bile. We speculated
that this model is intuitively the more suitable model to study the importance of biliary
bicarbonate output or, even more precisely, the buffering capacity adjacent to the
luminal membrane of the cholangiocyte provided by the presence of CAXIV. Since we
did not want extremely early damage, and also wanted to have more genetic variation
resembling the human situation, we outcrossed the mice on a mixed Sv129 and FVBN
background, which indeed slowed the progression of disease (see results).
We created four genotypes: Car14+/+Mdr2+/+ mice is control group, Car14+/+Mdr2-/-
and Car14-/-Mdr2-/- groups can display whether there is any difference of liver damage
in Mdr2-/- mice with lower bicarbonate bile (Car14-/-) and higher bicarbonate bile
(Car14+/+) at different ages. 3 weeks old mice are too young to survive the in vivo biliary drainage experiment, but we took the liver tissue and blood sample for histology,
70
immunohistochemistry and blood gas measurements. At this early age, only the
Car14-/-Mdr2-/- mice displayed occasional mild incrassation of periductal tissue and
infiltration of inflammatory cells.
In the 6 weeks old groups, we were able to perform in vivo biliary drainage
experiments, which is per definition the most sensitive test for measuring cholestasis
(reduction in bile flow). In the basal state, the only groups that significantly differed
were the Car14+/+/Mdr2-/- and the Car14-/-/Mdr2-/- mice, which displayed a significant
difference in bicarbonate output (Fig.23C). However, after TUDCA stimulation, the
Car14+/+/Mdr2-/- mice had significant higher bicarbonate output than all other groups
even in the basal state. The underlying mechanism could be severalfold and represent
either the sequelae of the toxic bile acid effect on the biliary tree that is believed to
cause the underlying damage in this mouse model, or it could represent a protective
response. The lack of phospholipid in bile in Mdr2−/− mice induces higher
concentration of hydrophobic bile salts and free bile acids (Fig.36). The damage of hydrophobic bile salts results in inflammation of bile epithelial cells, subsequently may develop paracellular leakage of bicarbonate. A similar mechanism was found to exist in the mildly inflamed proximal colon of TNF-α overexpressing mice96. However, the fact
that additional CAXIV deletion completely abolishes the increased bicarbonate output,
even to a lower value than in the single Car14-/- mice, argues against the paracellular leakage hypothesis.
71
Fig. 36 Mechanisms of bile duct injury in the Mdr2 (Abcb4) –/– mouse cholangiopathy model. Normal: In physiological situation, bile salts are secreted from hepatocytes into bile via bile salt export pump. Phospholipids (mainly phosphatidylcholine) are excreted by phospholipid flippase (Mdr2, Mdr3 in human). Bile salt usually mixed with phosphatidylcholine and cholesterol as a mixed biliary micelles which can prevents the toxic effects of bile salts. Mdr2-/- mouse: Because of the deficiency of Mdr2, the lack of Phospholipids in bile lead to higher concentration of hydrophobic bile salts(non-micelle bile salts) which causes cholangitis.
The large bile ducts (>15 μm) have the capacity to secret bicarbonate-rich bile in
response to pathological maneuvers such as extrahepatic bile duct ligation (BDL),
partial hepatectomy, acute carbon tetrachloride (CCl4) or chronic administration of
gamma-aminobutyric acid97. Some forms of liver damage such as alcohol, toxins, or drugs cause ductal proliferation associated with increased expression of secretin receptor (SR) which is only expressed in large bile duct, and with secretin-stimulated ductal secretory activity98–100. The group of Strazzabosco found that secretin stimulates bicarbonate output in cholangiocytes by activating cAMP synthesis that induces phosphorylation of PKA, opening of CFTR and activation of the apically located
72
AE2101–103. Thus, the proliferation of the bile duct epithelial cells observed in the
Mdr2-/- mice may result in an increase in bicarbonate output. However, one would
expect increased fluid secretion if CFTR activation was part of this increased
bicarbonate output. In addition, the group of Hugo deJonge provided evidence for a
CFTR-independent mechanism for the TUDCA-induced bicarbonate-rich choleretic
response90.
We lastly speculated that this higher bicarbonate output may be an adaptive
response to the noxious effects of the hydrophobic bile acids. Bile acids have been
shown to stimulate bicarbonate secretion in the human colon as early as 1971104, and
later have been shown to do the same in other epithelia including the biliary system.
The underlying mechanism how the altered bile composition, or even the alteration of
hepatocyte ion homeostasis due to the lack of Mdr2 results in a higher bicarbonate
output into bile despite no change in bile flow at early age needs to be further
investigated.
We also noticed that the bicarbonate output of all Car14-/- mice was significant
lower than Car14+/+ mice which accord with our previous study. But only
Car14-/-Mdr2-/- mice had lower bile flow exhibited that Mdr2 deficiency will lead to
much severer injury to bile duct with weaker bicarbonate protection. Accordingly, the
histological examination of the liver samples displayed a very mild pericholangic
infiltration in the Car14-/-Mdr2-/- mice at three weeks, but a difference in the severity of
the inflammatory alterations and the bile duct proliferation at 6 weeks, as well as a
difference in the development of fibrosis.
