THE NLRC4 INFLAMMASOME AND ITS REGULATION
OF LIVER DISEASE PATHOGENESIS
by
DAVID ANDREW DESANTIS
Submitted in partial fulfillment of the requirements
for the degree of Doctor of Philosophy
Dissertation Advisor: Colleen M. Croniger, Ph.D.
Department of Nutrition
CASE WESTERN RESERVE UNIVERSITY
August 2015 CASE WESTERN RESERVE UNIVERSITY
SCHOOL OF GRADUATE STUDIES
We hereby approve the dissertation of
DAVID ANDREW DESANTIS candidate for the degree of Doctor of Philosophy* ______
(signed) Noa Noy, PhD (Committee Chair)
Colleen Croniger, PhD
Laura Nagy, PhD
George Dubyak, PhD
Date of Defense May 13, 2015
*We also certify the written approval has been obtained for any proprietary material contained therein.
ii
DEDICATION
I dedicate this work to my mother, father, wife, and children. To my mother, who taught
me education is a priority. Knowledge has come first, everything else has followed. To
my father, who showed me the value of a strong work ethic. I derive all my persistence
from you.
To my wife, Caitlin, without whom I would most certainly not have achieved this
milestone. Your love, strength, and support has given me this wonderful opportunity. I
am forever grateful.
To my children, Sophie and Daniel. The moment I laid eyes on you I knew every achievement thereafter would be for you. This body of work is no different.
iii
TABLE OF CONTENTS
Table of Contents ...... iv
List of Tables ...... x
List of Figures ...... xi
Preface...... xv
Acknowledgements ...... xvi
List of Abbreviations ...... xviii
Abstract ...... xxiii
CHAPTER 1: THE LIVER
1.1. Human Liver Anatomy ...... 1
1.1.1. Gross Anatomy of the Human Liver ...... 1
1.1.2. Microscopic Anatomy of the Human Liver ...... 4
1.2. Mouse Liver Anatomy ...... 8
1.3. Cells of the Liver ...... 8
1.3.1. Hepatocyte ...... 8
1.3.1.1. Lipid Synthesis ...... 9
1.3.1.2. Lipid Oxidation ...... 9
1.3.1.3. Lipid Transport...... 10
1.3.1.4. Detoxification ...... 10
iv
1.3.2. Kupffer Cell ...... 13
1.3.3. Hepatic Stellate Cell ...... 14
1.3.4. Sinusoidal Endothelial Cell...... 16
CHAPTER 2: DISEASES OF THE LIVER
2.1. Overview ...... 18
2.2. Alcoholic Liver Disease ...... 18
2.3. Nonalcoholic Fatty Liver Disease ...... 19
CHAPTER 3: MURINE MODELS OF LIVER INJURY
3.1. Xenobiotic ...... 22
3.1.1. Carbon Tetrachloride ...... 22
3.1.2. Concanavalin A ...... 23
3.1.3. Ethanol ...... 25
3.2. Surgical ...... 28
3.2.1. Bile Duct Ligation...... 28
3.3. Dietary ...... 29
3.3.1. High-Fat Diet ...... 29
3.3.2. Methionine and Choline Deficient Diet ...... 30
CHAPTER 4: LIVER REGENERATION
4.1. Overview ...... 32
4.2. Two-Thirds Partial Hepatectomy ...... 35
v
4.3. Other Models of Liver Regeneration ...... 44
4.3.1. Concanavalin A ...... 44
4.3.2. Carbon Tetrachloride ...... 45
4.3.3. D-galactosamine ...... 45
CHAPTER 5: MOUSE GENETICS
5.1. Inbred Genetic Mouse Strains ...... 47
5.2. Chromosome Substitution and Congenic Mouse Strains ...... 48
CHAPTER 6: IMMUNITY
6.1. Overview ...... 53
6.2. The Innate Immune Response ...... 53
6.2.1. Neutrophils ...... 54
6.2.2. Cytokines ...... 55
CHAPTER 7: THE INFLAMMASOME
7.1. Overview ...... 57
7.2. Inflammasome Structure ...... 58
7.3. NLR Family ...... 60
7.4. Inflammasome Function ...... 65
7.4.1. DAMPs and PAMPs ...... 66
7.4.2. Caspase-1 ...... 68
7.4.3. Apoptosis ...... 69
vi
7.4.3.1. Overview ...... 69
7.4.3.2. Extrinsic Apoptotic Pathway ...... 71
7.4.3.3. Intrinsic Apoptotic Pathway ...... 72
7.4.3.4. Perforin-Granzyme-Dependent Apoptotic Pathway ...... 73
7.4.4. Pyroptosis ...... 74
7.4.5. Interleukin-1β ...... 74
7.4.6. Interleukin-18 ...... 79
7.4.7. Interleukin-6 ...... 81
7.5. NLRC4 Inflammasome ...... 87
7.5.1. Overview ...... 87
7.5.2. Structure and Assembly of the NAIP/NLRC4 Inflammasome ...... 87
7.5.3. NLRC4-Mediated Pyroptosis...... 91
7.5.4. Current Model ...... 93
CHAPTER 8: PUBLICATION
8.1. DeSantis DA, Lee P, Doerner SK, Ko CW, Kawasoe JH, Hill-Baskin AE,
Ernest SR, Bhargava P, Hur KY, Cresci GA, Pritchard MT, Lee CH, Nagy
LE, Nadeau JH, Croniger CM. Genetic resistance to liver fibrosis on A/J
mouse chromosome 17. Alcohol Clin Exp Res. 2013 Oct;37(10):1668-79...... 94
CHAPTER 9: PUBLICATION
9.1. DeSantis DA, Ko CW, Wang L, Lee PL, Croniger CM. Constitutive
Activation of the Nlrc4 Inflammasome Prevents Hepatic Fibrosis and
vii
Promotes Hepatic Regeneration after Partial Hepatectomy. Submitted to
Fibrogenesis Tissue Repair. 4/15/15...... 144
CHAPTER 10: DISCUSSION, IMPLICATIONS, AND FUTURE
DIRECTIONS
10.1. Identification of Novel Candidate Genes in the Progression of Liver Injury ...202
10.1.1. Discussion and Implications ...... 202
10.1.2. Future Directions ...... 207
10.2. The NLRC4 inflammasomes role in hepatic fibrogenesis and liver
regeneration...... 208
10.2.1. Discussion and Implications ...... 209
10.2.2. Future Directions ...... 212
10.3. Constitutive activation of NLRC4 and its role in hepatic regeneration ...... 213
10.3.1. Discussion and Implications ...... 213
10.3.2. Future Directions ...... 216
10.4. Overall Summary and Implications ...... 218
CHAPTER 11: APPENDIX
11.1. DeSantis DA, Ko CW, Liu Y, Liu X, Hise AG, Nunez G, Croniger CM.
Alcohol-induced liver injury is modulated by Nlrp3 and Nlrc4
inflammasomes in mice. Mediators Inflamm. 2013;2013:751374...... 221
11.2. Permission For The Use Of Copyrighted Material ...... 264
viii
CHAPTER 12: LITERATURE CITED ...... 330
ix
LIST OF TABLES
CHAPTER 1
Table 1.1. Cells found in the liver ...... 12
CHAPTER 7
Table 7.1. Nomenclature of NLR...... 62
CHAPTER 8
Table 8.1. Lieber–DeCarli diet study ...... 143
Table 8.2. Summary of candidate genes ...... 144
CHAPTER 11
Table 11.1. Food intake and blood alcohol ...... 263
x
LIST OF FIGURES
CHAPTER 1
Figure 1.1. Segments of the human liver ...... 3
Figure 1.2. Three-dimensional structure of a liver acinus ...... 7
CHAPTER 3
Figure 3.1. General scheme of ethanol oxidation ...... 27
CHAPTER 4
Figure 4.1. Prometheus ...... 34
Figure 4.2. Schematic drawings of mouse liver anatomy and positioning of silk
threads for knots ...... 38
Figure 4.3. Kinetics of BrdU incorporation after 2/3 PH ...... 39
CHAPTER 5
Figure 5.1. Selected analytical tools in mouse genetics ...... 50
CHAPTER 7
Figure 7.1. IL-6 classic-signaling and IL-6 trans-signaling ...... 84
Figure 7.2. Overall Structure of mNLRC4 ...... 89
CHAPTER 8
xi
Figure 8.1. Congenic strains from chromosome substitution strain-17 (CSS-17) ....131
Figure 8.2. CYP2E1 induction with alcohol consumption ...... 132
Figure 8.3. Measurements of liver injury ...... 133
Figure 8.4. Response to chronic CCl4 administration in congenic strains ...... 134
Figure 8.5. Reduced liver fibrosis in 17C-6 congenic strain ...... 135
Figure 8.6. Immunohistochemistry with α-smooth muscle actin α-SMA ...... 136
Figure 8.7. Detection of α-SMA ...... 137
Figure 8.8. Expression of candidate genes in congenic strains ...... 138
Figure 8.9. Expression of Noxo1 in cells isolated from the liver ...... 139
Figure 8.10. Expression of Nlrc4 in cells isolated from the liver ...... 140
Figure 8.S1. Expression levels for candidate genes after ethanol feeding study ...... 141
Figure 8.S2. Noxo1 mRNA expression in bone marrow derived macrophages
(BMDM) ...... 142
CHAPTER 9
Figure 9.1. Nlrc4 promoter and Cdx-1 binding sites ...... 188
Figure 9.2. Cdx-1 binds to the mouse Nlrc4 promoter ...... 189
Figure 9.3. Cdx-1 governs Nlrc4 gene expression in murine macrophage cell
line RAW264.7 ...... 190
Figure 9.4. Bone marrow-derived macrophages from 17C-6 congenic mouse
produce more IL-1β over time ...... 191
Figure 9.5. 17C-6 congenic mice have increased plasma IL-1β after chronic
CCl4 ...... 192
xii
Figure 9.6. 17C-6 have increased regenerative capacity after 2/3 partial
hepatectomy ...... 193
Figure 9.7. 17C-6 have increased mRNA of inflammation and cellular
proliferation genes after sham 2/3 partial hepatectomy ...... 194
Figure 9.8. 17C-6 have increased Cyclin D1 protein after 2/3 partial
hepatectomy ...... 195
Figure 9.9. 17C-6 have increased BrdU incorporation after 2/3 partial
hepatectomy ...... 196
Figure 9.10. 17C-6 have increased plasma IL-18 and IL-6 after 2/3 partial
hepatectomy ...... 197
Figure 9.11. 17C-6 have increased plasma IL-1β after 2/3 partial hepatectomy ...... 198
Figure 9.12. 17C-6 congenic mice have similar liver damage to B6 after a single-
dose of CCl4 ...... 199
Figure 9.13. 17C-6 congenic mice have increased levels of Cyclin D1 after a
single-dose of CCl4 ...... 200
Figure 9.14. Proposed model of Nlrc4 induced liver regeneration ...... 201
CHAPTER 11
Figure 11.1. Analysis of steatosis with alcohol feeding ...... 254
Figure 11.2. Measurements of liver injury ...... 255
Figure 11.3. CYP2E1 induction with alcohol consumption ...... 256
Figure 11.4. Measurement of αSMA and hydroxyproline ...... 257
Figure 11.5. Expression of hepatic TNF-α ...... 258
xiii
Figure 11.6. Hepatic expression of Nlrp3, Nlrc4, and Naip5 mRNA ...... 259
Figure 11.7. Altered IL-1β and IL-18 expression in the liver in Nlrp3-/- and
Nlrc4-/- mice ...... 260
Figure 11.8. Increased caspase-3/7 in the liver of Nlrp3-/- mice ...... 261
Figure 11.9. Decreased phosphorylation of STAT3 in the liver in Nlrp3-/- and
Nlrc4-/- mice ...... 262
xiv
PREFACE
This dissertation is original work by the author, David A. DeSantis.
xv
ACKNOWLEDGEMENTS
The completion of this degree would not be possible without the assistance, guidance, and mentorship of many people throughout my academic career. First, I would like to thank Dr. John Kopchick and Dr. Darlene Berryman of the Edison Biotechnology
Institute in Athens, Ohio. Dr. Kopchick was an inspiration in the early years of my undergraduate research. His passion for science and the charismatic way in which he conveyed his team’s discoveries left a lasting impression. Dr. Berryman was the single most influential person in my professional career from my time at Ohio University. She provided an accepting atmosphere to novice students who were eager to learn. She is responsible for nurturing the foundation of my academic career and I am incredibly lucky to have had the opportunity of her mentorship. I was exposed to a wonderful, tightknit research team at EBI which truly was a motivation to continue my education.
Secondly, I would like to thank the academic community of Case Western Reserve
University. This institute is a cauldron of knowledge and truly was the best location to further my academic pursuits. I would like to acknowledge the faculty and staff of the
CWRU Department of Nutrition for the support and opportunity they have provided.
Thank you to the CWRU Mouse Metabolic Phenotyping Center staff including Dr.
Michelle Puchowicz, and Lan Wang for their advice and technical support. I would like to thank past and present members of the Croniger lab who have assisted me in countless experiments, specifically Dr. Chang-Wen Hsieh and Chih-wei Ko. I would also like to
xvi
thank past and present CWRU Department of Genetics students and staff, specifically Dr.
Jason Heaney and Dr. Stephanie Doerner for their support and expertise in many projects.
Third, I would like to thank the members of my dissertation committee: Dr. Noa Noy, Dr.
Laura Nagy, and Dr. George Dubyak. My committee continually tested my thought
processes and critically evaluated my work, all the while showing genuine interest in my research and growth as a scientist. I am truly thankful to each member for their time, encouragement, and advice.
Last, I would like to thank Dr. Colleen Croniger. I have had the great fortune to call Dr.
Croniger my mentor and friend for the extent of my time at CWRU. From the beginning she has shown unwavering support of my academic pursuits, continually urging me to take the next step. She has seen the highest of my accomplishments, and lowest of my failures, always there to offer kind words, support, and motivation to dust myself off and try again. She is an absolutely brilliant scientist, passionate teacher, but an even better person. I’m am grateful for the immeasurable time and effort she has spent mentoring me. I would not be writing this dissertation if she was absent from my life, of that I am certain. Thank you Dr. Croniger. Thank you.
xvii
LIST OF ABBREVIATIONS
2/3PH two-thirds partial hepatectomy
ALT alanine aminotransferase
ANOVA analysis of variance
AP-1 activator protein 1
ASC apoptosis-associated speck-like protein containing a caspase activation and recruitment domain
AST aspartate aminotransferase
ATP adenosine triphosphate
B6 C57Bl/6J
BMDMs bone marrow-derived macrophages
BrdU bromodeoxyuridine
CAD caspase-activated DNase
CARD caspase activation and recruitment domain
CCl4 carbon tetrachloride
Cdx-1 caudal-related homeobox-1 c-Fos FBJ osteosarcoma oncogene
ChIP chromatin immunoprecipitation
CMV cytomegalovirus c-Myc myelocytomatosis oncogene
Con-A concanavalin A
xviii
Crp c-reactive protein
CSS chromosome substitution strain
CSS17 B6-Chr17A/J
CTL cytotoxic T lymphocytes
CYP2E1 cytochrome P450, family 2, subfamily E, polypeptide 1
DAMPs damage-associated molecular patterns
ECM extracellular matrix proteins
ELISA enzyme-linked immunosorbent assay
EtOH ethanol
FADD fas-Associated protein with death domain
GWAS genome-wide association studies
H&E hematoxylin and eosin
HAV hepatitis A virus
HBV hepatitis B virus
HCV hepatitis C virus
HFD high-fat diet
HFHS high-fat, high-sucrose containing diet
HFLS high-fat, low-sucrose containing diet
HFSC high-fat, simple-carbohydrate
HSC hepatic stellate cells
HSC-70 heat shock cognate-70
IHC immunohistochemistry
IL-1 interleukin-1
xix
IL-18 interleukin-18
IL-18Rα IL-18 receptor alpha chain
IL-18Rβ IL-18 receptor beta chain
IL-1β interleukin-1β
IL-6 interleukin-6
IκBα NF-Kappa-B Inhibitor-α
JAK janus kinase
KC Kupffer cells
LFSC low-fat, simple-carbohydrate
LPS lipopolysaccharide
LRR leucine-rich repeat
LSEC liver sinusoidal endothelial cells
MAPKs mitogen-activated protein kinases
MCD methionine and choline deficient diets
M-CSF macrophage colony-stimulating factor mIL-6R membrane bound interleukin-6 receptor
MyD88 myeloid differentiation primary response 88
NACHT nucleotide-binding and oligomerization
NAD nicotinamide adenine dinucleotide
NAFLD nonalcoholic fatty liver disease
NAIP neuronal apoptosis inhibitory protein
NAIP5 NLR family, apoptosis inhibitory protein 5
NASH nonalcoholic steatohepatitis
xx
NBD nucleotide-binding domain
NEFA non-esterified fatty acids
NLRC4 NOD-like receptor containing a caspase activating and recruitment domain 4
Nlrc4-MAS Nlrc4-macrophage activation syndrome
NLRP3 NLR family pyrin domain containing 3
NLR NOD-like receptors
NOD nucleotide-binding oligomerization domain
NOMID neonatal onset multisystem inflammatory disease
NOXO1 NADPH Oxidase Organizer 1
Obrq obesity resistant QTLs
PAMPs pathogen-associated molecular patterns
PCR polymerase chain reaction
PF pair-fed
PRR pattern recognition receptors
PYD pyrin domain qPCR real-time quantitative reverse transcription PCR
QTL quantitative trait loci rIL-6 recombinant IL-6
ROS reactive oxygen species
SEM standard error of the mean sIL-6R soluble IL-6 receptor siRNA small interfering RNA
xxi
SNP single nucleotide polymorphism
STAT3 signal transducer and activator of transcription 3
T3SS type III secretion systems
TLR toll-like receptor
TNF-α tumor necrosis factor α
VLDL very low-density lipoprotein
α-SMA α-smooth muscle actin
xxii
The NLRC4 Inflammasome and its Regulation of Liver Disease
Abstract
by
DAVID ANDREW DESANTIS
Nonalcoholic fatty liver disease (NAFLD) is form of chronic liver disease with high
prevalence in the general population. As NAFLD progresses, the liver become inflamed, fibrotic, and eventually cirrhotic. The NLRC4 inflammasome produces inflammatory cytokines in response to bacterial component stimuli within the cell cytosol. We hypothesized that mild constitutive activation of the NLRC4 inflammasome would prevent hepatic fibrogenesis and induce hepatic regeneration after liver injury.
Substitution of the A/J region of the genome into C57BL/6 (B6) mouse that encompasses the gene NLRC4 has shown to be hepatoprotective when mice were administered chronic and acute doses of hepatotoxin carbon tetrachloride (CCl4). Congenic mouse 17C-6 have two single nucleotide polymorphisms in NLRC4. Congenic 17C-6 have a portion of their
17th chromosome substituting the B6 form for the A/J form. The first polymorphism
(rs74459439-T), previously unidentified, is situated in the promoter of NLRC4.
Transcription factor binding analysis indicates this site is a potential binding site for
xxiii
transcription factor Cdx-1. This location appears to govern transcription of NLRC4 as
the A/J allele displays increased transcriptional activity.
The second polymorphism is located within the coding region of NLRC4 producing a nonsynonomous mutation leading to an amino acid substitution in the leucine rich repeat
(LRR) domain of NLRC4. The LRR is known to regulate activation of NLRC4 and 17C-
6 mice demonstrate increased production and maturation of the NLRC4 inflammasome downstream product, IL-1β. Congenic 17C-6 are resistant to CCl4-induced hepatic fibrogenesis and have remarkably accelerated regeneration of hepatic mass after two-
thirds partial hepatectomy.
Resistance to liver disease progression coupled with the acceleration of hepatic
regeneration is attributable to the NLRC4 inflammasome response. Increased secretion
of proinflammatory cytokines stimulates the release interleukin-6 (IL-6), an established
initiator of hepatocyte cell survival and proliferation. Congenic 17C-6 have elevated plasma IL-6 after two-thirds partial hepatectomy. Through this increased activation of the
NLRC4 inflammasome, hepatocytes display an increased rate of mitotic division to
recover lost tissue after damage. This hepatoprotective and regenerative response through
stimulation of the NLRC4 inflammasome complex may be a potential therapeutic
mechanism for individuals with fibrotic liver disease potentially mitigating the progression of liver disease.
xxiv
LITERATURE REVIEW
CHAPTER 1
THE LIVER
1.1. Human Liver Anatomy
1.1.1. Gross Anatomy of the Human Liver
The human liver is subdivided into 2 lobes and is the largest gland in the human body
comprising 2% to 3% of average body weight [1]. Anatomically it’s positioned directly
inferior and slightly to the right of the diaphragm. It is surrounded by the ribcage and is
encapsulated by a fibrous membrane known as the Glisson’s capsule. It’s also covered in
the anterior and posterior by the visceral peritoneum. The liver is supported from the
superior by the coronary ligament suspended from the diaphragm which divides into the
right and left triangular ligaments. At its anterior it is braced by the falciform ligament
and at the porta hepatis by the gastrohepatic and hepatoduodenal ligaments [2]. The
lesser curvature of the stomach is connected to the left hepatic lobe by the gastrohepatic
ligament which also contains the hepatic portion of the vagus nerve. The duodenum of
the small intestine associates with the liver through the hepatoduodenal ligament and
porta hepatis [1].
From the portal trunk the hepatic portal vein divides into the left portal vein branch and
the right portal vein branch. The right portal vein branch further divides into the right
anterior portal vein (feeding liver segments V and VIII) and the right posterior vein
1
(feeding liver segments VI and VII). The left portal vein feeds the left lobe (liver
segments II, III, and IV) (Figure 1.1). Segment I of the left lobe is fed by both the left
and right portal veins [3]. Portal vein variants are known to occur in 20% to 35% of the
general population [4].
Multiple ducts make up the intrahepatic biliary tree responsible for the transport of bile
synthesized in the liver. Simply put, the left hepatic duct drains the left segments of the liver and the right hepatic duct, the right segments of the liver. The right duct is known to have some anatomical variability, as noted by surgeons. The two ducts join at the hilar plate and form the common hepatic duct. The common hepatic duct continues caudally to join the cystic duct forming the common bile duct. The gallbladder sits on the posterior face of segments IV and V. The gallbladder is connected to the common bile duct via the cystic duct. The common bile duct proceeds along the hepatoduodenal ligament, draining into the duodenum through the ampulla of Vater [1].
2
Figure 1.1. Segments of the human liver.
The human liver is divided by the falciform ligament into an anatomical right lobe and left lobe. However, the liver also has a functional right and left side, divided by Cantlie's line: a hypothetical line from the gallbladder fossa to the middle hepatic vein. Each functional hemi-liver is composed of two sections: on the right, an anterior section (segments 5 and 8) and a posterior section (segments 6 and 7) separated by the right hepatic vein; and on the left, a lateral section (segments 2 and 3) and a medial section (segment 4) separated by the left hepatic vein and the falciform ligament (not shown). Each segment can be individually resected. Black dashed lines show the demarcations between sections. Reprinted by permission from Macmillan Publishers Ltd: Nature Reviews Clinical Oncology. Siriwardena AK. et al. Management of colorectal cancer presenting with synchronous liver metastases. Nat Rev Clin Oncol. 2014 Aug;11(8):446- 59. doi: 10.1038/nrclinonc.2014.90. Copyright 2014
3
Arterial vasculature of the liver originates from the descending abdominal aorta at the site of the celiac trunk. Branching off from this site are the left gastric artery, splenic artery, and common hepatic artery. The common hepatic artery then branches into the proper hepatic artery and gastroduodenal artery. The proper hepatic artery further divides into the left and right hepatic arteries, supplying their respective liver lobes [1].
Venous vasculature of the liver originates from multiple sources converging to form the portal vein. The superior mesenteric vein, splenic vein, coronary (left gastric) vein, cystic vein, and the right gastric and pancreaticoduodenal veins. The portal vein, which is valveless, continues toward the liver supported by the hepatoduodenal ligament and divides into the left and right portal veins near the liver hilum [1] (Figure 1.1). Supply of blood to the liver is a mixture with roughly 25% of all blood going into the liver through the hepatic artery and the other 75% through the portal vein. The blood then mixes in the hepatic sinusoids [5].
1.1.2. Microscopic Anatomy of the Human Liver
The human liver, and most other liver-containing species, is subdivided into polygonal structures known as hepatic lobules. In the center of the structure is the central vein and at the peripheral points of the polygon lies the portal area also known as the portal triad.
The portal triad is actually made up of four vascular structures. The proper hepatic artery, hepatic vein, common bile duct, and lymphatic vessels. The reasoning for the use of “triad” is the lymphatic vessels were not identified as part of the structure until after its
4 naming, and the name persisted [5, 6]. Structurally speaking, the hepatic lobule is considered the functional unit of the liver. It’s estimated that the entire human liver contains one million lobules [7]. Blood (arterial and venous) moves from the portal triad region, across the lobule and into the central vein. Bile on the other hand runs opposing.
Secreted from the canaliculi it runs toward the portal triad. The liver sinusoids are formed by columns of hepatocytes spanning from the central vein outward to the portal region. This sinusoid is lined with sinusoidal endothelial cells. Between the sinusoids and the hepatocytes lies a pericapillary space known as the space of Disse, where the flowing blood actually makes contact with the hepatocytes for nutrient/waste exchange.
The resident liver macrophage, Kupffer cells, reside in the sinusoidal space. Also present in the sinusoidal space are the Hepatic Stellate Cells (HSC) [5, 6] (Figure 1.3).
The liver acinus structure is more relevant to hepatic function and is arguably the functional unit of the liver in a metabolic sense. Its irregular shape is defined by an ellipsoidal (sometimes describes as diamond-shaped) arrangement of hepatocytes centered on the line connecting two portal triads, extending outward toward two central veins [8]. The acinus is divided into zones based upon their distance from arterial blood supply. Zone I (periportal) is comprised of hepatocytes receiving the most oxygen.
These hepatocytes are involved in oxidative liver functions like gluconeogenesis, β- oxidation, and cholesterol synthesis. Zone I receives the most oxygenated blood due to hepatocytes proximal location to entering vascular blood and as a consequence of this location in the acinus, least susceptible to hypoxia. Being situated proximal to venous blood supply also means they are first to realize toxins absorbed from the gut through the
5
portal blood supply. Zone II (midzonal) lies between zones I and III. Zone III
(centrilobular) hepatocytes are located at the periphery of the acinus. This is the least oxygenated zone and therefore susceptible to hypoxia. Hepatocytes in this zone highly
express enzymes involved in glycolysis and drug metabolism [5] (Figure 1.2).
6
Figure 1.2. Three-dimensional structure of a liver acinus. Reprinted by permission from Macmillan Publishers Ltd: Nature Reviews Immunology. Adams DH and Eksteen B. Aberrant homing of mucosal T cells and extra-intestinal manifestations of inflammatory bowel disease. Nat Rev Immunol. 2006 Mar;6(3):244-51. doi:10.1038/nri1784. Copyright 2006.
7
1.2. Mouse Liver Anatomy
The mouse liver is anatomically similar to the human liver, although some differences exist. The mouse liver is greater in size as a percentage of total body mass compared to humans. It weights roughly 2 grams, approximately 6% total body weight. The liver spans the entire subdiaphramatic space in the cranial abdomen. The mouse (and rat) liver has four lobes. They are the median, left, right, and caudate lobes. The right lobe contains a transverse septum making it appear as two distinct lobes. The medial lobe sits ventrally and positioned dorsal to this lobe is gallbladder. The left lobe is the largest of the four while the caudate is the smallest [9].
1.3. Cells of the Liver
The liver is comprised primarily of four cell types. The hepatocyte is the parenchymal cell making up roughly 80% of the liver volume. The other principal cells of the liver include the Kupffer cell (the resident macrophage), hepatic stellate cell (HSC), and epithelial cell (biliary and sinusoidal). These primary cells make up the majority of the liver but at least 15 different cell types are known to be found in the normal liver [6]
(Table 1.1).
1.3.1. Hepatocyte
8
The hepatocyte, the parenchymal cells of the liver, are considered the key effector cell as
they are responsible for the majority of functions attributed to the liver [10]. The hepatocytes are organized into plates (or chords) one cell thick [11]. They are polyhedral
in shape with sides facing the perisinusoidal space, bile canalicular space, and adjacent
cells. The perisinusoidal face of the hepatocyte has short microvilli. Many hepatocytes
in mammals are binuclear. The size and number of mitochondria within a hepatocyte
corresponds to its location in the liver acinus. Periportal (zone I) mitochondria are larger
and more numerous than centrilobular hepatocytes (zone III) [10].
1.3.1.1. Lipid Synthesis
The hepatocytes produce the bile components bile acid, cholesterol, phospholipids, and
conjugated bilirubin. In human, the average bile flow produced is about 620 mL/day
[12]. Hepatocytes take up non-esterified fatty acids (NEFA) via fatty acid transport
protein or the fatty acid translocase, CD36. Some NEFAs may enter the hepatocyte
through diffusion. Within the hepatocyte, NEFAs are bound to fatty acid binding protein
or acyl-CoA synthetases and directed to cellular sites of storage or metabolism. The
hepatocyte is also capable of denovo lipogenesis in the cytosol using the precursor acetyl-
CoA (derived from glucose or acetate) to produce fatty acids. In humans, the liver is the
major source of denovo lipogenesis [13] while in rodents the adipose and liver are both
large contributors [14].
1.3.1.2. Lipid Oxidation
9
Fatty acid oxidation in the hepatocyte occurs either through intra-mitochondrial oxidation
or peroxisomal oxidation. Intra-mitochondrial oxidation of fatty acyl-CoA occurs through β-oxidation to acetyl-CoA, which may then be completely oxidized to CO2 in the
tricarboxylic acid cycle [15]. Very long chain fatty acids are oxidized in the peroxisome
by the cytochrome P450 CYP4A ω-oxidation system yielding dicarboxylic acids and
principally serves as mechanism to manage excess fatty acids within the cell. The
peroxisomal oxidation does not have the mitochondrial electron transport chain and
therefore is less efficient in the production of adenosine triphosphate (ATP) [15].
Hepatocytes are also capable of incomplete β-oxidation converting acetyl-CoA to ketone
bodies (acetoacetate and β-hydroxybutyrate) [15, 16].
1.3.1.3. Lipid Transport
Triglyceride export through very low-density lipoprotein (VLDL) synthesis and secretion
is done in hepatocytes through lipid components produced in the smooth endoplasmic
reticulum, transported to the golgi apparatus and formed to glycosylated apoproteins.
Secretory vesicles bud off the golgi apparatus, migrate and fuse to the sinusoidal
membrane, releasing VLDL into the blood [15].
1.3.1.4. Detoxification
The liver is also responsible for uptake, concentration, metabolism, and excretion of
exogenous drugs or toxic compounds in the body. The location of the liver in the
gastrointestinal tract results in this large organ serving as the first to encounter foreign
10 compounds absorbed from the gut. In pharmacokinetics, the “first pass effect” describes the significant reduction in oral drug concentration before it reaches the systemic blood.
Drug delivery mechanisms must take this loss into account and other routes of administration are utilized to bypass the liver and avoid this phenomenon [17]. Hepatic drug metabolism occurs in three, distinct phases.
Phase I of hepatocyte-mediated xenobiotic metabolism includes bioactivation and toxification reactions. The exogenous drugs or toxins are biotransformed by the addition of functional hydroxyl, carboxyl, amino, or thiol groups. This biochemical reaction allows for the targeted compound to be more hydrophilic. These chemical reactions are executed by the cytochrome P450 superfamily. There are approximately 60 genes coding for CYP enzymes in the human genome [18]. Most of these enzymes are located on the cytoplasmic side of the endoplasmic reticulum or mitochondria. Free radicals and toxic electrophilic compounds can be a byproduct of phase I reactions.
Phase II involves conjugation with highly polar endogenous compounds, further increasing the size and hydrophilic nature of the target. The completion of this step usually results in an excretable and nontoxic compound. This phase occurs within the hepatocyte cytoplasm through diphosphate (UDP)-glucuronyl transferases, sulfotransferases, and glutathione S-transferases [19]. Phase III constitutes cellular excretion into hepatic canaliculi or bile and ultimately clearance through the kidneys or bile [20, 21].
11
Cells found in the Liver
Hepatocytes (60% of cells and 80% of volume of parenchyma) Biliary epithelia Endothelia Sinusoids Blood vessels (arteries and veins) Lymphatics Kupffer cells Hepatic stellate cells (also known as Ito or fat-storing cells) Lymphocytes (Pit cells) Progenitor cells Oval cell—rodent models Hepatoblasts—humans Fibroblasts Smooth muscle cells (blood vessels) Mesothelia Nerves (unmyelinated) Neuroendocrine cells Hematopoeitic cells Blood (erythrocytes, leukocytes, etc.)
Table 1.1. Cells found in the liver. Reprinted with permission from: Malarkey DE, Johnson K, Ryan L, Boorman G, Maronpot RR. New insights into functional aspects of liver morphology. Toxicol Pathol. 2005;33:27-34. doi: 10.1080/01926230590881826
12
1.3.2. Kupffer Cell
The resident macrophage of the liver, the Kupffer cell, resides predominately in the
lumen of the sinusoids. These cells were first identified in 1876 by Karl Wilhelm von
Kupffer who described them as "Sternzellen" (star cells or hepatic stellate cells) incorrectly defining them as endothelial cells [22]. In 1974, E. Wisse demonstrated definitively that endothelial cells and Kupffer cells were two distinct cells [23]. Their origin is debatable but the mononuclear phagocytotic system proposes that monocytes originate from precursor cells in the bone marrow and translocate to the liver, only then differentiating into Kupffer cells [24]. This implies they are not self-renewing and their population must be restored from the bone-marrow. Others argue they are self-renewing, dividing as fully differentiated macrophages or perhaps derived from a localized progenitor cell [25]. Supporting this hypothesis, Kupffer cell populations have been
shown to divide after hepatic partial hepatectomy [26]. Both hypothesis could
technically be correct, Klein et al. demonstrated two distinct Kupffer cell populations in
the liver. One derived from bone-marrow and recruited to the liver by inflammatory
stimuli, the other a fixed macrophage with long-lifespan but ultimately derived from
bone-marrow progenitors [27].
Kupffer cells account for roughly 15% of all liver cells, 30% of the sinusoidal cells [28],
and make up more than 50% of all resident tissue macrophages of the body [22]. They
are described as star-shaped or amoeboid-shaped and adhere to the surface of sinusoidal
endothelial cells. They contain oval nuclei, lysosomes, and multiple vesicles. They are
13
capable of phagocytizing large particles including erythrocytes and bacteria [22] and are
twice as numerous in the periportal region of the hepatic lobule sinusoids as the
centrilobular region [29]. Kupffer cell size and function are related to the position at
which they occupy the sinusoid. Cells near the periportal are larger and more phagocytic
but produce less superoxide anion free radicals than their centrilobular counterparts [30].
Monocytes leave the bone marrow within 24 hours of their production in mice, and
circulate the bloodstream with a half-life of around 17.4 hours. They may migrate into
the liver tissue and become Kupffer cells with an estimated lifespan of 3.8 days [22]. In
human transplanted livers, the donors Kupffer cell population has been shown to remain for up to a year [22]. Their local population is regulated partially though apoptosis.
Interleukin-1 (IL-1), IL-4, TNF-α, and IFN-γ cytokines have demonstrated regulation of
apoptotic cell death in Kupffer cells in vitro [31, 32].
The parenchymal cells of the liver play a role in detoxification of portal blood while the
Kupffer cells intercept infecting microorganisms infiltrating through the gastrointestinal
tract. They are critical components of the immune response through their production of
inflammatory compounds as a host defense. These include proinflammatory cytokines,
superoxide, nitric oxide, eicosanoids, chemokines, and lysosomal and proteolytic
enzymes [33].
1.3.3. Hepatic Stellate Cell
14
The hepatic stellate cell (HSC) is a found in the perisinusoidal space (space of Disse) in
the liver. They have also been called perisinusoidal cells, Ito cells, lipocytes, interstitial
cells, and fat-storing cells. They constitute about 3% to 8% of the total liver cells and are
mostly located in the centrilobular (zone III) of the hepatic acinus. HSC in the periportal
region (zone I) tend to be smaller while midzonal (zone 2) HSCs tend to have the largest lipid droplets [34]. Just like the resident macrophage of the liver, HSC discovery is credited to Karl Wilhelm von Kupffer who described star-shaped cells after gold and chloride staining of hepatic lobules. He also observed phagocytosis in what he thought were the same cells and called them “specialized endothelial cells of the sinusoids”. In
1952, Toshio Ito clarified that the HSCs are cells abundant in fat droplets while the phagocytizing cells were another cell type (Kupffer cell).
Hepatic stellate cells contain large stores of vitamin A in the form of retinyl esters. They hold nearly 80% of the retinoid stores in the murine body [35]. The retinyl esters are contained within the HSC in large lipid droplets. The lipid droplet are comprised of roughly 42% retinyl ester, 28% triglyceride, 13% cholesterol, and 4% phospholipid [36].
The retinyl esters in the HSC are derived from the hepatocytes. In the intestine, retinol is esterified and packaged into chylomicrons. The chylomicrons enter the liver through the lymph and are primarily taken up by hepatocytes. Hepatocytes are then able to transfer retinyl esters to HSC via retinol binding protein and receptor mediated endocytosis. The retinyl esters may be hydrolyzed for mobilization from the HSC lipid droplet into circulation in times of retinoid-insufficiency [37].
15
Hepatic stellate cells play a significant role in the production of extracellular matrix proteins (ECM) such as collagen and α-smooth muscle actin (α-SMA). They are therefore implicated as key regulators of fibrogenesis within the liver [38]. Upon activation, HSC proliferate and their endoplasmic reticulum increase in volume. They lose retinoid stores, and their overall structure changes from star-shaped to a more fibroblast-like form. Triggers for activation include inflammatory mediators, growth factors, reactive oxygen species, as well as apoptotic bodies from dying hepatocytes.
HSC isolated, purified, and subsequently cultured have been observed to undergo transition to a fibrogenic phenotype signifying HSC are able to stimulate their own
“activation” [39]. This spontaneous activation in vitro may be suppressed if the HSC are
cultured on a basement matrix gel. These cells consequently produce less ECM [40].
1.3.4. Sinusoidal Endothelial Cell
Liver sinusoidal endothelial cells (LSEC) form the single cell, fenestrated lining of the
hepatic sinusoid separating hepatocytes from the sinusoidal lumen. They lack a basement
membrane while forming a protective barrier to the hepatic parenchyma from pathogenic
agents in the blood supply. LSEC are active in receptor-mediated endocytosis mainly
using the mannose receptor, scavenger receptor, and the Fcγ receptor Ilb2 [41]. In
addition to receptor-mediated endocytosis, LSEC are able to phagocytose particles and
present antigen to lymphocytes [42]. They have been shown to express CD4, CD11b,
and CD11c, which are all markers more commonly related to T cells, myeloid cells, and
dendritic cells [43].
16
Cirrhosis of the liver, no matter the cause, is characterized by the defenestration of the sinusoidal endothelium as well as the presence of a sub-endothelial basement membrane.
Activation of HSC has been shown to be prevented by LSEC. Preceding hepatic fibrosis,
LSEC have an altered phenotype known as “capillarization” in which they lose the ability to prevent HSC activation [44]. LSEC serve as a selective barrier between the blood and the parenchyma, mediating the exchange of particles traveling both directions. They also serve as scavengers, clearing waste products [45].
17
CHAPTER 2
DISEASES OF THE LIVER
2.1. Overview
The historical stereotype for the source of liver disease has been the abuse of alcohol and
drugs. There are over 100 known forms of liver disease with an assortment of factors contributing to their pathogenesis. They are widespread across the population, affecting young infants to the elderly. Primary contributors to liver disease include viral infection, obesity, alcohol, drugs, toxin exposure, autoimmune disorders, genetics, and cancer [46].
2.2. Alcoholic Liver Disease
Excessive alcohol intake is associated with the development of alcoholic fatty liver disease with progression to alcoholic hepatitis and cirrhosis. Abuse of alcohol (>30 grams/day) is associated with an increased risk for alcohol-attributable cirrhosis. Genetic predisposition appears to play a strong role in susceptibility to alcohol-attributable cirrhosis as only 1% of individuals who drink 30 to 60 grams alcohol/day develop cirrhosis [47]. Even with a small percentage at risk for cirrhosis, the widespread use of alcohol caused an estimated 493,300 deaths from alcohol-attributable cirrhosis worldwide
in 2010 [48].