In 11 weeks old mice, the interesting change is that the high bicarbonate output in
Car14+/+Mdr2-/- mice is abolished. The bicarbonate output is still slightly higher than
Car14-/-Mdr2-/- mice, but there is no statistical difference. Compared with 6 weeks old
Car14+/+Mdr2-/- mice, we can find the much severe pathological changes in histology as
well. Bile flow of 11 weeks Car14+/+Mdr2-/- mice was significantly reduced in contrast
with control mice. We believe that these changes were caused by the progression of
73
cholangiopathy. Higher biliary bicarbonate in Car14+/+Mdr2-/- mice can neutralize some
but not all of the toxic of hydrophobic bile salts which means these mice has slower
progression of cholangitis in comparison with Car14-/-Mdr2-/- mice, but the availability
of biliary bicarbonate cannot protect against the disease in this mouse model. Once
damage to the cholangiocytes has happened, this will reduce the bicarbonate output in
bile even in the presence of CAX IV.
4.5. The protective function of “bicarbonate umbrella” in liver
Hydrophobic (monomeric, non-micellar) bile salts induce cytotoxicity in many cell types, including hepatocytes, even at low micromolar concentrations105–107. The
long-time exposure of hepatocytes and cholangiocytes to high concentrations of
hydrophobic bile acids induces cholangiocyte impairment, which is thought to be the
increased impairment. In previous studies, it has already been confirmed that more bile
acid will get into hepatocytes and cholangiocytes in lower PH environment108. But
interestingly, human BECs are exposed to high millimolar concentrations of
hydrophobic bile salts since new born without any signs of cytotoxicity109.
Our findings perform the “reality check” for the “biliary bicarbonate umbrella”
hypothesis, which has been formulated based on studies in a cultured cholangiocyte cell
line110. We were fortunate to find an animal model in which biliary bicarbonate output
was affected without a concomitant change in bile flow, and this allowed us to
selectively study the influence of bicarbonate on the progression of cholangiopathy in
three different mouse models: A slowly progressive peripheral small duct fibrosing
cholangiopathy in the exclusive absence of the carbonic anhydrase XIV, a progressive
inflammatory and fibrosis cholangiopathy in the Mdr2-/- mouse, which was accelerated
and aggravated by the concomitant lack of the apical carbonic anhydrase XIV, and a
very rapidly progressive severe toxic hepato-cholangiopathy inflicted by feeding DDC,
in which the presence or absence of the carbonic anhydrase did not affect the disease
74
progression.
While our findings are in line with the general idea of a protective alkaline
pH-microclimate in the juxtamucosal space, i.e. directly above the epithelial cells, they
differ in the mechanism how such a microclimate is established and provide evidence
that the static “unstirred layer” hypothesis of the protective mucus bicarbonate barrier
in the gastric, duodenal or colonic mucosa111. In the stomach and duodenum, the pH of
gastric acid is 1.5 to 3.5 in the lumen which is extremely over-acidic to the gastric
epithelial cells112. The most abundant epithelial cells are mucous cells, which cover the
entire luminal surface and extend down into the glands as "mucous neck cells". These
cells secrete a bicarbonate-rich mucus that coats and lubricates the gastric surface, and
serves an important role in protecting the epithelium from acid and other chemical
insults113. Similar mechanisms have been described for the duodenum and the colon, and the general idea is that the mucus gel allows the buildup of a pH gradient within an unstirred fluid layer, and that this permits the neutralization of the back-diffusing H+
ions, even if their luminal concentrations are high and that of the measured bicarbonate
output rate relatively low. The theory has been controversial, but recent sophisticated
studies using two-photon microscopy support the concept114. However, the finding of
extracellular carbonic anhydrase activity in gastrointestinal mucus115,116 and the fact
that the presence of bicarbonate accelerates the diffusions of protons through mucus117
suggests that the presence of extracellular carbonic anhydrases dissipate the
pH-gradient and this was seen as a potential danger to the mucosa. However, our results
demonstrate that the presence of an extracellular carbonic anhydrase in the glycocalyx
increases the rate of bicarbonate output, and that the rates of gastrointestinal
bicarbonate secretion measured in vitro in Ussing Chambers and also in vivo by
application of luminal perfusion techniques have strongly underestimated the rate of
- bicarbonate secretion, because for technical reasons, no CO2/HCO3 can be present in
the luminal perfusate. We therefore believe that the rates of epithelial bicarbonate
secretion may well match those of proton diffusion though the mucus/glycocalyx in
75
gastrointestinal epithelia, and that the so-called “proton back diffusion”118, may be a
readout of the physiological neutralization of protons by bicarbonate in the mucus layer,
with subsequent intraluminal CO2 generation and diffusion of the freely permeable
nontoxic CO2.
4.6. Conclusion
In this thesis, we investigated the bicarbonate umbrella theory in liver via the
zinc-containing enzyme carbonic anhydrase XIV. The deficiency of CAXIV in hepatocytes significantly reduced the bicarbonate output in bile. Car14-/- mice induced
a long-time lower bicarbonate milieu finally developed cholangitis of primary bile duct
in physiological condition. In higher intraluminal hydrophobic bile salts situation such
as Mdr2-/- mice, the deficiency of CAXIV also resulted faster and severer cholangitis
compared to Car14+/+Mdr2-/- mice. But biliary bicarbonate didn’t offer much protective
effect to inflammation induced cholangitis in DDC-fed mice. These results indicate that biliary bicarbonate (bicarbonate umbrella) is a specific protection against the toxicity of hydrophobic bile salts. Therefore, bicarbonate umbrella is an important and indispensable protective mechanism in biliary physiology. But like the other protective
- mechanisms in organism, bicarbonate umbrella also has its limitation. The HCO3 /BA
ratio is important for the protective function of the bicarbonate umbrella119. If biliary
bicarbonate is too low or hydrophobic bile salts is too high will finally lead to BECs
damage. We believe that biliary bicarbonate can provided effective protection before or
at the early stage of BECs damage.