18
Hepatic steatosis (the accumulation of fat within the liver) is seen in 90% of heavy alcohol drinkers [49]. A third of individuals who continue heavy alcohol intake will progress to steatohepatitis (inflammation of the fatty liver) [50]. Continued alcohol intake results in cirrhosis of the liver in roughly 16% of those with steatohepatitis [51].
Cessation of alcohol intake has been shown to improve long-term prognosis in individuals with alcoholic liver disease including limiting the progression to end-stage cirrhosis [52] as well as an improved survival rate [53].
The metabolism of ethanol is covered in detail in chapter 3. The primary metabolic pathway for ethanol metabolism in hepatocytes yields acetaldehyde and reduced nicotinamide adenine dinucleotide (NAD) [54]. This metabolic reaction leads to an altered redox state within hepatocytes impairing β-oxidation and the tricarboxylic acid cycle leading to hepatic steatosis [55]. Of those patients who progress to alcoholic steatohepatitis the primary factors contributing to the progression are neutrophil infiltration and activation [56], hepatic endotoxin (lipopolysaccharide) intrusion via the gut [57], release of proinflammatory cytokines (TNF-α [58], TRAIL [59]), and oxidative stress [60]. Activated HSC, a source of extracellular matrix proteins, are associated with human alcoholic fibrosis [61].
2.3. Nonalcoholic Fatty Liver Disease
Nonalcoholic fatty liver disease (NAFLD) is an form of chronic liver disease characterized by accumulation of fat within the liver independent of any other cause for
19
steatosis (i.e. alcohol consumption). The initial stage of NAFLD is simple steatosis in the absence of hepatic inflammation but may progress to nonalcoholic steatohepatitis
(NASH). NASH is a stage of hepatic inflammation that is histologically
indistinguishable from alcoholic steatohepatitis [62]. NASH may progress to cirrhosis in
up to 20% of patients [63]. Obesity, hypertriglyceridemia, elevated serum alanine
aminotransferase (ALT), and type II diabetes are associated with NAFLD [64, 65].
NAFLD incidence is most common in Western industrialized nations as it’s associated
with the various components of metabolic syndrome, a common occurrence in these
nations [66]. Most studies conducted within the Unites States of America using clinical
diagnostic modalities associated with NAFLD report 10% to 35% prevalence in the
general population [67]. The prevalence of NASH is reported as 3% to 5% of the general
population [67].
The proposed mechanism for the initial stage of NAFLD, steatosis, is through excessive
deposition of triglycerides in the hepatocytes transported from the adipose tissue [68]. A
decrease in hepatic export of triglyceride [69] as well as a diminished capacity for β-
oxidation within the hepatocytes are also possible contributors [70]. Peripheral insulin
resistance may contribute to steatosis through increased lipolysis, triglyceride synthesis,
and the hepatic uptake of free fatty acids [64, 71]. Perisinusoidal fibrosis may develop as
the severity of NAFLD progresses. The chronic inflammatory state in NASH contributes
to the activation of hepatic HSC, inducing fibrosis [72]. Increased risk of hepatocellular
20 carcinoma is associated with NAFLD-mediated cirrhosis [73]. Depending on the presence of progressive risk factors (age, BMI, Type II diabetes, visceral adiposity) and the stage at which NAFLD is diagnosed, the prognosis is variable. There is an overall increase in mortality in patients with advanced fibrosis when compared to non-fibrotic
NAFLD patients [74].
21
CHAPTER 3
MURINE MODELS OF LIVER INJURY
3.1. Xenobiotic
3.1.1. Carbon Tetrachloride
The hepatotoxin carbon tetrachloride (CCl4) is commonly used in both acute and chronic
models of liver injury. It induces a centrilobular necrosis followed by hepatic fibrosis
[75]. Early work using CCl4 to induce liver injury suggested that the toxin damaged the
hepatocyte mitochondria. Perturbations in normal lipid oxidation pathways lead to an
interruption of energy substrate availability. Later work demonstrated that triglyceride
accumulation and endoplasmic reticulum degeneration occurred before mitochondrial
degeneration, therefore disproving the previous hypothesis [76].
It is now known that CCl4 facilitates its hepatotoxic effects through metabolic activation by cytochrome P450, family 2, subfamily E, polypeptide 1 (CYP2E1) into the reactive
intermediate, trichloromethyl radical, that covalently binds lipid membranes leading to
loss of membrane integrity by peroxidation. Increased cellular permeability results in
loss of calcium homeostasis which leads to the induction of calcium dependent derivative
enzymes and cellular necrosis [77]. The liver specificity of CCl4 is attributed to the state
at which it becomes metabolically active. CYP2E1 is highly expressed in hepatocytes
with only limited expression elsewhere in the body [78].
22
In mice, 0.5 to 2.0 mL/kg body weight of CCl4 (diluted in oil) is administered into the
intraperitoneal (IP) cavity 2 to 3 times a week for 4 to 6 weeks to induce liver fibrosis.
Susceptibility is largely dictated by genetic background of the mouse strain, with
BALB/c mice showing most sensitivity, while FVB/N are less sensitive. C57BL/6 mice
are regarded as having intermediate sensitivity [79]. Chronic administration of CCl4
results in liver fibrogenesis similar to the human condition of liver fibrosis with
inflammation, fibrosis, and tissue regeneration [79].
Fibrosis development is divided into three stages. The first is acute injury, followed by fiber formation, and last is advanced fibrosis. A single injection of CCl4 is a reproducible
model of acute injury to the liver. At this early stage, CCl4 induces an inflammatory
response with the activation of liver resident Kupffer cells and the secretion of cytokines
and chemokines leading to activation/recruitment of other proinflammatory immune cells
such as monocytes, neutrophils, and lymphocytes [79].
The second stage, in which the liver has the observable accumulation of extracellular
matrix fibers occurs 2 to 3 weeks after chronic administration of CCl4. The final stage,
where bridging fibrosis is present, occurs 4 to 6 weeks after CCl4 administration [79].
Mice are able to recover from CCl4-induced fibrogenesis if the toxic agent is removed
and they are allowed to recover for at least one month [80].
3.1.2. Concanavalin A
23
Concanavalin A (Con-A) is a plant lectin used to induce a T-cell dependent acute liver
injury resulting in hepatitis that closely mimics human autoimmune hepatitis. Lectins are proteins that bind sugar residues on cell surfaces and potentiate a mitogenic stimulation of T lymphocytes [81]. They are abundant in nature, found in plants, animals as well as
bacteria [82].
The current hypothesis for T-cell mediated hepatic inflammation is that activated T-cells
stimulate Kupffer cells to produce the cytokine TNF-α, thereby damaging surrounding
hepatic sinusoidal endothelial cells. FasL-expressing lymphocytes are recruited to the
area with TNF-α and FasL inducing hepatocyte damage [83].
Con-A dosages between 10 to 20 µg/g body weight induces liver injury and this injury is
T-cell dependent as T-cell deficient mice have shown resistance [81]. CD4+ T cells
permeate the liver within 6 to 8 hours after Con-A treatment. Liver damage is shown to
occur before CD4+ T cell infiltration indicating that local T cells producing IFN-γ as well
as Kupffer cells producing TNF-α initiate the liver damage [84]. Pretreatment of
Interleukin-6 (IL-6) and/or IL-10 prevents hepatitis through inhibition of IFN-γ and TNF-
α mediated damage [85, 86]. Serum transaminases begin to decline 24 hours after Con-A administration and C57BL/6 mice have shown resistant to secondary doses of Con-A up to 8 days after the original application [83]. The features of Con-A induced hepatitis are similar to those of human autoimmune hepatitis and currently serves as a model of T-cell mediated hepatic injury seen in the human condition.
24
3.1.3. Ethanol
Ethanol consumption among humans is a leading cause of chronic liver disease, therefore
chronic administration of ethanol to rodents has become a model for the induction of liver disease. Multiple ethanol administration paradigms have been used with various advantages and disadvantages. A common method used in mouse administration models consists of ad libitum feeding using a liquid ethanol containing diet (Lieber-DeCarli high- fat ethanol liquid diet) with 5% (v/v) ethanol (36% ethanol-derived calories) for 4 to 6 weeks with isocaloric, non-ethanol fed mice serving as controls [87, 88]. This leads to a mild elevation of plasma ALT and hepatic steatosis with limited inflammation and no hepatic fibrosis [89]. Intragastric feeding is shown to induce a slightly more severe response, with severe steatosis, some inflammation, as well as mild fibrosis [90]. A chronic-binge feeding model (a.k.a. the National Institute on Alcohol Abuse and
Alcoholism (NIAAA) model), which closely aligns with human alcohol consuming
behavior, consists of 10 day ad libitum feeding at 5% (v/v) ethanol containing liquid diet
followed by a one-time ethanol intragastric gavage (5 g kg−1) resulting in increased serum
transaminases, steatosis, but no liver fibrosis [89]. The short experimental length of
treatment is a strong advantage to this model.
Upon ingestion, ethanol concentrations increase in the portal/hepatic blood vessels.
Ethanol is converted to acetaldehyde by alcohol dehydrogenase and CYP2E1 and then to
acetate by aldehyde dehydrogenase (ALDH) [91] (Figure 3.1). Gastric metabolism of
25 ethanol by alcohol dehydrogenase has been shown but ultimately its contribution to overall ethanol metabolism appears minute [92]. When hepatocytes metabolize ethanol to acetate, NADH is produced, increasing the NADH/NAD+ ratio within the metabolizing cell. This increased ration inhibits β-oxidation of fatty acids leading to an accumulation of lipids within the liver. Lipid accumulation in the liver can occur as a large lipid droplet (macrovesicular) or as multiple smaller lipid droplets (microvesicular) within the hepatocyte [93, 94].
Ethanol consumption is also closely linked to increases in bacterial lipopolysaccharide
(LPS) concentrations in serum. This is seen in both human subjects as well as acute and chronic animal models of ethanol consumption [57, 95, 96]. This increased serum LPS is ascribed to an increase in intestinal permeability. Ethanol consumptions disrupts the apical junction proteins between enterocytes of the intestine, leading to gut “leakage” and
LPS infiltration into portal circulation [97]. LPS is capable of triggering an immune response within the liver, resulting in the release of proinflammatory compounds and the development of hepatic inflammation.
26
Figure 3.1. General scheme of ethanol oxidation General scheme for alcohol oxidation. Alcohol is oxidized by alcohol and aldehyde dehydrogenases eventually to acetyl CoA. Depending on the nutritional, hormonal, energetic status, the acetyl CoA is converted to the indicated products. Reprinted from Alcohol Metabolism. Cederbaum A.I. Alcohol metabolism. Clin. Liver Dis. 2012;16:667–685, Copyright 2012, with permission from Elsevier.
27
3.2. Surgical
3.2.1. Bile Duct Ligation
The liver produces bile acids that are then stored in high concentrations in the gallbladder and excreted (as bile) through the common bile duct into the intestinal lumen for the
purpose of emulsification of ingested lipids, cholesterol, and fat-soluble vitamins
facilitating their absorption through the small intestine [98]. Bile acids have been shown
to play a role in the regulation of lipid, glucose, and drug metabolism. They are also
intracellular ligands for activation of nuclear receptors farnesoid X receptor, pregnane X
receptor, and vitamin D receptor as well as membrane bound G protein–coupled
receptors such as G protein–coupled bile acid receptor (TGR5 and Gpbar-1) [99].
Surgical ligation of the common bile duct induces obstructive cholestatic injury to the
liver capable of inducing hepatic fibrosis within 21 to 28 days. Complete obstruction of
the common bile duct results in biliary epithelial cell proliferation, myofibroblastic
differentiation of portal fibroblasts, and the excessive deposition of ECM within the
portal tracts [100]. Total bile duct ligation results in extensive liver necrosis within 3
days, but not liver cirrhosis [101].
Partial surgical ligation of the bile duct in mice has been performed in which the
gallbladder is removed to prevent cholecystitis and a 0.2mm needle is inserted into the
bile duct followed by ductal ligation around the needle. The needle is then removed
resulting in partial bile duct occlusion. This technique more closely aligns with human
28
cholestasis as it results in increased pressure in the biliary system and liver damage with
survivability. Mice experience the larger degree of liver injury 3 to 5 days after surgery
and are free of histological anomalies by day 14 [101]. Total bile duct ligation remains
the best surgical option for increased portal pressure resulting in hepatic damage and
fibrosis.
3.3. Dietary
3.3.1. High-Fat Diet
Nonalcoholic fatty liver disease (NAFLD) can result from excess fat intake, exacerbated
by a sedentary lifestyle. NAFLD is associated with components of metabolic syndrome
including obesity [102], insulin resistance [103], hyperlipidemia [104], and hypertension
[105]. Hepatic steatosis in humans, an early stage of NAFLD, is characterized by
macrovesicular triglyceride accumulation within the hepatocytes in the absence of
inflammation [106] and its generally considered reversible. As NAFLD progresses,
inflammatory foci appear with the ballooning degeneration of hepatocytes, presence of
Mallory bodies, and fibrosis [107]. Animal models replicating the human condition of
NAFLD at its progressive stages allows for the investigation of underlying NAFLD pathologies and potential therapeutic modalities.
A “two-hit” mechanism for NAFLD progression was proposed by Day et al. with the first
“hit” resulting in liver steatosis, the accumulation of fat in the liver. The second “hit” leads to progression of the disease to hepatic inflammation and fibrosis [108]. The
29
steatosis is elicited by excess fat intake. The excess lipid within the hepatocytes leads to
“second hit” oxidative stress, release of proinflammatory cytokines and free fatty acids,
and cellular apoptosis causing nonalcoholic steatohepatitis (NASH) [109]. Lieber et al.
report NASH may be induced using a high-fat diet (HFD) model in rats consisting of
71% energy from fat, 11% from carbohydrates, and 18% from protein ad libitum for 3 weeks. This causes steatosis, increased hepatic TNF-α, and extracellular collagen type 1 deposition as well as elevated CYP2E1 mRNA in liver tissue [110]. The HFD model produces variable degrees of NASH based upon rodent species, dietary fat content, fat origin, and duration of treatment [111]. Inbred mouse strain BALB/c accumulate more hepatic triglyceride than inbred strain C57BL/6J mice [112].
3.3.2. Methionine and Choline Deficient Diet
The methionine and choline deficient diets (MCD) is a diet deficient in the essential amino acid methionine and essential nutrient choline. These compounds are necessary for hepatic β-oxidation and VLDL packing and secretion [109]. Lack of these essential compounds leads to impaired lipid secretion from the liver resulting in excess hepatic lipid deposition, oxidative stress, increased serum ALT and the adiponectin, as well as the induction of hepatic CYP2E1 [113, 114]. Steatosis is present after 10 days of treatment [115] and perisinusoidal fibrosis evident after 8 to 10 weeks [113]. Similar to the HFD-induced model, the severity of NASH development from the MCD diet is also dependent on animal species and strain as well as diet duration. Inbred mouse strain
C57BL/6 are highly sensitive to the MCD diet, developing hepatic inflammation and necrosis [111]. This diet does not align with human NASH in some respects, as mice fed
30 a MCD diet lose weight, display low fasting blood glucose, and maintain normal peripheral insulin sensitivity [116].
31
CHAPTER 4
LIVER REGENERATION
4.1. Overview
The human digestive system is arranged in a manner so that the liver stands between
blood from the digestive organs and systemic circulation. The portal vein, feeding into
the liver, carries blood from the small and most of the large intestine, the spleen, and the
pancreas. Anything that makes its way into the blood stream from the intestine
(macronutrients, micronutrients, chemical toxins, bacteria, etc.) must go through the liver
before traveling elsewhere. This arrangement is critical to the many essential liver
functions related to digestion, nutrient storage, metabolism and host immunity [117].
These responsibilities make the liver absolutely indispensable while at the same time
highly susceptible to injury. To maintain this vital organ and its numerous biological
functions, the liver maintains an evolutionarily conserved capacity to regenerate [118].
Liver regeneration after loss of liver mass is seen in all vertebrate organisms [117].
The mystery behind the liver and its remarkable ability to regenerate has puzzled humanity since antiquity. Prometheus, a titan in Greek Mythology, is said to have stolen the privilege of fire from the Gods of Mount Olympus and given it to mankind for the
benefit of humanity. For this betrayal, Zeus banished Prometheus to Mount Caucasus,
chaining him to a rock. As further punishment, Prometheus must endure his liver being
eaten by an eagle though due to his immortality the lost liver tissue is regenerated every
night only for the process to begin again the next day [119]. The myth of Prometheus
32 colorfully illustrates mankind’s long held understanding of the importance of the liver and its extraordinary regenerative ability (Figure 4.1).
33
Figure 4.1. Prometheus. Prometheus. Cornelis Bloemaert, Dutch, c. 1603 – 1692, Date unknown. Philadelphia Museum of Art: The Muriel and Philip Berman Gift, acquired from the John S. Phillips bequest of 1876 to the Pennsylvania Academy of the Fine Arts, with funds contributed by Muriel and Philip Berman, gifts (by exchange) of Lisa Norris Elkins, Bryant W. Langston, Samuel S. White 3rd and Vera White, with additional funds contributed by John Howard McFadden, Jr., Thomas Skelton Harrison, and the Philip H. and A.S.W. Rosenbach Foundation, 1985. Photograph and Digital Image © Philadelphia Museum of Art.
34
To investigate the mechanism of liver regeneration the process may be induced by many
methods including the administration of chemical toxins to the liver leading to cell death,
clearance, and subsequent space for new cells. The use of CCl4 [120], ethanol [121],
acetaminophen [122], and bile duct ligation [123] have all been documented. The most
common method though is surgical removal of roughly two-thirds of the liver mass, a
process known as two-thirds partial hepatectomy (2/3PH) [124]. After 2/3PH the
remaining liver tissue undergoes hyperplasia, growing to the original liver mass within
approximately two weeks, in rodents [125].
4.2. Two-Thirds Partial Hepatectomy
The first manuscript describing 2/3PH in rats was published in 1931 by G.M. Higgins and
R.M. Anderson [124]. Interestingly, they use the term “liver restoration” instead of “liver regeneration” which is perhaps a more accurate description of this hepatoproliferative effect since the removed lobes do not regrow but the remaining lobes swell to make up
the missing hepatic mass [124, 126]. A benefit of this technique is that resection of lobes
of the liver does not cause damage to the remaining lobes [126]. As the technique
became more popular for investigating liver regeneration various practices deviating from
the described technique were adapted and is likely a significant reason for discrepancies
in data between investigators. Wustefeld et al. provide an example in their 2000
publication describing in detail their methodology for 2/3PH. A small incision just below the xiphoid process of the sternum is made to gain access to the liver. The liver is then pulled through this opening and the actual resectioning of the liver is performed outside the body. The remaining liver lobes are then reinserted back through the opening and the
35 incision closed [127]. This single small incision is an obvious advantage of this procedure although this method has been criticized [128] as imprecise due to limited visibility, which may lead to complications.
C. Mitchell and H. Willenbring noticed the inconsistency in 2/3PH techniques and published an extremely detailed process that produces consistent results in mice. They first call attention to the choice of anesthesia used during the procedure. Avertin, or tribromoethanol, is a common anesthetic that is injection into the intraperitoneal cavity at concentrations of 1.25%. Unfortunately this compound is highly hepatotoxic [128], as is another commonly used an intraperitoneal anesthetic cocktail of Ketamine/xylazine usually administered at 100mg/kg body weight and 10mg/kg body weight respectively
[129]. The inhalant anesthetic isoflurane (2-chloro-2-(difluoromethoxy)-1,1,1-trifluoro- ethane) commonly used in veterinary medicine has limited hepatotoxic effects and is currently the anesthetic of choice for survival surgeries, including 2/3PH [128]. Next, they indicate that all surgeries should be conducted at the same time of day as mitotic activity is influenced by circadian rhythm [130]. Mice should not be fasted, as fasting induced the accumulation of lipids in the liver which may impair liver regenerative capacity [131, 132]. Animals should be of the same gender and age (8-14 weeks) [128].
They document a very comprehensive surgical procedure which leads to resection of the left lateral and median lobes of the liver (Figure 4.2). This procedure should be accomplished in 15-20 minutes after which the animal is placed on a warming pad for recovery [128]. Anticipated results are also given which indicate (through BrdU
36 incorporation) that in C57BL/6 mice peak in hepatocyte mitosis between 36-48 hours after surgery [128] (Figure 4.3).