Based on the positive results from our in vivo study, we can make further analysis
applying molecular biology techniques such as quantitative PCR and western blot of
inflammation cytokines to support our theory of CAXIV, by which the biliary
bicarbonate mechanism can be further elucidated.
76
5. References
1. Sear, J. Anatomy and physiology of the liver. Baillieres. Clin. Anaesthesiol. 6, 697–727 (1992).
2. Barrett, K. E. & Ganong, W. F. Ganong’s review of medical physiology.P512. Ophthalmic
surgery 17, (2010).
3. Hall, J. E. & Guyton, A. C. Guyton and Hall Textbook of Medical Physiology. P.784. Journal of
Chemical Information and Modeling 53, (2011).
4. Hofmann, A. F. & Borgstroen, B. The Intraluminal Phase of Fat Digestion in Man: the Lipid
Content of the Micellar and Oil Phases of Intestinal Content Obtained During Fat Digestion and
Absorption. J. Clin. Invest. 43, 247–257 (1964).
5. Attili, A. F., Angelico, M., Cantafora, A., Alvaro, D. & Capocaccia, L. Bile acid-induced liver
toxicity: Relation to the hydrophobic-hydrophilic balance of bile acids. Med. Hypotheses 19, 57–
69 (1986).
6. Hofmann, a F. The continuing importance of bile acids in liver and intestinal disease. Arch.
Intern. Med. 159, 2647–2658 (1999).
7. Rodrigo-Torres, D. et al. The biliary epithelium gives rise to liver progenitor cells. Hepatology
60, 1367–1377 (2014).
8. Collin de l’Hortet, A. et al. Liver Regenerative-Transplantation: Regrow and Reboot. Am. J.
Transplant. Dec, 1–9 (2015).
9. Tabibian, J. H., Masyuk, A. I., Masyuk, T. V., O’Hara, S. P. & LaRusso, N. F. Physiology of
cholangiocytes. Compr. Physiol. 3, 541–565 (2013).
10. Mulkey, D. K., Henderson, R. a, Ritucci, N. a, Putnam, R. W. & Dean, J. B. Oxidative stress
decreases pHi and Na(+)/H(+) exchange and increases excitability of solitary complex neurons
from rat brain slices. Am. J. Physiol. Cell Physiol. 286, C940–51 (2004).
11. Cardone, R. a, Casavola, V. & Reshkin, S. J. The role of disturbed pH dynamics and the Na+/H+
exchanger in metastasis. Nat. Rev. Cancer 5, 786–795 (2005).
12. Casey, J. R., Grinstein, S. & Orlowski, J. Sensors and regulators of intracellular pH. Nat. Rev.
Mol. Cell Biol. 11, 50–61 (2010).
77
13. Beuers, U., Maroni, L. & Elferink, R. O. The biliary HCO3− umbrella. Curr. Opin.
Gastroenterol. 28, 253–257 (2012).
14. Chu S, M. M. The glow of the colonic pH microclimate kindled by short-chain fatty acids,
chloride and bicarbonate. J Physiol. Jun 1;517, 315 (1999).
15. Seidler, U. & Sjöblom, M. in Physiology of the Gastrointestinal Tract 1311–1339 (Elsevier,
2012). doi:10.1016/B978-0-12-382026-6.00048-8
16. Concepcion, A. R., Lopez, M., Ardura-Fabregat, A. & Medina, J. F. Role of AE2 for pHi
regulation in biliary epithelial cells. Front. Physiol. 4 JAN, 1503–12. (2014).
17. Banales, J. M. et al. Bicarbonate-rich choleresis induced by secretin in normal rat is
taurocholate-dependent and involves AE2 anion exchanger. Hepatology 43, 266–275 (2006).
18. Abuladze, N. et al. Expression and localization of rat NBC4c in liver and renal uroepithelium.
Am. J. Physiol. Cell Physiol. 287, C781–9 (2004).
19. Medina, J. F. Role of the anion exchanger 2 in the pathogenesis and treatment of primary biliary
cirrhosis. Dig. Dis. 29, 103–12 (2011).
20. Banales, J. M. et al. Up-regulation of microRNA 506 leads to decreased Cl -/HCO 3 - anion
exchanger 2 expression in biliary epithelium of patients with primary biliary cirrhosis.
Hepatology 56, 687–697 (2012).
21. Arenas, F. et al. Combination of ursodeoxycholic acid and glucocorticoids upregulates the AE2
alternate promoter in human liver cells. J. Clin. Invest. 118, 695–709 (2008).
22. Uriarte, I. et al. Bicarbonate secretion of mouse cholangiocytes involves na-hco3 cotransport in
addition to na-independent cl/hco3 exchange. Hepatology 51, 891–902 (2010).
23. Li, C., Roy, K., Dandridge, K. & Naren, A. P. Molecular assembly of cystic fibrosis
transmembrane conductance regulator in plasma membrane. J. Biol. Chem. 279, 24673–24684
(2004).
24. Lamprecht, G. & Seidler, U. The emerging role of PDZ adapter proteins for regulation of
intestinal ion transport. Am. J. Physiol. Gastrointest. Liver Physiol. 291, G766–G777 (2006).
25. Guggino, W. B. & Stanton, B. A. New insights into cystic fibrosis: molecular switches that
regulate CFTR. Nat. Rev. Mol. CELL Biol. 7, 426–436 (2006).
78
26. Banales, J. M., Prieto, J. & Medina, J. F. Cholangiocyte anion exchange and biliary bicarbonate
excretion. World J. Gastroenterol. 12, 3496–3511 (2006).
27. Tietz, P. S. et al. Agonist-induced coordinated trafficking of functionally related transport
proteins for water and ions in cholangiocytes. J. Biol. Chem. 278, 20413–20419 (2003).
28. Strazzabosco, M., Mennone, a & Boyer, J. L. Intracellular pH Regulation in Isolated Rat Bile
Duct Epithelial Cells. J. Clin. Invest 87, 1503–1512 (1991).
29. Pushkin, A. & Kurtz, I. 3 , CO 3 ) transporters : classification , function , structure , genetic
diseases , and knockout models. 90095, 580–599 (2006).
30. Gross, E. & Kurtz, I. Structural determinants and significance of regulation of electrogenic
Na(+)-HCO(3)(-) cotransporter stoichiometry. Am. J. Physiol. Renal Physiol. 283, F876–87
(2002).
31. Donowitz, M., Ming Tse, C. & Fuster, D. SLC9/NHE gene family, a plasma membrane and
organellar family of Na +/H+ exchangers. Molecular Aspects of Medicine 34, 236–251 (2013).
32. Kiela, P. R., Xu, H. & Ghishan, F. K. Apical NA+/H+ exchangers in the mammalian
gastrointestinal tract. in Journal of Physiology and Pharmacology 57, 51–79 (2006).
33. Spirli, C. et al. Functional polarity of Na+/H+ and Cl-/HCO-3 exchangers in a rat cholangiocyte
cell line. Am J Physiol Gastrointest Liver Physiol 275, G1236–1245 (1998).
34. Mennone, a et al. Role of sodium/hydrogen exchanger isoform NHE3 in fluid secretion and
absorption in mouse and rat cholangiocytes. Am. J. Physiol. Gastrointest. Liver Physiol. 280,
G247–54 (2001).
35. Strazzabosco, M. et al. Na+-dependent and -independent Cl-/HCO3 exchange mediate cellular
HCO3/- transport in cultured human intrahepatic bile duct cells. Hepatology 25, 976–985 (1997).
36. Grichtchenko, I. et al. Cloning, Characterization, and Chromosomal Mapping of a Human
Electroneutral Na+-driven Cl-HCO3 Exchanger. J. Biol. Chem. 276, 8358–8363 (2001).
37. Whittingtons, D. A. Expression, Assay, and Structure of the Extracellular Domain of Murine
Carbonic Anhydrase XIV: IMPLICATIONS FOR SELECTIVE INHIBITION OF
MEMBRANE-ASSOCIATED ISOZYMES. J. Biol. Chem. 279, 7223–7228 (2003).
38. Fujikawa-Adachi, K., Nishimori, I., Taguchi, T. & Onishi, S. Human carbonic anhydrase XIV
79
(CA14): cDNA cloning, mRNA expression, and mapping to chromosome 1. Genomics 61, 74–
81 (1999).
39. Mori, K. et al. Isolation and characterization of CA XIV, a novel membrane-bound carbonic
anhydrase from mouse kidney. J. Biol. Chem. 274, 15701–15705 (1999).
40. Kaunisto, K. et al. Carbonic anhydrase XIV: Luminal expression suggests key role in renal
acidification. Kidney Int. 61, 2111–2118 (2002).
41. Mori, K., Mukoyama, M. & Nakao, K. Searching for novel intercellular signal-transducing
molecules in the kidney and their clinical application. Clinical and Experimental Nephrology 14,
523–527 (2010).
42. Parkkila, S. et al. Expression of membrane-associated carbonic anhydrase XIV on neurons and
axons in mouse and human brain. Proc. Natl. Acad. Sci. U. S. A. 98, 1918–1923 (2001).
43. Casey, J. R., Sly, W. S., Shah, G. N. & Alvarez, B. V. Bicarbonate homeostasis in excitable
tissues: role of AE3 Cl-/HCO3- exchanger and carbonic anhydrase XIV interaction. Am. J.
Physiol. Cell Physiol. 297, C1091–102 (2009).
44. Chegwidden, W. R., Carter, N. D. & Edwards, Y. H. The Carbonic Anhydrases: New Horizons.
(Springer Science & Business Media.P.143-155, 2000).
45. Parkkila, S. et al. The plasma membrane carbonic anhydrase in murine hepatocytes identified as
isozyme XIV. BMC Gastroenterol. 2, 13 (2002).
46. Crosignani, a et al. Changes in bile acid composition in patients with primary biliary cirrhosis
induced by ursodeoxycholic acid administration. Hepatology 14, 1000–7 (1991).
47. Stiehl, A. et al. Effect of ursodeoxycholic acid on liver and bile duct disease in primary
sclerosing cholangitis. A 3-year pilot study with a placebo-controlled study period. J. Hepatol.
20, 57–64 (1994).
48. Denk, H., Stumptner, C. & Zatloukal, K. Mallory bodies revisited. Journal of Hepatology 32,
689–702 (2000).
49. Fickert, P. et al. Bile acid-induced Mallory body formation in drug-primed mouse liver. Am. J.
Pathol. 161, 2019–26 (2002).
50. Fickert, P. et al. A New Xenobiotic-Induced Mouse Model of Sclerosing Cholangitis and Biliary
80
Fibrosis. Am. J. Pathol. 171, 525–536 (2007).
51. Strazzabosco, M., Fabris, L. & Spirli, C. Pathophysiology of cholangiopathies. J. Clin.
Gastroenterol. 39, S90–S102 (2005).
52. Lazaridis, K. N., Strazzabosco, M. & Larusso, N. F. The cholangiopathies: Disorders of biliary
epithelia. Gastroenterology 127, 1565–1577 (2004).