37
Figure 4.2. Schematic drawings of mouse liver anatomy and positioning of silk threads for knots. (a) The thread for the first knot should be positioned between the caudate and the left lateral lobes at the base of the latter. (b) The second knot should be tied within the dashed area, above the gall bladder but not too close to the suprahepatic vena cava. The tip of the right lobe can be used as a reference point for placing the knot. Reprinted by permission from Macmillan Publishers Ltd: Nature Protocols. Mitchell C. and Willenbring H. A reproducible and well-tolerated method for 2/3 partial hepatectomy in mice. Nat Protoc. 2008;3(7):1167-70. doi: 10.1038/nprot.2008.80. Copyright 2008.
38
Figure 4.3. Kinetics of BrdU incorporation after 2/3 PH. The graph represents data pooled from three different experiments performed on 8- to 12–week-old wild-type mice of the C57Bl/6 strain (n≥10 mice per time point). Reprinted by permission from Macmillan Publishers Ltd: Nature Protocols. Mitchell C. and Willenbring H. A reproducible and well-tolerated method for 2/3 partial hepatectomy in mice. Nat Protoc. 2008;3(7):1167-70. doi: 10.1038/nprot.2008.80. Copyright 2008.
39
Modern investigative techniques have allowed for further investigation of liver
regeneration after partial hepatectomy. Historically, it was assumed that the hepatocytes
of the liver would undergo mitotic cell division to regain tissue mass lost after surgery. It
was estimated that one-to-two rounds of cell division per hepatocyte would be sufficient
to restore liver mass [117]. This hypothesis was recently disproved using hydrodynamic
tail vein injection delivery of plasmids into hepatocytes that revealed 40% of hepatocytes remaining after 70% partial hepatectomy did not divide, even though the liver regenerated to original liver mass within the typical 2 week period [126]. Enlargement of the hepatocytes was found to be the other contributing factor to regeneration of mass. The hepatocyte hypertrophy peaks one day after 70% partial hepatectomy [126] indicating this is an early cellular response and this enlargement is augmented if hepatocyte division is inhibited via dexamethasone [133].
Regeneration of the liver still largely depends on proliferation of hepatocytes. Under normal conditions hepatocytes are non-mitotic (G0 phase of mitosis). They are not
terminally differentiated but considered in a state of proliferative quiescence [134]. The
signal to proliferate comes from non-parenchymal cells through cytokine signaling of
TNF-α and IL-6. The Kupffer Cell is a large contributor of cytokines during the
initiation phase (or priming phase) of liver regeneration when stimulated by LPS [135],
C3a, C5a [136], and ICAM [137]. ICAM-1 deficient mice have impaired hepatocyte
regeneration due to ICAM-1’s role in stimulating TNF-α and IL-6 derived from Kupffer
cells [134]. These cytokines (TNF-α [138], Interleukin-18 (IL-18) [139], and IL-1β
[140]) are capable of binding to their complementary receptors on the same cell that
40 secreted them (autocrine) as well as adjacent cells (paracrine). Their receptor signaling cascades primarily result in transcription factor NF-kB [141] translocation to the nucleus inducing increased transcription of their own genes (positive feedback) to amplify the inflammatory response. The cytokine IL-6 also has a NF-kB response element regulating its transcription and is secreting upon NF-kB signaling activation [141]. The communication between Kupffer cells and hepatocytes at the early time points of liver regeneration is facilitated by inflammatory cytokines TNFα, IL-β, and IL-18 resulting in increased IL-6 secretion and “priming” of hepatocytes for division [142, 143].
During initiation of hepatocyte mitosis, the hepatocytes are primed for replication. This is done by initiation factors IL-6 and TNF-α. TNF-α is an early initiator while IL-6 seems to be the principal initiator [144]. IL-6 binds to its receptor complex and signals
STAT3 activation driving hepatocyte mitosis by initiating G0/G1 cell cycle phase transition [134]. This initiation stage is characterized by the release of immediate early- phase genes within 2 hours after partial hepatectomy that drives hepatocyte proliferation.
These immediate early-phase genes include c-fos, c-jun, and c-myc [145].
c-myc expression is strongly induced in hepatocytes almost immediately after partial hepatectomy [146]. c-myc heterodimerizes with its partner MAX and together associate with enhancer-box sequences [147]. Eukaryotic translation initiation factor 4E (eIF4E) and Eukaryotic Initiation Factor 2 alpha are facilitators of translation and are direct transcriptional targets of c-myc [148]. c-myc also drives cell cycle progression through
41
transcriptional regulation of Cyclin D2 and Cyclin Dependent Kinase 4 [149]. Mice with
targeted knockdown of c-myc in hepatocytes expressing just 10% normal c-Myc levels
show normal regeneration of liver mass after partial hepatectomy. Although these mice
seem to have regained liver mass within the normal 7 day period following partial hepatectomy, the liver architecture at the hepatic plate was disorganized with the presence of abnormally large hepatocytes [150].
c-jun may homodimerize with itself or heterodimerize with c-fos to form the transcription factor Activator Protein 1 (AP-1) [151]. The c-jun homodimer may bind to an AP-1 response element but the c-jun/c-fos AP-1 transcription factor has an affinity 25 times stronger for the same AP-1 response elements [151]. c-jun is required for hepatocyte survival and proliferation. Mice lacking c-jun are embryonic lethal at mid-gestation
[152]. Mice with conditional inactivation of c-jun after birth show impaired hepatocyte proliferation and increased hepatocyte death after partial hepatectomy, impairing their ability to restore liver mass [153]. Mice with hepatocyte specific deletion of c-jun are viable but exhibit decreased hepatocyte proliferation and decreased body weight [153].
c-jun is required for the progression of hepatocytes through G1 phase of the cell cycle
through its direct transcriptional control of Cyclin D1 [154]. c-fos-/- and c-jun-/- mice
fibroblasts both fail G1 phase progression and have decreased transcription of Cyclin D1
[155]. Cyclin D1 is a regulator of cyclin dependent kinases 4 and 6. Activation of these
kinases causes release of a RB-dependent cell cycle inhibitor that regulates cell cycle
42
progression. Cyclin dependent kinases 4 or 6 phosphorylates tumor suppressor protein
RB resulting in the release of E2F transcription factors that promote E2F responsive genes essential for DNA synthesis [156]. The promotion of these early response genes (c- fos, c-jun, c-myc) by IL-6 through STAT3 culminates in a strong survival and proliferation response within the target hepatocytes.
In mice, the proliferation phase begins between 24-48 hours after partial hepatectomy, with hepatocyte division beginning in zone 1 (periportal) and moving to zone 3
(centrilobular) [134]. Proliferation continues under the control of cyclins and cyclin-
dependent kinases and various growth factors are released including hepatocyte growth
factor and transforming growth factor alpha [134]. Termination of hepatocyte
proliferation, to prevent an overshoot of original hepatic mass is governed by suppressors of hepatocyte proliferation, transforming growth factor beta 1 and Activin which are expressed about 5 hours after partial hepatectomy and remains expressed through the termination phase [157]. Hepatocyte specific transforming growth factor beta receptor I deficient mice exhibit normal termination of liver regeneration after partial hepatectomy unless the transforming forming growth factor beta receptor I deficiency is combined with inactivation of the Activin receptor [158]. It’s suggested that these suppressors keep
hepatocytes under constant antagonism and that proliferation is only started when their suppression is overcome by growth factors. The molecular mechanism behind termination of liver regeneration is still largely not understood. A phenomenon has been documented in which the liver of a small animal (small dog to large dog) is transplanted
43 into the body of a larger animal after which the liver will continue to grow to the appropriate size for the large animal [159].
Partial hepatectomy is an excellent technique for the study of liver regeneration but in order to translate mechanistic discoveries to novel treatment methods for diseases of the liver (NAFLD, ALD, and Viral Hepatitis) investigations using other models of liver injury are also necessary. It is well established that IL-6 plays a critical role in recovery of the liver after hepatectomy but it has also been shown to be hepatoprotective in other animal models of liver injury [134].
4.3. Other Models of Liver Regeneration
4.3.1. Concanavalin A
Concanavalin A (Con-A) induced hepatitis as a consequence of T-cell activation is a model resembling human viral hepatitis. See Chapter 3 on Murine Models of Liver
Injury for more information on Con-A induced hepatitis. T-cell activation leads to secretion of TNF-α which is followed by secretion of IL-1 (no distinction was made between IL-1α and Interleukin-1β (IL-1β)) and IL-6 [160]. Con-A induced hepatitis is T- cell dependent as pretreatment of mice with T-cell specific, immunosuppressant drug
FK506 inhibits hepatitis development [81, 160]. It is also a TNF-α dependent model since passive immunization using an anti-TNF monoclonal antibody is capable of inhibiting TNF-α and TNF-β completely, protecting the liver from Con-A induced hepatitis development. Recombinant IL-6 (rIL-6) treatment was also able to prevent
44
Con-A induced hepatitis but must be prophylactically administered at least 3 hours before
Con-A injection [160]. It was thought that IL-6 mediates its protective effects by negatively regulating TNF-α [161] but it has been shown that a single injection of rIL-6 completely protects against Con-A induced hepatitis while only reducing the TNF-α response by half [160]. Another potential mechanism by which IL-6 is protective is through its induction of the acute phase response [162]. With IL-6 mediating the production and secretion of acute phase proteins promptly, the liver is more prepared to handle the burden of Con-A induced hepatitis [160].
4.3.2. Carbon Tetrachloride
Carbon tetrachloride (CCl4) chemical hepatotoxin is capable of inducting fibrosis by altering membrane permeability [163]. See Chapter 3 on Murine Models of Liver Injury for more information on CCl4-induced hepatitis. TNF-α levels have been shown to increase with CCl4 administration further enhancing hepatic injury [164] that may be inhibited with anti-TNF-α antibody administration or genetic ablation of the TNF-a gene
[164, 165]. A deficiency of IL-6 (IL-6-/-) in mice has been shown to impair liver regeneration and increase the injury sustained to the liver after CCl4 administration [120]. rIL-6 treatment 20 minutes before CCl4 injection not only corrects the increased hepatic damage observed in IL-6-/- mice but also improved liver regeneration in wild-type animals [120]. In conclusion, IL-6 is protective to CCl4-induced hepatic fibrosis.
4.3.3. D-galactosamine
45
D-galactosamine, which impairs hepatic RNA and protein synthesis, sensitizes the liver
to damage [166]. When given in conjunction with LPS (LPS/GalN) it quickly induces
hepatitis and is a model of endotoxic shock mediated [167] primarily through TNF-α
[168]. Mice (C57BL/6J) given a low dose of LPS (100ng/mouse) in conjunction with D- galactosamine (8mh/mouse) have a mortality rate of 90% with 24 hours [168].
Administration of anti-IL-6 (Monoclonal Ab 20F-3) enhanced mortality of these
LPS/GalN treated mice [168]. Mice pretreated with rIL-6 (as low as 132ng) one hour before LPS/GalN injections are protected against mortality, but only if an anti-TNF antibody was used to partially attenuate TNF-α [168]. IL-6 appears to play a critical role in mitigating the TNF-α induced endotoxic shock that leads to mortality when given
LPS/GalN.
46
CHAPTER 5
MOUSE GENETICS
5.1. Inbred Genetic Mouse Strains
The laboratory mouse has been used to model human disease for the past century. The
cost of housing and maintenance is relatively low when compared to other mammals
frequently used for scientific research. Mice are anatomically and genetically similar to humans. A 2002 study showed that humans and mice share roughly 97.5% of their
protein-coding DNA [169]. This is not far from the similarity humans share with the
chimpanzee, which is around 98.5% identical [170]. This genetic relationship humans
share with mice makes them a valuable tool for genomic studies.
A mouse strain is regarded as inbred when it has been mated brother/sister for at least
twenty generations [171]. After 20 generations, 98.6% of the gene loci are homozygous
[172]. The longer a particular mouse strain has been inbred, the more identical the
genome. Hundreds of inbred strains are available to test the genetic variability between these strains and the phenotypical consequences [173]. C57BL/6 inbred mouse strain is
the most common mouse used in scientific laboratory research. This strain was
established in the 1920’s by Dr. C.C. Little and a particular sub-strain has been
maintained at The Jackson Laboratory since 1948 [174]. The A/J inbred mouse strain was
originally generated by L.C. Strong in 1921 and a sub-strain adopted and maintained at
The Jackson Laboratory since 1947 [175]. 47
5.2. Chromosome Substitution and Congenic Mouse Strains
Manipulation of mouse genetics is an excellent tool used to establish the genetic foundations of disease. A resource for gene discovery, gene characterization, and gene relationship analysis. A systematic manipulation of entire chromosomes within the mouse genome may be used to establish gene variants existing in different inbred mouse populations and their relationship to phenotypical observations. A mouse chromosome substitution strain (CSS) (also known as a consomic strain) is a mouse strain that is created by replacing an individual chromosome in a host mouse strain with the corresponding chromosome from a donor strain [176] (Figure 5.1). This allows for each segment of the donor strain to be assessed against a consistent, identical background.
The CSSs are made by intercrossing a homozygous host strain with a homozygous donor strain. The F1 generation will be heterozygous. The resulting hybrid is backcrossed into the donor strain for 10 generations. During the backcrossing, each new generation is genotypically screened and the mice that inherit a non-recombinant chromosome are selected for the next round of breeding. On average, one-in-five mice within a litter will
carry the desired non-recombinant chromosome [176]. After 10 generations of
backcrossing subsequent brother-sister breeding will maintain the CSS.
Entire CSS panels have been made. They consist of complete substitutions for each 19
autosomes, one for each sex chromosome (X & Y), as well as one for the mitochondrial
48
DNA [176]. The laboratory of Dr. Joseph H. Nadeau constructed a CSS panel using
inbred mouse strain A/J as the donor, and C5BL/6J as the host strain. They chose these
mouse strains due to their perceived phenotypical differences. The A/J and C5BL/6J
mouse strains differ in behavioral, immunological, and morphological characteristics
[177]. The Nadeau lab estimates that generation of this CSS panel took over 17,000 mice
and about 7 years to generate [177]. For example, a mouse homozygous for the 2nd chromosome while the rest of the genome is C57BL/6J would be designated B6-Chr2A/J.
The CSS panel allows the identification of quantitative trait loci (QTL). These are segments of DNA associated with genes that underlie a measurable trait. Mapping of these regions attempts to correlate a particular section of DNA with a quantitative trait
[178].
49
Figure 5.1. Selected analytical tools in mouse genetics. The generation of coisogenic, recombinant inbred, consomic and congenic strains is illustrated. Reprinted by permission from Macmillan Publishers Ltd: Nature Reviews Immunology. Rogner UC, Avner P. Congenic mice: cutting tools for complex immune disorders. Nat Rev Immunol. 2003 Mar;3(3):243-52. Copyright 2003
50
Limitations exist for CSSs. Their substitution of an entire chromosome allows for
multiple QTLs to be present within the substituted chromosome. Although the rest of the
genome is consistent, the presence of an entire chromosome from the donor mouse allows
for limited precision in targeting individual QTLs. Once a QTL is identified, assessing
its boundary is difficult, although this may be overcome using congenic strains.
CSS strain were made to implicate entire chromosomes of a given inbred genetic strain
with a quantitative trait. Taking this process a step further, to allow for a narrowing of
the field would increase the sensitivity (and power) of the individual trait being
characterized. A congenic strain, and their derivatives (subcongenic, sub-subcongenic) is
a substitution of only a segment of a given chromosome from a donor strain with its
corresponding segment in the host strain. Using an existing CSS strain, congenic mice
are made by crossing the CSS with the inbred host strain (Figure 5.1). The progeny are
then backcrossed to the host strain with selection for progeny that carry a recombinant
chromosome within a targeted region. This mouse carrying the targeted recombinant is
subsequently backcrossed to the host strain and then intercrossed to create a homozygous
mouse [178, 179].
A 2004 study investigated diet-induced obesity susceptibility of a CSS panel with A/J as the donor and C57Bl/6J (B6) as the host strain. Mice were weaned at 4 weeks of age, fed a high-fat, low-sucrose (HFLS) containing diet (Research Diets, #D12330) of a high-fat,
high-sucrose containing diet (HFHS) (Research Diets, #D12331). 17 of the 22 CSS
51
strains fed the HFHS diet gained significantly less weight that the B6 control strain
including B6-Chr17A/J (CSS17). A panel of 7 congenic mice derived from CSS17 strain
was subjected to a high-fat, simple-carbohydrate (HFSC) diet (same as HFHS diet,
Research Diets, #D12331) or a low-fat, simple-carbohydrate (LFSC) diet (Research
Diets, #D12329). After 16 weeks on the diets, CSS-17 and 5 of the 7 congenic strains fed
the HFSC diet weighed significantly less than C57BL/6J mice [180]. QTL that associate
with diet-induced obesity resistance are termed obesity resistant QTLs (Obrq). From this
study, three QTLs (Obrq13, Obrq14, and Obrq15) located on the 17th chromosome were identified that confer resistance to diet-induced obesity when the A/J derived allele is present [180]. Obrq13 also protects against susceptibility to liver steatosis and elevated fasting blood glucose and insulin on a HFSC diet. Obrq15 also protects against liver steatosis on a HFSC diet [180].
52
CHAPTER 6
IMMUNITY
6.1. Overview
The immune system functions to serve as a host defense against invading organisms and
foreign materials exposed to the body through inhalation, ingestion, and/or penetration of
the skin or mucous membrane. The immune system is divided into two subsystems
defined by the speed and specificity in which the response to exposure occurs. The
innate immune response is categorized by an immediate response in host defense. The
other subsystem of immunity, the adaptive immune response consists of antigen-specific
targeting of invaders by T and B Lymphocytes. The adaptive response specificity allows
for precise targeting of these intruders with little collateral damage to healthy surrounding
tissues. This response may take days to weeks to fully mature [181].
6.2. The Innate Immune Response
The response of the innate immune system to infection is immediate. The host defense
mechanism would require this immediate reaction in order to effectively control any
potential bacterial or viral infection while the lymphocytes of the adaptive immunity
response differentiate into effector cells to attack the trespasser. Although this is a critical role of the innate immune system, it is not the only role [182].
53
6.2.1. Neutrophils
A critical response of innate immunity is the recruitment and stimulation of neutrophils to
kill infectious agents. The initial recruitment of the neutrophil is attributable to
chemokine signaling released from macrophages present at the site of infection.
Pathogen-associated molecular patterns (PAMPs) and damage-associated molecular
patterns (DAMPs) are recognized by the macrophages using pattern recognition receptors
(PRRs) such as the cell surface bound toll-like receptors (TLR) or cytoplasmic sensors
nucleotide-binding oligomerization domain (NOD)-like receptors (NLRs). These DAMPs
and PAMPs attributable to infection or cell trauma are detected and the macrophages
release cytokines IL-1, IL-6, and TNF-α as well as several types of chemokines [183,
184]. Neutrophils are mobile, freely traveling throughout the body and upon sensing one of these chemoattractants they begin to translocate to the site of chemokine origination.
The exact mechanism by which neutrophils navigate to the area of injury is not fully understood but it is known that they utilize a chemokine gradient as well as the ability to distinguish between intermediary chemoattractants and end-target chemoattractants
[185]. Upon arrival at the site of infection, the neutrophils begin to phagocytize infectious organisms and kill the infecting agent through two distinct mechanisms. The first, is an oxygen-dependent response known as a respiratory burst, in which toxic oxygen metabolites such as hydrogen peroxide and hydroxyl free radicals are used to destroy the infectious agent [186]. The second mechanism is through the use of neutrophil produced bacteriolytic enzymes and cationic antimicrobial peptides which attack bacterial membranes, triggering bacteria cell death [187].
54
6.2.2. Cytokines
Cytokines are messenger proteins secreted by a cell to send a signal in an autocrine, paracrine, or endocrine manner. Essentially, they function as a hormone for the immune system. Cells of the immune system must be capable of “talking” to each other in order to mount a sufficient host defense. They can affect multiple biological processes from embryonic development to disease pathogenesis. Cytokines send their signal by binding to cell surface receptors. The term cytokine encompasses a broad range of messengers including interferons, interleukins, chemokines, mesenchymal growth factors, tumor necrosis factors, and adipokines. The nearly endless responses cytokines are capable of generating has triggered a large amount of medical research the past few decades in hopes of finding viable disease treatments in the fields of inflammation, auto-immune diseases, and tissue differentiation among many others [181, 188]. It is thought that cytokines evolved from early transcription factors. The method in which they carry out their processes is thought of as a soluble protein secreted from a cell. The protein then binds to a specific membrane-bound receptor of the cell of which it was secreted, nearby cells, or even cells across the body utilizing the vasculature. This is not always the case, in fact, some cytokines never leave the cell. Remarkably, every cell in the human body is capable of producing and responding to cytokines except the red blood cell [188].
A cytokine is a double-edged sword of sorts. Cytokines are absolutely essential to a healthy innate immune response but they may also be attributable to disease initiation and progression. Dysregulation of cytokine production and/or secretion leads to autoimmune diseases. As a consequence, inhibitors of cytokine signaling is a key area of research for
55 viable treatment options. For instance, sepsis causes over 500,000 deaths each year within the United States and clinical investigators thought that by blocking the actions of proinflammatory cytokines such as IL-1 and TNF-α they may decrease mortality. After clinical trials in over 12,000 patients, metaregression analysis showed that only individuals at high risk of death received any benefit and that treatment may have actually been detrimental to patients with low risk [189]. On the other hand, inhibition of cytokines to treat autoimmune diseases such as rheumatoid arthritis and Crohn’s disease have been decidedly successful [190, 191]. For further reading on far-reaching effects of cytokines and how they came to be, a 2011 review titled “Historical Insights into
Cytokines” published by Charles Dinarello in the European Journal of Immunology extensively covers the history and biological impacts of cytokines extensively [188].
56
CHAPTER 7
THE INFLAMMASOME
7.1. Overview
NLR family proteins in humans consists of 22 members playing a role in innate immunity
whose purpose is to detect pathogen-associated molecular patterns (PAMPs) and damage associated molecular patterns (DAMPs). These proteins sense signals associated with pathogenic infection, cell damage, cell death, and overall cellular environmental stress.
Hematopoietic immune cells such as macrophages, neutrophils, and dendritic cells respond to danger signals using a variety of sensors. These sensors can be situated on the plasma membrane of the cell as transmembrane receptors like members of the TLR and
C-type lectin receptor families, surveying the extracellular space. NLR family proteins are responsible for the detection of intracellular danger signals [192]. When this
intracellular sentry binds its corresponding PAMP/DAMP, it activates a cascade of events
leading to the formation of a multimeric protein structure known as the inflammasome.
This inflammasome ultimately leads to the production of inflammatory cytokines to
allow the cell to respond to the danger signal [193]. There have been several different names given to these proteins including but not limited to, NOD-LRR, NACHT-LRR,
NOD-like receptor, CARD, NALP, NOD, and PYPAF [194]. More recently, literature
has begun to designate this family of proteins using the standardized nomenclature
“NLR”, standing for nucleotide-binding domain and leucine-rich repeat containing gene
family. This nomenclature was approved by the HGNC and the Mouse Genomic
57
Nomenclature Committee and seems to have been universally adopted as the term of
choice when describing NLR family proteins [194].
7.2. Inflammasome Structure
These NLR family immune regulatory proteins are characterized by the presence of a nucleotide-binding and oligomerization (NACHT) nucleotide-binding domain (NBD) followed by C-terminal leucine-rich repeats. The N-terminus of the protein has either a caspase activation and recruitment domain (CARD) or pyrin (PYD) domains [195]. The
domains, it is thought, serve different roles allowing for NLR’s to sense its ligand,
autoregulate, and interact with other proteins to activate inflammasome assembly.
The leucine-rich repeat (LRR) domain situated at the C-terminus of the NLR protein is
made up of a variable number of leucine repeats. LRRs by definition are 20-29 protein
residue sequence motifs. Most proteins containing an LRR are eukaryotic (but not all)
and most of these sequences seem to be associated with protein-protein recognition
N processes. The LRR motif contains a conserved 11-residue segment, LxxLxLxx /CxL.
N “x” can be any amino acid, “ /C” designates a conserved asparagine or cysteine residue, and “L” may be leucine, valine, isoleucine, and phenylalanine. The 3-dimenstional structures of many proteins containing LRRs have been determined and provide insight into the structure and function of LRR regions in NLR proteins. Most LRR consist of repeating β strand and α helix connected by loops forming a diverse range of structures
similar in conformation to a horseshoe or fish hook with the β sheet lining the inner
58
concave surface and the α helices lining the outer edge or convex side [196]. This
structure is rigid enough for protein-protein interaction but maintains a structural
flexibility allowing for interactions with diverse binding partners [197]. Many parts of the LRR surface are utilized for ligand interaction but the concave surface and neighboring loops of the LRR provide an excellent target interface with the a-helix of another interacting protein [198, 199]. For the NLR family proteins, the LRR is thought
to serve the purpose of autoregulation, protein-protein interaction, and PAMP sensing
[200].
The central nucleotide-binding and oligomerization (NACHT) domain is the only domain
common to all NLR family members. The principal purpose of this domain is to enable
activation of the inflammasomal signaling complex through interaction with ATP.
Homology modeling indicates that minor sequence variability among NLR family
proteins within the NACHT domain results in differences in the ATP binding pocket.
The 3-dimensional structure of the nucleotide binding region from APAF1 [201] and
NOD-like receptor containing a caspase activating and recruitment domain 4 (NLRC4)
[202] have been documented. This area is characterized by α/β-fold with five parallel β
strands in a central core surrounded by four α helices on each side. This domain contains
conserved motifs including an ATPase-specific P-loop and a Mg2 coordination loop
attributed to nucleotide binding and/or hydrolysis [203]. For NLRC4, adenosine
diphosphate-mediated interaction within the ATP binding pocket of the NACHT domain
was found to be indispensable in stabilizing the “closed” conformation of NLRC4,
disallowing self-oligomerization and subsequent inflammasome activation [202]. Single
59 nucleotide polymorphisms (SNP) have been documented in human NLR genes within the
NACHT domain. Some of these alterations cause genetic diseases attributable to an altered inflammatory response. Mutations R260W, D303N, and L353P are associated with Muckle-Wells syndrome and chronic infantile neurological cutaneous and articular syndrome [204].
Situated at the C-terminus of the NLR family proteins lies a CARD domain or PYD domain. Both these domains mediate protein-protein interaction for inflammasome signaling. The PYD interacts with the adaptor apoptosis-associated speck-like protein containing a caspase activation and recruitment domain (ASC). ASC has a PYD for interaction with the PYD of NLRs but is also contains a CARD for interaction with the
CARD of procaspase 1. ASC essentially forms a bridge between an NLR family protein and capsase-1. Those NLR family proteins containing C-terminal CARD instead of the
PYD can bind procaspase 1 independent of ASC [205].
7.3. NLR Family
There have been a variety of names used to describe the human NLR gene family as well as its subfamilies and individual genes. These names include CATERPILLAR, NOD-
LRR, NACHT-LRR, CARD, NALP, NOD, PAN, and PYPAF. In 2008, the HUGO
Gene Nomenclature Committee unified the nomenclature of these family of proteins under the designation “NLR”, standing for nucleotide-binding domain and leucine-rich
60
repeat containing [194]. The subfamilies were organized by their N-terminal effector domains.
61
Nomenclature of NLR New approved symbol Other names and aliases Domain organization CIITA C2TA AD-NACHT-NAD-LRR NAIP BIRC; CLR5.1 BIR3x-NACHT-NAD?-LRR NOD1 CARD4; CLR7.1 CARD-NACHT-NAD-LRR NOD2 CARD15; CD, BLAU, IBD1, PSORAS1; CARD2x-NACHT-NAD- CLR16.3 LRR NLRC3 NOD3; CLR16.2 CARD-NACHT-NAD-LRR NLRC4 CARD12; CLAN; CLR2.1, IPAF CARD-NACHT-NAD-LRR NLRC5 NOD27; CLR16.1 CARD-NACHT-NAD-LRR NLRX1 NOD9; CLR11.3 X-NACHT-NAD-LRR NLRP1 NALP1; DEFCAP; NAC; CARD7; PYD-NACHT-NAD-LRR- CLR17.1 FIIND-CARD NLRP2 NALP2; PYPAF2; NBS1; PAN1; PYD-NACHT-NAD-LRR CLR19.9 NLRP3 CIAS1; PYPAF1, cryopyrin; CLR1.1, PYD-NACHT-NAD-LRR NALP3 NLRP4 NALP4; PYPAF4; PAN2; RNH2; PYD-NACHT-NAD-LRR CLR19.5 NLRP5 NALP5; PYPAF8; MATER, PAN11; PYD-NACHT-NAD-LRR CLR19.8 NLRP6 NALP6; PYPAF5; PAN3; CLR11.4 PYD-NACHT-NAD-LRR NLRP7 NALP7; PYPAF3; NOD12; PAN7; PYD-NACHT-NAD-LRR CLR19.4 NLRP8 NALP8; PAN4; NOD16; CLR19.2 PYD-NACHT-NAD-LRR NLRP9 NALP9; NOD6; PAN12; CLR19.1 PYD-NACHT-NAD-LRR NLRP10 NALP10; PAN5; NOD8; Pynod; PYD-NACHT-NAD CLR11.1 NLRP11 NALP11; PYPAF6; NOD17; PAN10; PYD-NACHT-NAD-LRR CLR19.6 NLRP12 NALP12; PYPAF7; monarch-1; RNO2; PYD-NACHT-NAD-LRR PAN6; CLR19.3 NLRP13 NALP13; NOD14; PAN13; CLR19.7 PYD-NACHT-NAD-LRR NLRP14 NALP14; NOD5; PAN8; CLR11.2 PYD-NACHT-NAD-LRR
Table 7.1. Nomenclature of NLR. The HUGO recommended nomenclature for NLR protein family. The domain structure of each gene is listed. AD: acid transactivation domain; BIR: baculoviral inhibitory repeat; PYD: pyrin domain; NACHT: NAIP, CIITA, HET-E, and TP1; CARD: caspase recruitment domain; NAD: NACHT-associated domain; LRR: leucine-rich repeats. Reprinted from Immunity. Ward PA. The NLR gene family: A Standard nomenclature. Immunity 2008; 28(3):285-287. doi:10.1016/j.immuni.2008.02.005. Copyright 2008, with permission from Elsevier.
62
The entire human NLR family is divided into four sub families.
1.) NLRA subfamily is comprised of only CIITA (class II, major histocompatibility
complex, transactivator) due to a transactivator domain as its N-terminal effector
domain. It is expressed in macrophages, B cells, T cells, and dendritic cells [206].
The acidic transactivation domain recruits an enhanceosome for the purpose of
regulating major histocompatibility complex class I and class II gene
transcription. The enhanceosome consists of DNA binding transcription factors
RFX, cyclic AMP response element binding protein, and NF-Y. CIITA is well
established as a master regulator of MHC gene transcription but has also been
shown to regulate over 60 immune related genes including cytokine IL-4 and IL-
10 [194, 207, 208].
2.) NLRB subfamily is comprised of only NAIP (NLR family, apoptosis inhibitory
protein) with its N-terminal effector domain containing 3 consecutive baculoviral
inhibitor of apoptosis repeat (BIR) domains. Mice are known to express NAIP1,
NAIP2, NAIP5, and NAIP6 while humans are known to only express a single
NAIP. In mice, NAIP5 and NAIP6 detect flagellin while NAIP2 detects type III
secretion system rod protein and NAIP1 detects bacterial type III secretion
systems (T3SS) needle protein. Human NAIP detects T3SS needle protein as well
[209]. Upon ligand binding to NAIP, a trimeric complex is formed with NLRC4,
forming a NAIP/NLRC4 inflammasome that activate caspase-1 [194, 210].
3.) NLRC subfamily contains NOD1 (Nucleotide-binding oligomerization domain-
containing protein 1), NOD2 (Nucleotide-binding oligomerization domain-
containing protein 2), NLRC3 (NLR family, CARD domain containing 3),
63
NLRC4 (NLR family, CARD domain containing 4), NLRC5 (NLR family,
CARD domain containing 5). The N-terminus contains a CARD (caspase
activation and recruitment domain) domain which, as its name implies, is a
protein motif responsible for association with caspase-1 leading to capase-1
activation. Active caspase-1 induces rapid cell death through cellular pyroptosis
as well as proteolytically activates proinflammatory cytokines IL-1β and IL-18.
The overall purpose is to kill cells infected with bacteria and recruit immune cells
to the area of infection [194, 211, 212].
4.) NLRP subfamily contains NLRP1 (NOD-like receptor family, pyrin domain
containing 1), NLRP2 (NOD-like receptor family, pyrin domain containing 2),
NLRP3 (NOD-like receptor family, pyrin domain containing 3), NLRP4 (NOD-
like receptor family, pyrin domain containing 4), NLRP5 (NOD-like receptor
family, pyrin domain containing 5), NLRP6 (NOD-like receptor family, pyrin
domain containing 6), NLRP7 (NOD-like receptor family, pyrin domain
containing 7), NLRP8 (NOD-like receptor family, pyrin domain containing 8),
NLRP9 (NOD-like receptor family, pyrin domain containing 9), NLRP10 (NOD-
like receptor family, pyrin domain containing 10), NLRP11 (NOD-like receptor
family, pyrin domain containing 11), NLRP12 (NOD-like receptor family, pyrin
domain containing 12), NLRP13 (NOD-like receptor family, pyrin domain
containing 13), and NLRP14 (NOD-like receptor family, pyrin domain containing
14). Members of the NLRP subfamily contain a PYRIN domain in their N-
terminus which is a putative protein-protein interactive domain that initiates
apoptotic and inflammatory signaling pathways [194, 213]. The Pyrin domain is
64
highly evolutionarily conserved in both human and murine species and is the most
abundant member of the death domain superfamily [214]. Variations in pyrin
domains across NLRPs include length of α-helices forming the death domain fold
as well as variability in the surface charge at interfaces of protein-protein
interaction [215].
It is important to note there is one remaining NLR family protein that is not included in the four sub-families. NLRX1 (NLR family member X1) bares no strong homology to the
N-terminal domain of any other NLR family member [194]. NLRX1 is expressed in the mitochondria and has been shown to play a role as a negative regulator of antiviral signaling [216]. Arnoult and colleagues further characterized this protein through identification of a mitochondrial-addressing sequence in the N-terminus allowing for direct targeting to the mitochondrial matrix and through bcl complex resulting in generation of reactive oxygen species [217].
7.4. Inflammasome Function
Inflammasomes are initiate a class of inflammatory caspases. These caspases in humans and murine species are caspase-1 and caspase-12 [218]. The family members of interleukin-1β (IL-1β)-converting enzyme (ICE)-like proteases are named in the order of their publication, therefore the first member of the family is caspase-1. Caspases are synthesized in an inactive form, known as proenzymes. They may be activated following the cleavage of specific aspartate cleavage sites [219].
65
7.4.1. DAMPs and PAMPs
Pathogen-associated molecular patterns (PAMPs) are biological compounds present in
organisms that are not present in the host organism. They serve as markers for foreign
body infiltration into the host and potentiate a stimulation of the innate and adaptive immune system responses. A common example of a PAMP is LPS presence in human blood. This is a compound found in outer membrane of gram-negative bacteria and would serve as an indicator of bacterial infection, which would necessitate an immune system response [220]. PAMPs include microbial nucleic acids (DNA, dsRNA, ssRNA,
and 5′-triphosphate RNA), lipoproteins, surface glycoproteins, membrane components
(LPS), and glycosylphosphatidylinositol [221]. These molecules are recognized by
pattern recognition receptors (PRRs) such as TLR, AIM2 like receptors, retinoid acid- inducible gene I -like receptors, and NOD-like receptors (NLRs) [221, 222].
A heavily researched PAMP, LPS, binds to its PRR, TLR4. TLR4 activation leads to the
production pro-inflammatory cytokines and chemokines [223]. TLR4 agonist compounds
(derivatives of LPS) have recently been used in human vaccines for hepatitis B and
human papilloma virus as an adjuvant to trigger T-helper 1 and T-helper 2 immune
response, improving the efficacy of vaccinations [224]. TLR4 signaling is a myeloid
differentiation primary response 88 (MyD88)-dependent pathway that induces
transcription factor NF-κB translocation to the cell nucleus and induces pro-inflammatory
cytokine gene transcription [225].
66
DAMPs are host cell derived molecules that initiate an immune response. DAMPs are commonly nuclear and cytosolic proteins (including some membrane proteins) that are released into the extracellular matrix upon cell lysis [226]. Non-protein DAMPs exist, some being ATP, uric acid, heparin sulfate, RNA, and DNA. DAMPs interact with
DAMP specific receptors activate cell signaling cascades including mitogen-activated protein kinases, NF-κB, and PI3K/AKT.
In 1989 Charles Janeway suggested antigen-presenting cells recognized infectious pathogens using membrane bound receptors. He named this “signal 0”, which has since evolved to encompass PAMPs and DAMPs[220]. PAMPs were quickly identified as components of microorganisms that allowed for host cells to distinguish self from non- self and to generate and immune response [221]. DAMPs were originally proposed by
Polly Matzinger as molecules activating antigen-presenting cells to tissues damage through internal host stressors or microbial-induced damage [227].
PAMPs are commonly referred to as exogenous signal 0 molecules while DAMPs are endogenous signal 0 molecules. Signal 0, is in reference to a molecular signal that binds to a specific cell membrane receptor that induce autophagy. Autophagy is a lysosomal degradation pathway in response to environmental and cellular stress to invoke cell survival [221]. PAMP and DAMPS binding to PRRs induce pathways synthesizing
67
cytokines, chemokines, cell adhesion molecules, and immunoreceptors critical in the
adaptive immune response [221].
In mammalian cells, NLR family proteins induce autophagy upon sensing bacterial compounds. Neuronal apoptosis inhibitory protein (NAIP) proteins are cytosolic
receptors. In mice, NAIP5 functions as a cytosolic sensor of bacterial flagellin. NAIP1
and NAIP2 detect needle and inner rod proteins which are gram-negative bacteria
components of bacterial type III secretion systems (T3SS). It is important to note that in
humans, the only NAIP protein is responsible for detection of the T3SS compounds, not
flagellin [228]. It is hypothesized that alternative pathways are utilized in human cells to
detect bacterial flagella [229]. NAIPs bound to their bacterial ligand associate in the
cytosol with adapter protein NLRC4. NLRC4 must be distinguished as an adapter
protein, as previous literature has referred to this protein as the sensor but more recent
findings reveal NAIP as the initial sensor, with NLRC4 functioning as an adapter
responsible for the recruitment of caspase-1 to facilitate the downstream signal
(proinflammatory IL-1β, IL-18) in response to the bacterial infection [230].
7.4.2. Caspase-1
Caspase-1 is formerly known as IL-1 converting enzyme. Initial investigation of this
cysteine protease focused around its ability to cleave the inactive precursor cytokine IL-
1β (31 kDa) at Asp-116-Ala-117 producing a matured, active (17 kDa) IL-1β
inflammatory cytokine [231]. A single 45 kDa caspase-1 proenzyme is cleaved into two
68
subunits to become proteolytically active. The subunits, 20 kDa and 10 kDa, are
commonly referred to as p20 and p10. The four cleavage sites required for generation of
these 2 subunits were originally suggested to be autoproteolytically cleaved due to their
cleavage at Asp-Xaa bonds [232]. Further investigation has revealed that caspase-1 is
activated by recruitment to a large multimeric molecular platform called the
inflammasome [195].
This activation of caspase-1, facilitated by the inflammasome, occurs through dimerization of the uncleaved caspase-1 through an autocatalytic reaction generating the active caspase-1 molecule comprised of two p20 and two p10 subunits [233]. The
consequence of caspase-1 activation is rapid cellular secretion of IL-1β as well as
programmed cell death through pyroptosis. The half-life of active caspase-1 is roughly 9
minutes at a physiological temperature of 37ºC [234]. An active caspase-1 enzyme is also
capable of activating pro-IL-18. Although over 600 targets for apoptosis-associated
effector caspases have been documented, IL-1β and IL-18 are only two target substrates
for the inflammatory caspase, caspase-1, confirmed in vitro and in vivo [234]. Several
other cleavage targets of caspase-1 have been identified through proteomic analysis but
their significance under physiological conditions has not been elucidated [235].
7.4.3. Apoptosis
7.4.3.1. Overview
69
The term apoptosis was first described in 1972 describing cell death as a counterbalance to mitosis in the regulation of cell populations [236]. Its characteristics include nuclear and cytoplasmic condensation, membrane bound fragmentation of cell compartments, formation of intact apoptotic bodies and degradation of these bodies by other cells through lysosomal enzymes [236].
The term apoptosis is derived from Greek with the prefix apo- representing “separation” and -ptosis meaning “falling off” and is generally considered to refer to “falling off of a petal from a flower” or “falling off of leaves from a tree” [237, 238]. It is now generally recognized as the process of programmed cell death resulting in little to no inflammatory response [239]. Taking into account this non-inflammatory characteristic it’s considered a normal housekeeping mechanism allowing for the death of older cells to make way for new [240], a graceful end to the life of a cell.