53. Wynn, T. A. Cellular and molecular mechanisms of fibrosis. J. Pathol. 214, 199–210 (2008).
54. Trauner, M., Fickert, P. & Wagner, M. MDR3 (ABCB4) defects: A paradigm for the genetics of
adult cholestatic syndromes. Seminars in Liver Disease 27, 77–98 (2007).
55. Smit, J. J. M. et al. Homozygous disruption of the murine MDR2 P-glycoprotein gene leads to a
complete absence of phospholipid from bile and to liver disease. Cell 75, 451–462 (1993).
56. Fickert, P. et al. Regurgitation of bile acids from leaky bile ducts causes sclerosing cholangitis in
Mdr2 (Abcb4) knockout mice. Gastroenterology 127, 261–274 (2004).
57. Dinh Duc Vu, Tuchweber, B., Raymond, P. & Yousef, I. M. Tight junction permeability and
liver plasma membrane fluidity in lithocholate-induced cholestasis. Exp. Mol. Pathol. 57, 47–61
(1992).
58. Roma, M. G., Orsler, D. J. & Coleman, R. Canalicular retention as an in vitro assay of tight
junctional permeability in isolated hepatocyte couplets: effects of protein kinase modulation and
cholestatic agents. Fundam. Appl. Toxicol. 37, 71–81 (1997).
59. Alpini, G., McGill, J. M. & LaRusso, N. F. The pathobiology of biliary epithelia. Hepatology 35,
1256–1268 (2002).
60. Strazzabosco, M., Spirli, C. & Okolicsanyi, L. Pathophysiology of the intrahepatic biliary
epithelium. J. Gastroenterol. Hepatol. 15, 244–253 (2000).
61. Rivard, A. L. et al. Administration of tauroursodeoxycholic acid (TUDCA) reduces apoptosis
following myocardial infarction in rat. Am. J. Chin. Med. 35, 279–295 (2007).
62. Trauner, M. & Graziadei, I. W. Review article: mechanisms of action and therapeutic
applications of ursodeoxycholic acid in chronic liver diseases. Aliment. Pharmacol. Ther. 13,
979–996 (1999).
63. Plösch, T. et al. Abcg5/Abcg8-independent pathways contribute to hepatobiliary cholesterol
81
secretion in mice. Am. J. Physiol. Gastrointest. Liver Physiol. 291, G414–23 (2006).
64. Dikkers, A., Freak de Boer, J., Annema, W., Groen, A. K. & Tietge, U. J. F. Scavenger receptor
BI and ABCG5/G8 differentially impact biliary sterol secretion and reverse cholesterol transport
in mice. Hepatology 58, 293–303 (2013).
65. Desai, M. S. et al. Hypertrophic cardiomyopathy and dysregulation of cardiac energetics in a
mouse model of biliary fibrosis. Hepatology 51, 2097–2107 (2010).
66. Singh, A. K. et al. CFTR and its key role in in vivo resting and luminal acid-induced duodenal
HCO3- secretion. Acta Physiol. 193, 357–365 (2008).
67. Sommansson, A., Saudi, W. S. W., Nylander, O. & Sjöblom, M. The ethanol-induced
stimulation of rat duodenal mucosal bicarbonate secretion in vivo is critically dependent on
luminal Cl-. PLoS One 9, (2014).
68. Fickert, P. et al. Characterization of animal models for primary sclerosing cholangitis (PSC). J.
Hepatol. 60, 1290–1303 (2014).
69. Munck, A. et al. Management of pancreatic, gastrointestinal and liver complications in adult
cystic fibrosis. Rev. Mal. Respir. 32, 566–585 (2015).
70. Strazzabosco, M. Transport systems in cholangiocytes: Their role in bile formation and
cholestasis. in Yale Journal of Biology and Medicine 70, 427–434 (1997).
71. Beuers, U. et al. The biliary HCO3− umbrella: A unifying hypothesis on pathogenetic and
therapeutic aspects of fibrosing cholangiopathies. Hepatology 52, 1489–1496 (2010).
72. Baghdasaryan, A. et al. Dual farnesoid X receptor/TGR5 agonist INT-767 reduces liver injury in
the Mdr2 -/- (Abcb4 -/-) mouse cholangiopathy model by promoting biliary HCO3- output.
Hepatology 54, 1303–1312 (2011).
73. Schroeder, M. a. et al. Extramitochondrial domain rich in carbonic anhydrase activity improves
myocardial energetics. Proc. Natl. Acad. Sci. 110, E958–E967 (2013).
74. Sjöblom, M. et al. Duodenal acidity ‘sensing’ but not epithelial HCO3- supply is critically
dependent on carbonic anhydrase II expression. Proc. Natl. Acad. Sci. U. S. A. 106, 13094–9
(2009).
75. Furukawa, O. et al. Mechanism of augmented duodenal HCO(3)(-) secretion after elevation of
82
luminal CO(2). Am. J. Physiol. Gastrointest. Liver Physiol. 288, G557–G563 (2005).
76. Turnberg, L. A., Fordtran, J. S., Carter, N. W. & Rector, F. C. Mechanism of bicarbonate
absorption and its relationship to sodium transport in the human jejunum. J. Clin. Invest. 49,
548–556 (1970).
77. Xia, W. et al. The distinct roles of anion transporters Slc26a3 (DRA) and Slc26a6 (PAT-1) in
fluid and electrolyte absorption in the murine small intestine. Pflugers Arch. 466, 1541–56
(2014).
78. Tietge, U. J. F. et al. Secretory phospholipase A2 increases SR-BI-mediated selective uptake
from HDL but not biliary cholesterol secretion. J. Lipid Res. 49, 563–71 (2008).