As a complement to the general maintenance of cell populations, is functions as a host defense mechanism to destroy infected cells or those damaged beyond repair [241].
Excess heat, radiation exposure, hypoxia, and cytotoxicity from drug agents are all triggers of apoptosis not directly attributed to infective causes [240].
Early apoptotic cell death is morphologically identified by cell shrinkage and concentration of the cell organelles with loss of cytoplasmic volume. Chromatin becomes tightly compressed. Cells appear oval in shape with membrane blebbing leading to
70
identifiable formation of apoptotic bodies in a process called “budding” [236]. These
bodies are engulfed by phagocytic macrophage cells. Apoptotic bodies therefore do not spill their cellular contents into the extracellular space.
Three apoptotic mechanisms have been described to date. The first is an extrinsic or death receptor mediated pathway. The second, an intrinsic or mitochondrial mediated pathway, and the third, a T-cell mediated cytotoxicity perforin-granzyme-dependent
mediated pathway. These pathways are not distinct and share similar key mechanisms in
their coordinated network mediating cell death.
7.4.3.2. Extrinsic Apoptotic Pathway
The extrinsic pathway of apoptotic cell death involves extracellular triggers binding death
receptors in the TNF receptor gene superfamily, of which contain a cytoplasmic domain
known as the death domain [242]. The most heavily researched ligands for these
receptors are TNF-α and FasL. Upon ligand binding to the receptor, intracellular adapter
proteins containing corresponding death domains associate with the receptor. For
example, Fas-Associated protein with Death Domain (FADD) binds the FasL/FasR
complex and Tumor necrosis factor receptor type 1-associated DEATH domain protein
binds the TNF-α/TNFR complex with subsequent recruitment of FADD and the serine-
threonine kinase receptor-interacting protein [243]. Initiator caspase procaspase-8
associates with the structure, mediated through its death effector domain with FADD
followed by autocatalytic activation of caspase-8 [244]. Activated caspase-8 cleaves
71
effector (or executioner) caspases -1,-3,-6, and -7 [245]. These caspases, once activated,
cleave various substrates to induce cell death. Caspase-3 activates endonuclease
Caspase-Activated DNase (CAD) by releasing it from its inhibitor, inhibitor of Caspase-
activated DNase. CAD then degrades chromosomal DNA leading to the apoptosis
characteristic chromatin condensation [246]. This pathway can be inhibited by cellular
FLICE (FADD-like IL-1β-converting enzyme)-inhibitory protein (c-FLIP), which binds
to the FADD/caspase-8 complex [240, 247].
7.4.3.3. Intrinsic Apoptotic Pathway
The intrinsic pathway of inducing apoptosis, is a non-receptor-mediated pathway.
Intracellular signals serve as negative regulators, for example the absence of growth factors, hormones, and cytokines leading to a failure to suppress triggers of apoptosis.
Signals serving as positive regulators such a radiation, toxins, excess heat, viruses, and free radicals initiate apoptosis [240]. This stimuli, or lack of stimuli, leads to an opening of the mitochondrial pore and release of the pro-apoptotic protein, cytochrome-c.
Cytochrome-c binds Apoptotic protease activating factor 1 (APAF1) forming the wheel- shaped apoptosome [248] which also requires the presence of either 2'-deoxy ATP or
ATP [249]. The apoptosome cleaves initiator caspase, procaspase-9. Procaspase-9 normally exists as an inactive monomer and is activated by the apoptosome through dimerization and subsequent autocatalytic cleavage [250]. Caspase-9 then initiates cleavage and activation of effectors caspases, caspase-3 and caspase-7, inducing cell death [251]. This pathway is regulated through B-cell lymphoma 2 (Bcl-2) family proteins.
72
The Bcl-2 family proteins can be simply characterized as proapoptotic or antiapoptotic.
Proapoptotic Bcl-2 family members, BCL-2 antagonist killer 1 and BCL-2-associated x
protein, upon activation homodimerize to form pore structures in the outer mitochondrial
membrane to promote permeability, facilitating the release of cytochrome-c (and other
soluble proteins) and initiating the intrinsic apoptotic pathway. Antiapoptotic Bcl-2
family members inhibit the apoptotic pathway by directly inhibiting the proapoptotic Bcl-
2 proteins [252].
7.4.3.4. Perforin-Granzyme-Dependent Apoptotic Pathway
CD8+ cytotoxic T lymphocytes (CTL), or killer T cell, are able to selectively target
antigen presenting cells. When exposed to infected or dysfunctional somatic cells, cytotoxic T cells kill these cells through the extrinsic apoptotic pathway via production of
FasL and subsequent apoptotic death [253]. CTL are also capable of inducing targeted cell death through another, independent mechanism. CTL interact with the target cells membrane and secrete the transmembrane pore-forming molecule perforin. The CTL then release serine protease cytoplasmic granules granzyme A and granzyme B into the target cell cytoplasm. Granzyme A activates DNA nicking through DNAse NM23-H1
[254]. Granzyme B cleaves inhibitor of Caspase-activated DNase, freeing CAD to induce chromatin condensation [246] as well as cleavage of BID [255], and direct activation of caspase-3 [240]. Together, this targeted interaction with target cells by CTL and ensuing insertion of granzymes induces an apoptotic cell death.
73
7.4.4. Pyroptosis
Pyroptosis is a caspase-1 dependent form of programmed cell death resulting in rapid cell lysis. The term pyroptosis was first used by Brad T. Cookson in 2001 derived from a
conjugation of the Greek words “Pyro” meaning fire or fever and “ptosis” meaning
falling [256]. It was first described in 1992 as published literature investigating gram-
negative bacteria Shigella flexneri induction of an apoptosis like cell death in infected
macrophages [257]. Before pyroptosis, cell death was attributed to two principal
pathways, apoptosis and necrosis [258].
Opposed to apoptosis, is the inflammatory cell death mechanism pyroptosis which is independent of apoptotic caspases. The inflammatory caspase, caspase-1, activates proinflammatory cytokines IL-1β and IL-18. Apoptosis functions normally even in the absence of caspase-1, further delineating the processes involved in apoptotic and pyroptotic cell death [259]. Pyroptosis is characterized by DNA damage with chromatin condensation in the presence of an intact nucleus as well as membrane swelling that
eventually ruptures, releasing cell contents to the extracellular matrix. This rupture event
may be measured by lactate dehydrogenase content in the serum [260, 261]. Cellular lysis results in the proinflammatory characterization attributed to pyroptotic cell death.
7.4.5. Interleukin-1β
74
The term interleukin was first used by immunologists discussing soluble factors acting on lymphocytes [262]. IL-1β is an inflammatory cytokine and a member of the interleukin-1
(IL-1) family, the second largest of the interleukin families. The IL-1 family includes seven ligands (IL-1α, IL-1β, IL-18, IL-33, IL-36α, IL-36β, and IL-36γ), three receptor antagonists (IL-1Ra, IL-36Ra, and IL-38) and one anti-inflammatory cytokine (IL-37)
[263]. The IL-1 family receptors contain both activators and suppressors of inflammation. Remarkably, IL-1α and IL-1β bind the same receptor but share only 24% amino-acid sequence identity [262].
The IL-1 receptor family encodes ten receptors although some are orphan receptors
(receptors with no confirmed ligand thus far) [264]. The IL-1 family receptor chains contain three immunoglobulin (Ig)-like domains in their extracellular portion. The intracellular portion is characterized by a toll-interleukin receptor domain that interacts with the MyD88 adaptor protein. Interestingly, the TIR domain of the IL-1 family receptor is also shared by the toll-like receptor class of proteins.
IL-1β is produced by hematopoietic cells like blood monocytes, tissue macrophages, skin dendritic cells, and brain microglia. It is primarily produced in response to TLR signaling, activated compliment (C5a), and other cytokines including itself. Large amounts of IL-1β mRNA may be transcribed in response to a multitude of triggers while a majority of it is never translated and is quickly degraded. In fact, many microbial products induce IL-1β transcription, mostly through TLRs. The IL-1β (as well as the IL-
75
18 and IL-1F7) mRNA contains an instability element within the coding region responsible for a significant translational failure rate to a functional protein. The addition of TLR or other active IL-1 family cytokines however greatly increases the rate of successful IL- 1β translation [264]. In human peripheral blood mononuclear cell lines IL-
1β ligand was able to stimulate itself as induction of mRNA levels was observed for over
24 hours [265]. Human monocytes stimulated with TLR4 agonist endotoxin LPS only release roughly 20% of translated IL-1β over a 24-28 hour period [266]. A second stimulus, extracellular ATP, greatly increases the maturation and secretion of IL-1β from these monocytes [267].
IL-1β translated from its mRNA template exists as a precursor with no biological activity.
This IL-1β precursor must be cleaved to a bioactive form by caspase-1 (see the section on caspase-1 in this chapter for more detail). In order for IL-1β to stimulate its own transcription, it must bind to its receptor and form the heterodimer receptor complex IL-
1RI/IL-1RAcP. The complex formation results in the intracellular TIR domains recruitment of MyD88 adaptor protein. A subsequent cascade of signaling results in NF-
κB activation of IL-1β transcription [268]. Translation of the protein takes place in the cytosol (not the ER). Extracellular ATP acts as a trigger of the P2X7R to facilitate a rapid efflux of K+ [269]. The decrease in intracellular K+ leads to the activation of the
NLRP3 (NALP3) inflammasome and, with the adaptor protein ASC, processes
procaspase-1 into active caspase-1. NALP3 is a protein encoded by the NLRP3 gene in
humans [270]. IL-1β is then cleaved by this active caspase-1. IL-1β lacks a conventional
hydrophobic signal sequence utilized in the classical model of secretion of proteins.
76
Proteins lacking this signal sequence are dubbed “leaderless” and require a non-classical method of secretion [271]. There are five different non-classical release mechanisms suggested for IL-1β. First is exocytosis through secretory lysosomes [272]. The second,
shedding of IL-1β from the plasma membrane via microvesicles [273]. The third,
proposes fusion of microvesicles with the plasma membrane leading to the release of IL-
1β-containing exosomes [274]. The fourth, export through specific membrane
transporters [275]. The fifth, is simply the release of IL-1β upon cell lysis [276].
Inflammasome processing of IL-1β is the rate-limiting step in the processing and
secretion of IL-1β. Inhibitors of caspase-1 reduce the secretion of the mature IL-1β. As
mentioned above, the NLRP3 inflammasome mediated processing of IL-1β was the first
inflammasome to be identified as an activator of caspase-1 and subsequent activation of
IL1-β [277] but other identified inflammasome complexes are capable of activating
caspase-1 and therefore processing IL-1β.
The ability for the inflammasome to mediate caspase-1 activation was thought to require
activation of the P2X7 receptor via ATP leading to the efflux of intracellular K+.
Recently it was discovered that the P10 subunits of activated caspase-1 was detected in human blood mononuclear cells without ATP stimulation [278]. It appears that the activation of the inflammasome activation may vary from cell to cell. Differentiated macrophages have been shown to require extracellular ATP stimuli for inflammasome activation [278].
77
Humans with a mutation in inflammasome component NLRP3 (or NALP3) are associated with chronic autoinflammatory diseases (as a result of spontaneous release of
IL-1β from monocytes) including familial cold autoinflammatory syndrome, Muckle-
Wells syndrome, and chronic infantile neurologic, cutaneous, articular syndrome [279-
281]. These diseases present various symptoms associated with autoimmune disorders including fever, rash, skin lesions, and localized pain. The severity and age-of-onset varies. Each of these may be managed through treatment using IL-1Ra (Anakinra). This treatment method targets the IL-1β receptor, serving as an antagonist and blocking the overactive signaling of the IL-1β pathway [282].
Mice deficient in IL-1β surprisingly have no readily apparent phenotype. This clearly points to the redundancy of proinflammatory cytokine signaling. IL-1β-deficient mice have the same response to LPS challenge as wild-type mice [283]. Closer research revealed IL-1β to be important in inflammation (as one would expect) and multiple publications have established a measurable reduction in several proinflammatory response pathways in control of local and systemic inflammation with IL-1β deficiency
[264].
Non-caspase-1 processing of IL-1β has been documented. Sterile inflammation attributable to IL-1β originating from neutrophils causes an increase in IL-6 secretion and acute phase proteins [284]. This response may be blunted using inhibitors of IL-1β but
78 not caspase-1, indicating a non-caspase-1 mediated mechanism. In this case neutrophil processing of IL-1β is attributed to extracellular processing via proteinase-3 [285]. This proteinase has also been shown to participate in processing of IL-18 [286].
7.4.6. Interleukin-18
Interleukin-18 (IL-18), also known as interferon-gamma inducing factor, is a cytokine belonging to the IL-1 superfamily. Similar to IL-1β, IL-18 is synthesized as a 24 kDa precursor protein that requires caspase-1 mediated cleavage. It too lacks a conventional hydrophobic signal sequence for classical protein secretion. Unlike IL-1β, the IL-18 precursor is constituently synthesized in healthy human and murine blood monocytes, peripheral macrophages, keratinocytes, and endothelial cells lining the gastrointestinal tract [287, 288].
The IL-18 precursor is cleaved by caspase-1 into a 17 kDa active cytokine. Similar to IL-
1β, this maturation is mediated through caspase-1 activation via the inflammasome complex. Upon cleavage, the mature IL-18 protein is secreted from monocytes and macrophages but roughly 80% of IL-18 precursor remains inside the cell [287].
Inhibition of caspase-1 seems to completely abrogate IL-18 processing [289]. Recently, the IL-18 precursor released from dying cells has been shown to be processed extracellular via proteinase-3 [290], much like IL-1β. Fas ligand stimulation of caspase-1 deficient mice still show mature IL-18 secretion [291]. It is evident that caspase-1 is a critical regulator of IL-18 and IL-1β maturation but the scientific literature investigating
79
both cytokines indicates there are other passageways to producing mature IL-18 and IL-
1β cytokines. This is an area that should be further researched as fully understanding
these signaling pathways is critical for treatment of individuals afflicted with the genetic
mutations leading to unbridled proinflammatory cytokine production.
Secreted, mature IL-18 acts as an extracellular ligand through association with its
receptor IL-18 receptor alpha chain (IL-18Rα). Cells that only express this receptor
chain have limited IL-18 sensitivity due to a low-affinity relationship between the ligand
and the receptor. Cells that possess a co-receptor for IL-18, termed IL-18 receptor beta
chain (IL-18Rβ) have increased sensitivity, as the receptor complex with both IL-18Rα
and IL-18Rβ has a high affinity for the IL-18 ligand [287]. With the formation of this
receptor heterodimer, the intracellular TIR domains gain proximity and recruit adaptor
MyD88 as well as four IL-1 receptor-associated kinases and TRAF-6. Afterward, IκB is
degraded and transcription factor NF-κB liberated to translocate to the nucleus and bind
its corresponding response elements and regulate target gene transcription. This signaling cascade is very similar to the IL-1β process, as well as all other IL-1
superfamily cytokines. For comparison, IL-1β is able to stimulate cells at concentrations
of ng/mL range while IL-18 requires much greater concentrations in the 10-20 ng/mL
range of cells with greatest sensitivity (expressing both IL-18Rα and IL-18Rβ) [287, 292,
293].
80
A review published by Dinarello provides an excellent summary on mice deficient in IL-
18. Briefly, he describes IL-18-deficient mice that “overeat, become obese, and exhibit lipid abnormalities, increased atherosclerosis, insulin resistance, and diabetes mellitus reminiscent of the metabolic syndrome” [287, 294]. Remarkably, the insulin resistance may be alleviated with recombinant IL-18 treatment. Further investigation of the hyperphagia indicates it exerts its effects independent of the leptin signaling pathway as
IL-18-deficeint mice respond typically when given an exogenous leptin dose [295].
7.4.7. Interleukin-6
The cytokine Interleukin-6 (IL-6) belongs to the IL-6 family of cytokines which also includes leukemia inhibitory factor, oncostatin M, ciliary inhibitory factor, cardiotropin-
1, cardiotropin-like related cytokine and stimulating neurotrophin-1/B-cell stimulating factor 3, neuropoietin, Interleukin-27 , and Interleukin-31 [296]. IL-6, a 26 kDa compound, was discovered in 1986 as a B-cell differentiation factor. It facilitated the differentiation of activated B-cells into immunoglobulin-producing cells [297] and was originally termed B-cell stimulatory factor-2. It was later discovered to be identical to an already known compounds interferon-β2 [298], 25 kDa protein [299], hybridoma/plasmacytoma growth factor [162], and hepatocyte-stimulatory factor [162].
IL-6 is a glycosylated protein with a four-helix bundle structure typical to the IL-6 cytokine family [296]. This bundle structure is composed to four α-helices (A, B, C, D).
It has three binding regions designated sites 1, 2, and 3. The first, comprised of the C-
81 terminal residues of α-helix D and the C-terminal of the AB loop, binds to membrane bound interleukin-6 receptor (mIL-6R), the second site is located between α-helices A and C which binds gp130 (contacting between domains 2 and 3 of gp130), and the third site is located at the N-terminus of the AB loop which also binds gp130 (contacting domain 1 of gp130) [296].
IL-6 is secreted by many cell types including T-cells, B-cells, monocytes, fibroblasts, keratinocytes, endothelial cells, mesangial cells, adipocytes, and various tumor cells
[300]. IL-6 exerts its effects on cells through association with a receptor system complex comprised of two receptors. The first is mIL-6R (80kDa) which is type 1 transmembrane glycoprotein also known as IL-6Rα, CD126 and gp80 [296]. mIL-6R is thought to stabilize the receptor complex [301]. Interestingly, human IL-6 is able to bind both the human and murine mIL-6R while the murine IL-6 is only able to associate with the murine mIL-6R [302]. The second receptor in the IL-6 receptor complex system is a type
1 transmembrane signal transducer protein gp130 also known as IL-6Rβ and CD130
[296]. IL-6R is the non-signaling α-receptor target for IL-6 and upon ligand binding, β- receptor gp130 dimerizes with the complex and is induced [300]. Membrane bound gp130 is expressed in many cell lines and tissues throughout the body and is not restricted to cells known to be responsive to IL-6 [303]. mRNA transcripts were observed in myeloid leukemia cells, B cell lymphoma cells, plasmacytoma cells, myelomonocytic leukemia cells [304]. Murine tissues expressing gp130 mRNA include brain, heart, thymus, spleen, kidney, lung, and liver [304].
82
Many type 1 and type 2 transmembrane proteins exist in a soluble form that contains the
majority of the extracellular domain of its membrane bound partner [305]. Production of
these soluble receptor proteins can occur through translational differences from
alternatively spliced mRNAs [305]. They may also be produced through cleavage of the
membrane bound receptor protein by metalloproteinases, another membrane bound
protein [306]. The soluble IL-6 receptor (sIL-6R) can be produced by both of these
mechanisms [307, 308]. Many soluble receptors function as an antagonist to their complimentary membrane bound receptor, inhibiting their biological activity [309]. That is not the case with sIL-6R, as its role is to stimulate IL-6 signaling in cells. sIL-6R will mediate the IL-6 receptor complex signaling through the IL-6-sIL-6R-gp130 complex in cells that lack mIL-6R [305]. This is known as trans-signaling (Figure 7.1).
If gp130 is expressed ubiquitously throughout the body, potentially the IL-6-sIL-6R-
gp130 complex could signal throughout the body as well. A soluble gp130 receptor
exists in high concentrations in the blood and acts as an antagonist to this receptor complex, effectively weakening IL-6 sensitivity in most cells [310]. In the few cell types
that do express mIL-6R such as macrophages, neutrophils, T-cells, and hepatocytes the
classic-signaling complex of IL-6-mIL-6R-gp130 facilitates IL-6 signaling [296] (Figure
7.1).
83
Figure 7.1. IL-6 classic-signaling and IL-6 trans-signaling. IL-6 Classic-signaling requires membrane bound IL-6R and is restricted to hepatocytes, some epithelial cells and some leukocytes. IL-6 trans-signaling requires sIL-6R and is possible on all cells of the body since all cells express the gp130 protein. Reprinted with permission from: Rose-John S. IL-6 trans-signaling via the soluble IL-6 receptor: importance for the pro-inflammatory activities of IL-6. Int J Biol Sci. 2012; 8:1237–1247 doi: 10.7150/ijbs.4989. Distributed under the Creative Commons Attribution- NonCommercial-NoDerivatives 4.0 International License.