79. Tanaka, K. et al. Mechanisms of impaired glucose tolerance and insulin secretion during
isoflurane anesthesia. Anesthesiology 111, 1044–51 (2009).
80. Roos, A. & Boron, W. F. Intracellular pH. Physiol. Rev. 61, 296–434 (1981).
81. Trauner, M., Meier, P. J. & Boyer, J. L. Molecular pathogenesis of cholestasis. N Engl J Med 339,
1217–1227 (1998).
82. He, H., Mennone, A., Boyer, J. L. & Cai, S. Y. Combination of retinoic acid and ursodeoxycholic
acid attenuates liver injury in bile duct-ligated rats and human hepatic cells. Hepatology 53, 548–
557 (2011).
83. Úriz, M., Sáez, E., Prieto, J., Medina, J. F. & Banales, J. M. Ursodeoxycholic acid is conjugated
with taurine to promote secretin-stimulated biliary hydrocholeresis in the normal rat. PLoS One 6,
e28717 (2011).
84. Michalopoulos, G. K. Liver regeneration. Journal of Cellular Physiology 213, 286–300 (2007).
85. Taub, R. Liver regeneration: from myth to mechanism. Nat. Rev. Mol. Cell Biol. 5, 836–847
(2004).
86. Michalopoulos, G. K. & DeFrances, M. C. Liver regeneration. Science (80-. ). 276, 60–66
(1997).
87. Shoup, M. et al. Volumetric analysis predicts hepatic dysfunction in patients undergoing major
liver resection. J Gastrointest Surg 7, 325–330 (2003).
88. Schindl, M. J., Redhead, D. N., Fearon, K. C. H., Garden, O. J. & Wigmore, S. J. The value of
83
residual liver volume as a predictor of hepatic dysfunction and infection after major liver
resection. Gut 54, 289–296 (2005).
89. Alvaro, D. et al. Hormonal regulation of bicarbonate secretion in the biliary epithelium. in Yale
Journal of Biology and Medicine 70, 417–426 (1997).
90. Bodewes, F. a J. a et al. Ursodeoxycholate modulates bile flow and bile salt pool independently
from the cystic fibrosis transmembrane regulator (Cftr) in mice. Am. J. Physiol. Gastrointest.
Liver Physiol. 302, G1035–42 (2012).
91. Svichar, N. et al. Carbonic Anhydrases CA4 and CA14 Both Enhance AE3-Mediated
Cl--HCOFormula Exchange in Hippocampal Neurons. J. Neurosci. 29, 3252–3258 (2009).
92. Vargas, L. A. & Alvarez, B. V. Carbonic anhydrase XIV in the normal and hypertrophic
myocardium. J. Mol. Cell. Cardiol. 52, 741–752 (2012).
93. Hanada, S. et al. Tumor necrosis factor-alpha and interferon-gamma directly impair epithelial
barrier function in cultured mouse cholangiocytes. Liver Int 23, 3–11 (2003).
94. Rangaswami, H., Bulbule, A. & Kundu, G. C. Osteopontin: Role in cell signaling and cancer
progression. Trends in Cell Biology 16, 79–87 (2006).
95. Mauad, T. H. et al. Mice with homozygous disruption of the mdr2 P-glycoprotein gene. A novel
animal model for studies of nonsuppurative inflammatory cholangitis and hepatocarcinogenesis.
Am. J. Pathol. 145, 1237–45 (1994).
96. Juric, M. et al. Increased epithelial permeability is the primary cause for bicarbonate loss in
inflamed murine colon. Inflamm. Bowel Dis. 19, 904–911 (2013).
97. Glaser, S. S. et al. Morphological and functional heterogeneity of the mouse intrahepatic biliary
epithelium. Lab. Invest. 89, 456–69 (2009).
98. Alpini, G. et al. Upregulation of secretin receptor gene expression in rat cholangiocytes after bile
duct ligation. Am. J. Physiol. 266, G922–8 (1994).
99. Alpini, G. et al. Molecular and functional heterogeneity of cholangiocytes from rat liver after
bile duct ligation. Am J Physiol 272, G289–97 (1997).
100. Lesage, G. et al. Regrowth of the rat biliary tree after 70% partial hepatectomy is coupled to
increased secretin-induced ductal secretion. Gastroenterology 111, 1633–44 (1996).
84
101. Alpini, G. et al. Large but not small intrahepatic bile ducts are involved in secretin-regulated
ductal bile secretion. Am J Physiol 272, G1064–74 (1997).
102. Alpini, G., Lenzi, R., Sarkozi, L. & Tavoloni, N. Biliary physiology in rats with bile ductular cell
hyperplasia. Evidence for a secretory function of proliferated bile ductules. J. Clin. Invest. 81,
569–578 (1988).
103. Alpini, G. et al. Heterogeneity of the proliferative capacity of rat cholangiocytes after bile duct
ligation. Am. J. Physiol. 274, G767–G775 (1998).
104. Mekjian, H. S., Phillips, S. F. & Hofmann, A. F. Colonic secretion of water and electrolytes
induced by bile acids: perfusion studies in man. J. Clin. Invest. 50, 1569–1577 (1971).
105. Perez, M. J. & Britz, O. Bile-acid-induced cell injury and protection. World J. Gastroenterol. 15,
1677–1689 (2009).
106. Rust, C. et al. Bile acid-induced apoptosis in hepatocytes is caspase-6-dependent. J. Biol. Chem.
284, 2908–2916 (2009).