84
In the classic-signaling complex, IL-6 facilitates mIL-6R and gp130 dimerization, gp130
lacks an intrinsic kinase domain to facilitate signaling itself but instead contains
recruitment sites for the non-receptor tyrosine kinase Janus kinase (JAK) allowing for
proximity mediated transactivation of JAKs to induce phosphorylation of tyrosine
residues of gp130 and begin signal transduction [311, 312].
There are two signaling pathways mediated by gp130 JAK activation. The first is known
as the gp130 Tyr759-derived SHP-2/ERK (extracellular-signal-regulated kinase) MAPK
(mitogen-activated protein kinase) pathway. Phosphorylated gp130 residue Tyr759
recruits SHP-2 which is phosphorylated by JAKs and interacts with Grb2 (growth-factor-
receptor-bound protein 2) and the GDT/GTP exchanger for Ras, Sos (son-of-sevenless)
[300]. Ras then signals the ERK-MAPK signaling cascade activating transcription
factors C/EBP (CCAAT/enhancer-binding protein β) to effect transcription activity target genes [300, 313].
The second signaling pathway mediated by gp130 JAK activation is through
YXXQ/YXPQ motifs of gp130 recruitment of STAT proteins that are then targeted for phosphorylation by JAKs. Phospho-STATs heterodimerize (STAT1-STAT3) or homodimerize (STAT1-STAT1, STAT3-STAT3) and translocate to the nucleus to function as transcription factors for target gene response elements [300]. Suppressor of cytokine signaling proteins is a transcriptional target of JAK/STAT signaling and as their name suggests, negatively regulate the JAK/STAT pathway. Suppressor of cytokine
85
signaling contain SH2 domain and depending on its composition selectively targets
tyrosine residues of cytokine receptors. For example, SOCS3 targets the phosphorylated
tyrosine residue 757 of gp130 and functions as an E3 ubiquitin ligase facilitating protein
degradation [314].
IL-6 signaling has been shown to play a role in liver regeneration. IL-6 deficient mice
are more susceptible to hepatic injury induced by CCl4 [120], FasL [315], ethanol [316],
Con-A [317], and acetaminophen [318]. IL-6 deficient mice also display an impaired ability to regenerate their liver after 2/3PH [319]. Due to IL-6 signaling through classic-
signaling and trans-signaling mechanisms, it is necessary to study each mechanisms
contribution to hepatocellular proliferation. Peters et al. designed an IL-6/sIL-6R fusion
protein (Hyper-IL-6) that mimics IL-6 trans-signaling by directly stimulating gp130 in
the absence of IL-6R. After partial hepatectomy, they demonstrated that Hyper-IL-6, but
not IL-6 alone, led to early induction of hepatocyte proliferation implicating IL-6 trans-
signaling in the regulation of liver regeneration [320]. J. Gewiese-Rabsch et al. used a soluble gp130 protein (sgp130Fc) that inhibits IL-6 trans-signaling to further explain the role IL-6 signaling in liver regeneration after damage induced by hepatotoxin CCl4 [321].
IL-6 knock-out mice have been shown to be more sensitive to CCl4-induced liver damage
[322]. Blockade of only trans-signaling through sgp130Fc in CCl4 challenged mice leads
to increased ALT and aspartate aminotransferase (AST) levels in blood, increased
morphological liver damage (cellular necrosis) visualized through hematoxylin and eosin
(H&E) staining, and less intact cell nuclei visualized through 4',6-diamidino-2-
phenylindole staining [321]. These studies demonstrate the importance of IL-6 trans-
86
signaling in the hepatocellular response to chemically and surgically induced liver
damage.
7.5. NLRC4 Inflammasome
7.5.1. Overview
NOD-like receptor containing a caspase activating and recruitment domain 4 (NLRC4) is
activated in response to sensing bacterial intrusion into the cellular cytosol. Specific
Neuronal apoptosis inhibitory protein (NAIP) proteins bind various bacterial products
within the cytosol and then associate with NLRC4 inducing a conformational change in
NLRC4 leading to the assembly of the multimeric NLRC4 inflammasome complex. This
active complex then activates caspase-1, setting off a molecular cascade of events leading
to the secretion of inflammatory cytokines and/or cellular pyroptosis [229].
7.5.2. Structure and Assembly of the NAIP/NLRC4 Inflammasome
NLRC4 contains a N-terminal CARD, a central NACHT domain, and a C-terminal
leucine-rich repeat (LRR) domain [323] (Figure 7.2). NLRC4 interacts directly with caspase-1 through the CARD domain [324]. NACHT is an acronym combining NAIP
(neuronal apoptosis inhibitor protein), C2TA (MHC class 2 transcription activator), HET-
E (incompatibility locus protein from Podospora anserina) and TP1 (telomerase- associated protein) [325]. The NACHT domain is a member of the NTPase superfamily and is thought to mediate ATP-dependent oligomerization of the NLR protein [326]. As
87
documented in other NLR family proteins, deletion of the LRR domain leads to a
constitutively active form of NLRC4 [327].
The crystal structure of mouse NLRC4 was recently solved by Hu et al. describing the
entire protein except the N-terminal CARD domain [328]. The NLRC4 protein deficient
in CARD showed no defect in protein folding and the overall structure is described as an
inverted question mark [328]. It is comprised of a nucleotide-binding domain (NBD), a
helical domain designated HD1, a winged-helix domain, followed by a NBD, a helical
domain designated HD2, and then the LRR [328]. Three residues, His443 in the winged-
helix domain, Gly520 in the HD2, and Tyr617 in the LRR, result in constitutive
activation of IL-1β when mutated [328]. It was verified that the NLRC4 protein was an active ATPase and through mutation analysis implicated HD1 and LRR domains as potential inhibitors of the NBD that upon ligand binding undergo a conformational
change exposing the NBD [328].
88
Figure 7.2. Overall Structure of mNLRC4. The overall structure of mNLRC4ΔCARD shown in cartoon. The structural domains of mNLRC4ΔCARD are labeled, and the numbers following their labels indicate their boundaries. The bound ADP molecule is shown in stick and cyan. Some of the structural elements are labeled. The dashed line indicates the disordered region (residues 1011 to 1014). “N” and “C” represent N terminus and C terminus, respectively. From Hu Z. et al. Crystal structure of NLRC4 reveals its autoinhibition mechanism. Science. 2013 Jul 12;341(6142):172-5. doi: 10.1126/science. Reprinted with permission from AAAS.
89
Bacteria L. pneumophila and S. typhimurium are gram-negative flagellum bacteria. The flagellin of these bacteria are recognized by the extracellular membrane bound sensor called Toll-like receptor 5 (TLR5) [329]. Flagellin may also be injected into the cellular cytosol using a protein appendage found in various gram-negative bacterial secretion systems. S. typhimurium uses a type III secretion system salmonella pathogenicity island
1 to translocate proteins into eukaryotic cells [330] while L. pneumophila uses Type IV secretion system (T4SS) [331]. NAIP5 binds cytosolic flagellin of L. pneumophila using three leucine residues presence on the bacterial flagellin [332]. NAIP5 seems to be only partially responsible for the detection of S. typhimurium flagellin as NAIP5-/- mice were still able to generate an NLRC4-mediated response in macrophages [333]. It is suspected that NAIP6 could be responsible for the NLRC4-mediated response to S. typhimurium flagellin in the NAIP5-/- mice since NAIP6 has been shown to associate flagellin [229].
NAIP1 and NAIP2 mediate a NLRC4 inflammasome activation through sensing bacterial needle and inner rod protein PrgJ of the T3SS [209, 334]. There is only one known
NAIP in humans and it does not sense bacterial flagellin but appears to sense T3SS needle protein [209]. Small interfering RNA (siRNA) targeted knockdown of NLRC4 in human cells elicit enhanced bacterial growth, therefore implicating NLRC4 in sensing bacterial flagellin in humans [335].
In order for NAIPs to associate with NLRC4, they require NOD domain-associated α- helical domains, therefore this region is essential for NAIP/NLRC4 interaction but the exact molecular interaction between the two remains elusive [229]. The leucine rich repeat region (LRR) of both NAIPs and NLRC4 have been implicated in serving an
90 autoinhibition role [202, 336]. NAIP and NLRC4 interaction leads to conformational protein changes allowing for NLRC4 oligomerization and activation of the inflammasome complex [337]. The active NAIP/NLRC4 inflammasome leads to the recruitment and autoproteolytic cleavage of caspase-1. Detection of bacterial flagellin and T3SS rod component PrgJ induces the phosphorylation of Nlrc4 as Serine residue
533. This step is not sufficient to activate Nlrc4, NAIP/NLRC4 interaction is necessary to activate the NLRC4 inflammasome and caspase-1 activation [338].
The NLRC4 inflammasome functions with or without the adapter protein ASC [339].
ASC usually serves as a bridge for inflammasome activation of caspase-1 though its N- terminal PYD and a C-terminal CARD [340]. Since NLRC4 has a CARD it technically does not require ASC for caspase-1 recruitment, but there is evidence that an ASC- independent NLRC4 inflammasome activates caspase-1 without autoproteolysis, resulting in primarily stimulation of the pyroptotic pathway [339]. Recruitment of ASC to the Nlrc4 inflammasome has been shown to greatly enhance the maturation and secretion of IL-1β [341].
7.5.3. NLRC4-Mediated Pyroptosis
Inflammasome assembly is capable of activating caspase-1, leading to cleavage and activation of proinflammatory cytokines IL-1β and IL-18 [342]. The immune systems phagocytic attack of foreign trespassers and infected cells may not always be enough to clear infection. Evolutionarily conserved mechanisms for cells to detect infection and
91 undertake a programmed self-destruction benefits the host as it cuts short the infecting agents replicative cycle within the cell as exposes them to immune cells for attack [258].
Evolutionary studies indicate that cell death as a consequence of inflammasome activation was the principal mechanism for host defense, and inflammatory cytokine processing evolved out of this process [343].
Cell death through pyroptotic and apoptotic mechanisms both require the activation of caspases. In regard to pyroptosis, caspase activation is facilitated through the activated inflammasome complex’s recruitment and cleavage of capsase-1. NLR family proteins containing a PYD must first recruit adapter protein ASC to mediate a bridge complex, for example NLRP3-ASC-Caspase-1. The Nlrc4 inflammasome is somewhat unique in this activation as Nlrc4 itself contains a CARD to directly recruit caspase-1 and has been shown to function independent of ASC and capable of initiating pyroptosis. However, data has shown that the Nlrc4 inflammasome requires ASC to secrete active IL-1β [344].
This originally sounded implausible since IL-1β processing is achieved through cleaved caspase-1, as would the initiation of pyroptosis. How could the inflammasome activate caspase-1 differently to stimulate pyroptosis without stimulating IL-1β (and IL-18) processing? This was potentially solved recently as data indicates that NLRC4-induced pyroptosis does not require ASC or cleavage of caspase-1. Broz et al. showed that Nlrc4- mediated pyroptosis and cytokine processing occur in independent, distinct regions of the cell. ASC foci (requiring ASC) in the cytosol appear to be the location of cytokine processing while pyroptosis occurs in a diffuse pattern throughout the cell. They also demonstrate the Nlrc4-mediated pyroptosis is achieved through an uncleaved, but
92
catalytically active caspase-1 [339]. In mouse bone marrow derived macrophage cells
deficient in ASC, pyroptosis occurs at an accelerated rate upon L. pneumophila infection
compared to cells with ASC [345]. These results indicate that the activation of capase-1
in the presence of ASC not only potentiates cytokine processing but attenuates pyroptosis, acting in some sort of regulatory fashion.
7.5.4. Current Model
The current model of NLRC4 inflammasome assembly and activation postulates ligand binding, facilitated through NAIPs, to the LRR domain of NLRC4 leads to the exchange of adenosine diphosphate (ADP) for adenosine triphosphate (ATP). This then induces a conformation change consenting oligomerization with other active NLRC4 molecules to assemble an active inflammasome [328, 346-348].
93
CHAPTER 8
PUBLICATION
8.1. DeSantis DA, Lee P, Doerner SK, Ko CW, Kawasoe JH, Hill-Baskin AE, Ernest SR,
Bhargava P, Hur KY, Cresci GA, Pritchard MT, Lee CH, Nagy LE, Nadeau JH,
Croniger CM. Genetic resistance to liver fibrosis on A/J mouse chromosome 17.
Alcohol Clin Exp Res. 2013 Oct;37(10):1668-79.
94
8.1.
GENETIC RESISTANCE TO LIVER FIBROSIS ON A/J MOUSE
CHROMOSOME 17
David A. DeSantis1, Peter Lee1, Stephanie K. Doerner2, Chih-Wei Ko1, Jean H.
Kawasoe2, Annie E. Hill-Baskin2, Sheila R. Ernest2, Prerna Bhargava3, Kyu Yeon
Hur3, Gail A. Cresci4, Michele T. Pritchard5, Chih-Hao Lee3, Laura E. Nagy1,4,
Joseph H. Nadeau6 and Colleen M. Croniger1
Departments of 1Nutrition, 2Genetics, Case Western Reserve University School of Medicine, Cleveland, Ohio 3Department of Genetics and Complex Diseases, Harvard University School of Public Health, Boston, Massachusetts 4Departments of Pathobiology and Gastroenterology, Cleveland Clinic Foundation, Cleveland, Ohio 5Departments of Pharmacology, Toxicology and Therapeutics, The University of Kansas Medical Center, Kansas City, Kansas 6Institute for Systems Biology, Seattle, Washington
Address Correspondence to: Colleen M. Croniger, PhD, Department of Nutrition, Case
Western Reserve University, 10900 Euclid Ave, Cleveland, OH 44106; Tel.: 216-368-
4967; Fax: 216-368-6644; E-mail: [email protected]
95
Keywords
C57BL/6J; A/J; Congenic Strains; Liver Fibrosis; Nlrc4 Inflammasome; CCl4
Acknowledgements
This work was supported by the National Institute on Alcohol Abuse and Alcoholism
(NIAAA) P-20 grant AA017837 (LEN, CMC), AA017918 and COBRE grant
8P20GM103549-07 (MTP), 1F32AA021044-02 (GAC), and a grant from Howard
Hughes Medical Institute for Undergraduate Research (PL).
96
Abstract
Background
Because the histological and biochemical progression of liver disease is similar in
alcoholic steatohepatitis (ASH) and nonalcoholic steatohepatitis (NASH), we
hypothesized that the genetic susceptibility to these liver diseases would be similar. To
identify potential candidate genes that regulate the development of liver fibrosis, we
studied a chromosome substitution strain (CSS-17) that contains chromosome 17 from
the A/J inbred strain substituted for the corresponding chromosome on the C57BL/6J
(B6) genetic background. Previously, we identified quantitative trait loci (QTLs) in CSS-
17, namely obesity-resistant QTL 13 and QTL 15 (Obrq13 and Obrq15, respectively),
that were associated with protection from diet-induced obesity and hepatic steatosis on a
high-fat diet.
Methods
To test whether these or other CSS-17 QTLs conferred resistance to alcohol-induced liver
injury and fibrosis, B6, A/J, CSS-17, and congenics 17C-1 and 17C-6 were either fed
Lieber–DeCarli ethanol (EtOH)-containing diet or had carbon tetrachloride (CCl4) administered chronically.
Results
97
The congenic strain carrying Obrq15 showed resistance from alcohol-induced liver injury and liver fibrosis, whereas Obrq13 conferred susceptibility to liver fibrosis. From published deep sequencing data for chromosome 17 in the B6 and A/J strains, we identified candidate genes in Obrq13 and Obrq15 that contained single-nucleotide polymorphisms (SNPs) in the promoter region or within the gene itself. NADPH oxidase organizer 1 (Noxo1) and NLR family, CARD domain containing 4 (Nlrc4) showed altered hepatic gene expression in strains with the A/J allele at the end of the EtOH diet study and after CCl4 treatment.
Conclusions
Aspects of the genetics for the progression of ASH are unique compared to NASH, suggesting that the molecular mechanisms for the progression of disease are at least partially distinct. Using these CSSs, we identified 2 candidate genes, Noxo1 and Nlrc4, which modulate genetic susceptibility in ASH.
98
Introduction
The main causes of liver disease in industrialized countries include alcoholic
steatohepatitis (ASH), chronic hepatitis C virus (HCV) infection, and nonalcoholic
steatohepatitis (NASH; Gines et al., 2004). In patients who chronically consume alcohol
or are morbidly obese, the progression of liver disease is similar, originating with development of a fatty liver (hepatic steatosis) and then progresses to hepatitis, fibrosis, and finally cirrhosis. However, of those patients that develop hepatitis, only a small subset will progress further to cirrhosis. Approximately 20 to 50% of patients with ASH
and ~20% of patients with NASH will progress to cirrhosis (Diehl et al., 1988)
suggesting a gene–environment interaction contributes to disease progression.
Although most genetic screens focus on the behavioral aspect of alcohol dependence,
several investigated the genetic susceptibility to liver injury (Kimura and Higuchi, 2011;
Lin et al., 2012). Recently, both genetic and environmental risk factors are known to
influence not only addictive behavior but also the severity of steatosis, hepatitis, and liver
fibrosis. In human studies, single-nucleotide polymorphisms (SNPs) have been identified
in genes known for the metabolism of alcohol, such as alcohol dehydrogenase (Edenberg
et al., 2006; Xuei et al., 2006), aldehyde dehydrogenase (Vasiliou et al., 2000), and
cytochrome P450, family 2, subfamily E, polypeptide 1 (CYP2E1; Itoga et al., 2001;
Webb et al., 2011). This list of candidate genes for liver fibrosis is not comprehensive,
but it supports the concept that polymorphisms in multiple genes impact the progression
of liver injury and disease. In addition, there are other mechanisms besides genetic
modifications (such as SNPs) that could modulate development of disease. For example,
99 epigenetic regulation and microRNAs could also be genetically regulated and impact disease progression.
Using an unbiased approach to find genes that modulate the development of fibrosis, we surveyed chromosome substitution strains (CSSs) containing chromosome 17 from the
A/J inbred strain that was substituted for the corresponding chromosome on the
C57BL/6J (B6) genetic background. CSS-17 was chosen because this strain is resistant to obesity and steatosis on a high-fat, simple-carbohydrate (HFSC) diet, and we previously located several quantitative trait loci (QTLs) associated with obesity and hepatic steatosis
(Millward et al., 2009). Two of the identified QTLs associated with resistance to diet- induced obesity and hepatic steatosis on HFSC diet are Obrq13 and Obrq15 (s 8.1,
Table 8.2; Millward et al., 2009). The primary difference between these QTLs derived from A/J alleles is that Obrq15, develops insulin resistance on the HFSC diet, whereas Obrq13 remains insulin sensitive as measured by glucose tolerance tests and calculated homeostasis model of insulin resistance. Because progression of liver disease is similar in NASH and ASH, we hypothesized that the genetic susceptibility to these liver diseases would be the similar. We hypothesized that Obrq13 and Obrq15 would also confer resistance to alcohol-induced liver injury and fibrosis. In this study, congenic strains derived from CSS-17 that contain Obrq13 (called 17C-1) orObrq15 (called 17C-
6) were analyzed for their susceptibility to either alcohol-induced liver injury or carbon tetrachloride (CCl4)-induced liver fibrosis.
100
Experimental Procedures
Husbandry
B6, A/J, and B6-Chr 17A/J/NaJ (CSS-17) mice were generated and maintained at Case
Western Reserve University (Singer et al., 2004). The congenic strains derived from
CSS-17 were generated as previously described (Millward et al., 2009). Mice were raised
in microisolator cages with a 12-hour light/12-hour dark cycle. All mice were weaned at
3 to 4 weeks of age and raised on LabDiet #5010 autoclavable rodent chow (LabDiet,
Richmond, IN) ad libitum until studies were initiated.
Ethanol Feeding Diet Study
Eight- to 10-week-old female B6, A/J, CSS-17, and congenics 17C-1 and 17C-6 were fed
either Lieber–DeCarli ethanol (EtOH)-containing diet (+EtOH) or pair-fed (PF) control diet as previously described in Pritchard and Nagy (2005). For measurements of serum
EtOH concentrations, blood was taken from the tail vein 2 hours into the feeding cycle.
Female mice were used for this study because they are more susceptible to alcohol- induced liver injury and have a significantly higher risk of developing cirrhosis for any given level of alcohol intake (Sato et al., 2001).