107. Spivey, J. R., Bronk, S. F. & Gores, G. J. Glycochenodeoxycholate-induced lethal hepatocellular
injury in rat hepatocytes: Role of ATP depletion and cytosolic free calcium. J. Clin. Invest. 92,
17–24 (1993).
108. Hohenester, S., Maillette De Buy Wenniger, L., Jefferson, D. M., Oude Elferink, R. P. & Beuers,
U. Biliary bicarbonate secretion constitutes a protective mechanism against bile acid-induced
injury in man. Dig. Dis. 29, 62–65 (2011).
109. Hofmann, A. F. Bile acids: Trying to understand their chemistry and biology with the hope of
helping patients. Hepatology 49, 1403–1418 (2009).
110. Hohenester, S. et al. A biliary HCO 3 - umbrella constitutes a protective mechanism against bile
acid-induced injury in human cholangiocytes. Hepatology 55, 173–183 (2012).
111. Bachmann, O. & Seidler, U. News from the end of the gut--how the highly segmental pattern of
colonic HCO − transport relates to absorptive function and mucosal integrity. Biol. Pharm. Bull.
34, 794–802 (2011).₃
112. Hoehn, K. & Marieb, E. N. Human Anatomy & Physiology. Benjamin-Cummings Publ. Co. 306
(2012). doi:10.1007/BF00845519
85
113. Kauffman Jr., G. L. Gastric mucus and bicarbonate secretion in relation to mucosal protection.
J Clin Gastroenterol 3, 45–50 (1981).
114. Baumgartner, H. K. & Montrose, M. H. Regulated Alkali Secretion Acts in Tandem with
Unstirred Layers to Regulate Mouse Gastric Surface pH. Gastroenterology 126, 774–783
(2004).
115. Endeward, V., Kleinke, T. & Gros, G. Carbonic anhydrase in the gastrointestinal mucus of
mammals--possible protective role against carbon dioxide. Comp. Biochem. Physiol. A. Mol.
Integr. Physiol. 136, 281–7 (2003).
116. Kleinke, T. et al. A distinct carbonic anhydrase in the mucus of the colon of humans and other
mammals. J. Exp. Biol. 208, 487–96 (2005).
117. Desai, M. A. & Vadgama, P. M. Bicarbonate and other buffer systems can enhance the rate of H+
diffusion through mucus in vitro. BBA - Gen. Subj. 1116, 43–49 (1992).
118. TERNER, C. The reduction of gastric acidity by back-diffusion by hydrogen ions through the
mucosa. Biochem. J. 45, 150–8 (1949).
119. Baghdasaryan, A. et al. Inhibition of intestinal bile acid absorption improves cholestatic liver
and bile duct injury in a mouse model of sclerosing cholangitis. J. Hepatol. 64, 674–681 (2016).
86
6. Abstract Background and aim: One of the most important functions of the liver is to produce bile which aids in digestion via the emulsification of lipids. Bile acids are the major components of bile. But they are also important contributors to diseases of bile duct. The acidic toxicity of high-molarity hydrophobic bile salt monomers can injure biliary epithelial cells (BECs) which may develop pathological changes and ultimately lead to end-stage hepatic failure. The ‘biliary bicarbonate umbrella’ is considered to be a critical protective mechanism of BECs. Abundant apical secretion of bicarbonate from BECs may provide protection against uncontrolled penetration of protonated bile acid monomers and hydrophobic bile salt monomers, which are transferred from bile via apical membranes into the cholangiocytes. Carbonic anhydrase 14 (CAXIV) is a zinc-containing enzyme which catalyzes the reversible hydration of carbon dioxide. CAXIV has been proved that it plays the key element of normal bicarbonate homeostasis in excitable tissues. High mRNA expression of CAXIV is also found in liver tissues, and immunohistochemical studies localize CAXIV predominantly to the canalicular membrane of hepatocytes and the cholangiocyte apical membrane. We hypothesized that CAXIV played a critical role in bicarbonate protection of the liver. Methods: We measured the bile flow and bicarbonate output before and after tauroursodeoxycholic acid (TUDCA) stimulation in anesthetized Car14 knockout mice, 3,5-diethoxycarbonyl-1,4-dihydro-collidine (DDC) -fed mice and Car14/Mdr2 double knockout mice as well as the wild types at different ages. The liver tissues of these mice were fixed, sectioned and analysed with histological and immunohistochemical methods (H&E staining, Sirius Red staining, CK19 immunohistochemistry staining), as well as Western and quantitative PCR analysis were performed. Results: The Car14-/- mice had a significantly lower bicarbonate output in comparison with Car14+/+ mice. Histological changes consistent with fibrosis and sclerosing cholangiopathy, as well as a reduction of bile flow rates were observed in 1-year-old Car14-/- mice compare to younger mice both with Car14-/- and Car14+/+. Similar differences were observed between 6-week-old Car14+/+Mdr2-/- and Car14-/-Mdr2-/- mice, where lower bicarbonate output correlated with lower bile flow and histological features of cholangitis in the latter in the 3 and 6 week-old groups. But in the 11-week-old Car14+/+Mdr2-/- and Car14-/-Mdr2-/- mice, these differences became smaller, most likely because the progressive nature of the disease in the Mdr2-/- mice is only slowed, but not fully prevented by a higher bicarbonate output facilitated in the presence of CAXIV expression. In the mouse model with DDC induced liver damage,
87 no difference in bile flow and histology were observed between the Car14 knock-out and wild type mice. Conclusions: Deficiency of CAXIV in hepatocytes significantly reduced the bicarbonate output in bile which caused series of functional and histological changes. Biliary bicarbonate can provide effective protection to BECs, but cannot fully prevent injury caused by severe bile duct injury.
88
7. Eigenständigkeiterklärung
Erklärung nach § 2 Abs. 2 Nrn. 6 und 7 der Promotionsordnung der Medizinischen Hochschule Hannover Ich erkläre, dass ich die der Medizinischen Hochschule Hannover zur Promotion eingereichte Dissertation mit dem Titel Genetic deletion of the carbonic anhydrase XIV induces low biliary bicarbonate output and enhances the toxic effect of bile acids to the cholangiocytes - an “in vivo” confirmation of the “biliary bicarbonate umbrella” hypothesis in der Klinik für Gastroenterologie, Hepatologie und Endokrinologie unter Betreuung von Prof. Dr. med. Ursula Seidler mit der Unterstützung durch Dr. Henrike
Lenzen und Dr. Brigitte Riederer oder in Zusammenarbeit mit Zhenzhen Zhou ohne sonstige Hilfe durchgeführt und bei der Abfassung der Dissertation keine anderen als die dort aufgeführten Hilfsmittel benutzt habe.
Die Gelegenheit zum vorliegenden Promotionsverfahren ist mir nicht kommerziell vermittelt worden. Insbesondere habe ich keine Organisation eingeschaltet, die gegen Entgelt Betreuerinnen und Betreuer für die Anfertigung von Dissertationen sucht oder die mir obliegenden Pflichten hinsichtlich der Prüfungsleistungen für mich ganz oder teilweise erledigt.
Ich habe diese Dissertation bisher an keiner in- oder ausländischen Hochschule zur Promotion eingereicht. Weiterhin versichere ich, dass ich den beantragten Titel bisher noch nicht erworben habe.
Ergebnisse der Dissertation werden in keinem Publikationsorgan veröffentlicht.
Hannover, den ______
______(Jiajie Qian)
89
8. Curriculum Vitae
Jiajie Qian
Attending Doctor Department of Gastrointestinal Surgery The First Affiliated Hospital, College of Medicine, Zhejiang University 79 Qingchun Road, Hangzhou, Zhejiang province, P.R. China, 310003 Tel: +86-571-87236147 Fax: +86-571-87236570 Cell phone: +86-13958165545 Email: [email protected]
Personal Information:
Date of Birth July 28th, 1978
Place of Birth Hangzhou, Zhejiang Province
Gender Male
Degree Master of Medicine
Positions:
2007-Present Attending Doctor, Department of Gastrointestinal Surgery The First Affiliated Hospital, College of Medicine, Zhejiang
University
2002-2007 Resident Doctor, Department of General Surgery The First Affiliated Hospital, College of Medicine, Zhejiang
University
90
Education:
2013-2016 Doctoral student of Medicine, Department of Gastroenterology, hepatology & Endocrinology, Hannover Medical School
2007-2010 Master of Medicine, Department of General Surgery
College of Medicine, Zhejiang University
1997-2002 Bachelor of Medicine, Department of Clinical Medicine
Wenzhou Medical College
Research Experiences:
Take part in the research work granted by National Natural Science Foundation of
China (8107964)
Role of plasmacytoid dendritic cells in the induction of regulatory T cells generating
and accumulating in human gastric cancer microenvironment. In charge in the research work granted by Natural Science Foundation of Zhejiang
Province(Y2100092) Research of Autophagy related gene ATG-12 in chemotherapy resistance of Gastric Cancer
Publications:
1. Qian JJ, Li FQ, Yu JR, Yang ZL. Mining Predictive Biomarker for Neoadjuvant Chemotherapy in Gastric Cancer by Proteomics. Hepato-Gastroenterology 58 (2011):1828-1833 2. Qian JJ, Wu XY, Lin Y. Hepatitis B virus X protein disrupts DNA interstrand crosslinking agent mitomycin C induced ATR dependent intra-S-phase checkpoint. European Journal Of Cancer 44(2008):1596-1602 3. Xu GQ, Chen MH, Qian JJ, Ren GP, Chen HT. Endoscopic ultrasonography for the diagnosis and selecting treatment of esophageal leiomyoma. J Gastroenterol Hepatol. 2012 Mar;27(3):521-5.
91
9. Acknowledgement
First, I’d like to express my deepest respect to Prof. Dr. med. Ursula Seidler. She provided me the opportunity to study in Germany, taught me the professional knowledge, and guided me to do the research, and more important thing is her serious attitude to work, all these experiences will be beneficial to me in the future, and I hope I can keep them forever. It is absolutely necessary to thank all the support from our lab, many thanks to Brigitte Riederer, Brigitte Rausch, Anurag Kumar Singh, Sunil Yeruva, Xuemei Liu, Ming Luo, Yongjian Liu, Weiliang Xia and Taolang Li. They taught me the basic experimental techniques, managed all the supportive work. Most important, particular thanks to Prof. Dr. Christian Stock, Prof. Dr. Daniela Seidler, PhD. Katerina Nikolovska, PhD. Ke Xiao and Dr.med. Alok Garg, you are not only colleagues but dear friends to me, without your support and friendship, I cannot make through these tough days, I mean it. Thanks to Rong Song, Wen Zheng, Rongjun Chen and Xiaokun Liu for their kindly help during the life in Germany. Finally, I am so grateful to my family for their love, sacrifices, support and understanding in my life. I owe them so so much, beyond words. These years in Germany are not that easy. But all the people that I met here, the life and happiness that I had here, even the loneliness, struggle and upset that I have been through here, all of these definitely will be the treasure in my life, and I will keep it in my memory forever. So, goodbye, Hannover!
Jiajie Qian 07.10.2017
92