CCl4 Administration and Sample Collection
CCl4 (Sigma-Aldrich, St. Louis, MO) administration in male B6, A/J, CSS-17, and
congenics 17C-1 and 17C-6 strains was performed as previously described (Pritchard and
101
Nagy, 2010). We used male mice for this study because our previous study to identify
QTLs for resistance to diet-induced obesity was performed in male mice (Millward
et al., 2009) due to their greater propensity of gaining body weight than females on HFSC
diet (Hong et al., 2009). Therefore, to compare our previous high-fat diet study to
development of liver fibrosis, the CCl4 was administered to male mice.
Histology and Immunohistochemistry
For histological analysis, formalin-fixed tissues were paraffin embedded, sectioned
(5 μm) and stained with Sirius red stain for collagen as previously described (Pritchard
and Nagy, 2010). Formalin-fixed, paraffin-embedded liver sections were deparaffinized
and stained for α-smooth muscle actin (α-SMA) as previously described (Pritchard and
Nagy, 2010). The positive α-SMA areas and Sirius red stained areas were
morphometrically quantified using Image-Pro Plus software (Media Cybernetics,
Bethesda, MD) and analyzed. All images presented in the results are representative of at
least 3 images per liver and 4 mice per experimental condition.
Neutrophils were immunolocalized in liver tissue using an antineutrophil antibody,
NIMP-R14 (Abcam, Cambridge, MA). Briefly, formalin-fixed, paraffin-embedded liver tissues were deparaffinized. Liver sections were then incubated with the NIMP-R14 antibody (1:100) overnight at 4°C. After washing, the tissues were incubated with a biotinylated anti-rat secondary antibody according to the manufacturer's instructions
(Vectastain Elite ABC kit, rat IgG; Vector Laboratories, Burlingame, CA). Subsequently,
102
tissues were incubated with an avidin–biotin–HRP complex (Vectastain Elite ABC kit;
Vector Laboratories). ImmPACT NovaRed peroxidase substrate was utilized to visualize
positive staining in tissues. To quantify neutrophil content in tissues, 10 nonoverlapping
images of each section were captured and neutrophil presence was graded as none, rare,
few, or moderate. These subjective assessments were assigned a number (none = 0, rare = 2, few = 4, moderate = 6), and each section's numbers were averaged. The averages (a single number per mouse) were used for statistical analysis.
Measurement of Hepatic Triglycerides, Plasma EtOH, and Plasma Alanine
Aminotransferase Concentrations
For measurement of liver triglycerides, 100 to 200 mg of liver was saponified with an
equal volume by weight of 3 M KOH/65% EtOH as described (Salmon and Flatt, 1985).
We measured glycerol concentration against glycerol standards, using a commercially
available triglyceride glycerol phosphate oxidase reagent kit (Pointe Scientific, Lincoln
Park, MI), as previously described (Buchner et al., 2008). Plasma EtOH concentration
and alanine aminotransferase (ALT) were measured using commercially available
enzymatic assay kits (Sigma-Aldrich), according to manufacturer's instructions.
Isolation of Hepatocytes
Hepatocytes were isolated from B6, 17C-1, and 17C-6 mice by an in situ collagenase
(type VI; Sigma-Aldrich) perfusion technique, as described previously (West
et al., 1989). Hepatocytes were separated from the nonparenchymal cells by 2 cycles of
103
differential centrifugation (50×g for 2 minutes) and further purified over a 30% Percoll
gradient. The purity of these hepatocyte cultures exceeded 98% by light microscopy, and
viability was typically more than 95% by trypan blue exclusion assay. The cells were
isolated, and mRNA was harvested without culturing.
Isolation of Hepatic Stellate Cells
Mouse hepatic stellate cells (HSCs) were isolated from B6, 17C-1, and 17C-6 mice by enzymatic digestion and Percoll density gradient centrifugation, as previously described
(Blomhoff and Berg, 1990). HSCs were isolated, and mRNA was harvested without culturing the cells for the day 0 (quiescent HSC). HSCs were cultured for 7 days
(activated HSCs) in 6-well plates with Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum, glutamine, HEPES buffer, and antibiotics. The cells were incubated at 37°C in an air atmosphere containing 5% CO2. After 7 days, the cells
were isolated and we extracted mRNA for day 7 values (Blomhoff and Berg, 1990).
Isolation of Macrophages
Peritoneal macrophages were obtained by peritoneal lavage from B6, 17C-1, and 17C-6
mice that were injected with 2 ml of 4% thioglycollate. Thioglycollate-elicited peritoneal
macrophages were collected from the peritoneal cavity of mice 3 days after injection.
Macrophages were collected by centrifugation and plated (Corning, Lowell, MA) in
RPMI-1640 medium (Sigma-Aldrich) supplemented in 10% fetal calf serum, 300 mg/l l-
glutamine (Sigma-Aldrich), 100 U/ml penicillin/0.1 mg/ml streptomycin (Sigma-
104
Aldrich), and 1 mM sodium pyruvate (Sigma-Aldrich). The cells were incubated at 37°C in an air atmosphere containing 5% CO2. Unattached cells were removed by refreshing
the medium 4 hours after isolation. After culture for 20 hours, cell culture media were
removed and replaced with fresh media and were stimulated with 100 ng/ml
lipopolysaccharide (LPS). After 4-hour stimulation with LPS, cell culture media were
removed and cells were collected for isolation of RNA.
Isolation of Bone Marrow-Derived Macrophages and Measurement of Reactive Oxygen
Species
Bone marrow-derived macrophages (BMDMs) were differentiated from femoral bone
marrow cells with 20% of L929 conditioned medium in DMEM medium (Sigma-
Aldrich). Every other day, the medium was replaced with the fresh medium. On day 7 to
8, cells were collected with enzyme-free cell dissociation buffer (Gibco Life
Technologies, Grand Island, NY). To measure cellular reactive oxygen species (ROS), the cells were incubated with 10 μm of Cm-H2DCFDA (#C6827; Invitrogen, Grand
Island, NY), in HBSS for 1 hour. Fluorescence was measured by fluorescence-activated
cell sorting analysis. Values were expressed as Mean Fluorescent Intensity × %Gated.
Real-Time Quantitative Reverse Transcription PCR
Total RNA from 30 mg of liver or isolated cells was prepared with an RNeasy Mini Kit
(Qiagen, Valencia, CA) and was synthesized to single-strand cDNA from 500 ng of total
RNA with random hexamer primers and MMTV reverse transcriptase (Applied
105
Biosystems, Foster City, CA). We amplified cDNAs using SYBR Green PCR Core
reagent mix (Applied Biosystems) and performed real-time quantitative reverse
transcription PCR (real-time qRT-PCR) with a Chromo4 Cycler (MJ Research, St. Bruno,
Canada). The relative amounts of mRNA were determined by the ΔCt values
(mean Ct gene of interest and mean Ct from 18S rRNA or 36b4 [housekeeping gene] from the same mouse) as previously described (Millward et al., 2009; Pritchard
et al., 2011).
Protein Isolation and Western Blotting
Proteins were isolated, and Western blot analysis was performed from liver samples as
previously described (Millward et al., 2010). The membranes were incubated with
antibodies to CYP2E-1 (1:10,000; Fitzgerald Industries International, Concord, MA) or
α-SMA (1:5,000; Santa Cruz Biotechnology, Santa Cruz, CA) The immunoreactive
proteins were detected using the Super-Signal Chemiluminescent Substrate® Kit (Thermo
Scientific, Rockford, IL) and measured the density of the immunoreactive bands by
scanning densitometry (UN-SCAN-IT gel software, Orem, Utah). The membranes were
stripped and normalized for loading differences using heat shock cognate-70 (HSC-70)
(1:16,000; Santa Cruz Biotechnology) as previously described (Millward et al., 2010).
Statistical Analysis
106
The values reported are means ± standard error of the mean (SEM). Data were analyzed with Student's t-test or analysis of variance (ANOVA) with Bonferroni correction for multiple testing using GraphPad Prism (GraphPad Software, San Diego, CA).
107
Results
Genetic Susceptibility to Alcohol-Induced Liver Injury
We tested susceptibility of females from B6, CSS-17, and congenic strains (17C-1 and
17C-6) derived from CSS-17 to alcohol-induced liver injury after chronic administration
of Lieber–DeCarli EtOH-containing diet. No differences in daily food intake were found
in all of the strains (Table 1). To ensure that the various strains metabolized EtOH in a
similar manner, blood alcohol levels were measured 2 hours into the feeding cycle. A/J
mice had a greater increase in plasma EtOH concentrations, while all of the other strains
showed similar increased plasma EtOH concentrations compared to B6 mice (Table 1).
Because EtOH consumption induces CYP2E1 expression and activity (Morimoto
et al., 1995), we measured the expression of CYP2E1 with Western blot analysis. All
strains had increased expression of CYP2E1 with EtOH feeding (Fig. 8.2). Therefore, these results indicate that the genetic strains had similar increase in CYP2E1 expression, thereby validating these strains for testing of hepatic responses to perturbations.
Liver injury was characterized as an increase in hepatic triglyceride and an increase in
plasma ALT after EtOH consumption compared to pair-fed controls (Dooley and ten
Dijke, 2012). At the end of the alcohol-diet study, minimal hepatic lipid accumulation was detected in the A/J strain, whereas a similar increase in hepatic triglycerides was found in B6, CSS-17, 17C-1, and 17C-6 strains (Fig. 8.3). Plasma ALTs were measured after feeding of Lieber–DeCarli EtOH-containing diet, and these values were compared to their pair-fed controls (Fig. 8.3). A significant increase in ALTs was found in all
108
strains. Congenic strain, 17C-1, showed the greatest increase in ALTs (3.3-fold) over its pair-fed control, while congenic strain 17C-6 had reduced ALTs compared to B6 mice.
This suggested that the 17C-1 strain contains A/J allele(s) that result in increased susceptibility to alcohol-induced liver injury and 17C-6 strain contains A/J allele(s) that protect against alcohol-induced liver injury.
Genetic Susceptibility to Liver Fibrosis
Frank fibrosis is difficult to induce with alcohol feeding alone. Therefore, to identify genes modulating development of liver fibrosis, we analyzed B6, A/J, CSS-17, 17C-1, and 17C-6 using the established CCl4 protocol (Constandinou et al., 2005). Toxicity
results from the bioactivation of the CCl4 molecule to the trichloromethyl free radical by
cytochrome P450 isozymes (P450s) in the endoplasmic reticulum (Recknagel
et al., 1989), specifically CYP2E1 (Wong et al., 1998). To determine whether metabolism
of CCl4 was comparable in each strain, we measured CYP2E1 protein levels in olive oil
and CCl4-treated mice (Fig. 8.4). All 4 strains had similar CYP2E1 expression levels with
olive oil and they had reduced expression after CCl4 administration, thus supporting
comparable metabolism of CCl4. The 17C-1 strain had dramatically reduced CYP2E1
expression compared to B6 (Fig. 8.4). However, in the alcohol study, 17C-1 strain had
similar expression levels of CYP2E1 with alcohol consumption compared to B6 mice,
suggesting that CCl4 administration may cause more injury than alcohol consumption in the 17C-1 strain.
109
HSCs represent ~15% of the total cells in the liver and are located within the space of
Disse. Activated HSCs show profound phenotypic changes, including enhanced cell proliferation, de novo expression of α-SMA, and overproduction of the extracellular matrix (ECM; Lee et al., 2004). To determine whether mice developed fibrosis, liver sections from these animals were taken 3 days after the final CCl4 injection, when peak
fibrosis is evident (Higashiyama et al., 2007), and they were stained with Sirius red to
detect ECM. Images of the stained livers were scanned and quantified. B6 and 17C-1
mice showed a moderate degree of Sirius red staining, whereas A/J, CSS-17, and 17C-6
mice had noticeably reduced Sirius red staining (Fig. 8.5). The control olive oil injections
showed no liver injury (images not shown). The livers of B6, A/J, CSS-17, 17C-1, and
17C-6 mice were also analyzed by immunohistochemistry (IHC) with α-SMA antibody,
and the images were scanned and quantified as described in Materials and Methods. We
found that B6 mice showed a moderate degree of α-SMA expression, whereas A/J, CSS-
17, and 17C-6 mice had noticeably reduced α-SMA expression, and 17C-1 strain had
significantly more α-SMA expression (Fig. 8.6). α-SMA protein expression was also
measured by Western blot analysis. Congenic strain, 17C-6, had 50% less α-SMA protein
expression compared to B6 CCl4-treated mice, whereas 17C-1 had 87% more expression
(Fig. 8.7). These results suggest that a small region of the A/J chromosome in congenic
strain 17C-6 contains 1 or more genes that protect against liver fibrosis. Conversely, 17C-
1 contains at least 1 gene that results in increased susceptibility to liver fibrosis.
Candidate Genes for Development of ASH
110
Previously, we deep sequenced DNA from the B6 and CSS-17 strains (Sudbery
et al., 2009). We identified candidate genes in Obrq13 and Obrq15 that contained nonconsensus SNPs in 5′ untranslated region or within the gene. The candidate genes
were plasminogen (Plg), solute carrier family 5 (choline transporter), member 7 (Slc5a7),
fibronectin type III domain containing 1 (Fndc1), and NADPH oxidase organizer 1
(Noxo1) for 17C-1 (Obrq13), and DEAD/H (Asp-Glu-Ala-Asp/His) box helicase 11
(Ddx11), thyroid hormone receptor interactor 10 (Trip10), cyclin-dependent kinase-like
(Cdkl4), and NLR family, domain containing 4 (Nlrc4) for congenic 17C-6 (Obrq15;
Table 2). Plg, Fndc1, Slc5a7, Noxo1, Ddx11, Trip10, Cdkl4, and Nlrc4 mRNA
expression were measured in livers from B6, A/J, CSS-17, 17C-1, and 17C-6 mice at the
end of the treatment protocols. Although there were no statistical differences in mRNA
expression for Plg, Ddx11, Trip10, and Cdkl4, we did find increased Fndc1 expression in
17C-1 EtOH-fed mice and 17C-6 pair-fed mice. Slc5a7 was significantly reduced in
EtOH-fed 17C-6 mice (Fig. 8.S1). While these genes had altered expression, they did not
correlate with treatment. We detected altered expression for Noxo1 and Nlrc4 that was
due to both genetics and experimental treatment. After administering the Lieber–DeCarli
EtOH-containing diet, increased expression of Noxo1 was found in mice with the A/J
allele for Noxo1 (Fig. 8.8A), whereas increased Nlrc4 expression was found in both pair-
fed and EtOH-fed strains that contained the A/J allele for Nlrc4 (Fig. 8.8C). To determine
whether these changes in gene expression also occur during fibrosis, expression
of Noxo1 and Nlrc4 was analyzed after CCl4 treatment (Fig. 8.8B,D, respectively). The
congenic strain, 17C-1, which had increased liver injury also showed
111
increased Noxo1 expression after CCl4administration (Fig. 8.8D), whereas 17C-6 had increased expression of Nlrc4 after CCl4 administration.
Cell-Specific Expression and Function of Altered Expression for Noxo1 and Nlrc4
In Fig. 8.8, the expression data were from a whole-liver homogenates and thus contained all of the cell types in the liver (hepatocyte, Kupffer cells, NK cells, endothelial cells, and
HSCs). To determine whether the SNPs in the A/J allele for these genes modulate expression in a cell-specific manner, we isolated hepatocytes, thioglycollate-elicited peritoneal macrophages, and HSCs from B6, 17C-1, and 17C-6 mice fed normal chow diet. In the mice containing the A/J allele for Noxo1, the SNPs did not alter mRNA expression in isolated hepatocytes. However, Noxo1 mRNA expression was significantly increased in isolated macrophages treated with and without LPS compared to B6 macrophages (Fig. 8.9). Noxo1 mRNA expression in cultured BMDM (Fig. 8.S2) was also measured and showed a 10-fold increase in expression. In addition, the mRNA expression of Noxo1 was significantly increased in HSCs at day 0 compared to B6 HSCs.
To determine whether an increase in Noxo1 expression correlates with an increase in oxidative stress, we measured cellular ROS in BMDM from B6, CSS-17, and 17C-1 mice. The 17C-1 strain had dramatically increased ROS compared to B6 BMDM. Thus, we conclude that the SNPs in the A/J allele for the Noxo1 gene increase Noxo1 expression in macrophages and HSCs and result in increased cellular oxidative stress, which may contribute to increased fibrosis in 17C-1 mice.
112
In the mice containing the A/J allele for Nlrc4, SNPs did not alter expression in the isolated hepatocytes (Fig. 8.10). In HSCs, Nlrc4 mRNA expression was reduced compared to B6 HSCs. However, Nlrc4 expression was significantly increased in isolated thioglycollate-elicited peritoneal macrophages treated with and without LPS compared to
B6 macrophages. Because Nlrc4 is part of the inflammasome, we wanted to determine whether increased Nlrc4 expression correlated with increased inflammation. Therefore, neutrophil infiltration was measured with an antibody for NIMP-R14 in liver sections from B6 and 17C-6 mice. There was an increase in NIMP-R14 positive cells in 17C-6 strain compared to B6 and CSS-17 mice. As a result, we conclude that the SNPs in the
A/J allele for the Nlrc4 gene increased Nlrc4expression in macrophages and result in increased hepatic inflammation in the 17C-6 strain.
113
Discussion
The aim of the present study was to identify novel candidate genes involved in
development of liver injury and fibrosis. B6, A/J, CSS-17 and 17C-1 and 17C-6 congenic
strains were analyzed for hepatic steatosis and liver injury in response to chronic alcohol
feeding or administration of CCl4. We hypothesized that because progression of liver
disease from steatosis to cirrhosis is similar for patients with ASH and NASH, the genes
and pathways that were previously identified on a high-fat diet would be similar for
alcohol consumption and CCl4treatment. However, we found that Obrq13 promoted liver injury and fibrosis, and Obrq15 protected mice from alcohol-induced liver injury and fibrosis. Therefore, genetics and the molecular mechanisms for progression to ASH and
NASH appear to have distinctive features (Table 2).
Using the congenic strains derived from CSS-17, the consequences of a very small region
of the A/J chromosome 17 in context of the B6 chromosome were studied. This illustrates
one of many advantages in using CSS strains and congenic strains over a traditional F1
cross between B6 and A/J inbred strains. The isogenic A/J chromosome in CSSs enables
us to: (i) identify both dominant and recessive genes modulating liver injury, (ii) maintain
a genetically identical mouse strain for continued functional and mechanistic studies, and
(iii) decouple phenotypes to investigate their independent contributions to disease. With
the CSS paradigm, several candidate genes that were located in either Obrq13
(Fndc1, Plg, Slc5a7, and Noxo1) or Obrq15 (Trip 10, Ddx11, Cdkl4, and Nlrc4) were
identified. Little is known about the role of these genes in the liver. Fndc1 has been found
to be elevated in patients with skin tumors (Anderegg et al., 2005). Increased choline
114
transporter expression and activity from Slc5a7 have been found in patients with
Alzheimer's disease (Bissette et al., 1996). In the Fig. 8.S1, the expression of these genes
was not altered between strains except for Fndc1 and Slc5a7. The role of these genes in
the development of liver disease is not clear and requires more analysis. In this study, we focused on the 2 candidate genes that had altered gene expression due to genetic variance and experimental treatments.
Patients with chronic liver injury from alcohol or HCV infection have increased ROS and
oxidative stress (Parola and Robino, 2001). In the hepatocyte, ROS is generated from
CYP2E1 activation with alcohol consumption and the apoptotic bodies from damaged hepatocyte can activate quiescent HSCs to become activated HSCs that produce collagen and proliferates. ROS can also be produced by plasma membrane-associated NADPH oxidase (Babior et al., 2002). In the resident macrophage of the liver, Kupffer cells,
NADPH oxidase reduces molecular oxygen to generate superoxide, which is converted to hydrogen peroxide (Babior et al., 2002). The activation of oxidants increases redox-
sensitive transcription factor, nuclear factor kappa–light chain enhancer of activated B cells and increases expression of critical cytokines, such as tumor necrosis factor α
(TNFα) and interleukins. These cytokines activate a signaling cascade in the hepatocyte and ultimately cause apoptosis and can activate the HSC (Wheeler et al., 2001).
Using the CSS paradigm, we identified 2 candidate genes, Noxo1 and Nlrc4, that modulate liver injury. Identification of Noxo1 supports published evidence that the NOX
115 family of NADPH oxidases contributes to liver disease (Bataller et al., 2003). This gene family of proteins transfers electrons across biological membranes, and the product of the electron transfer reaction is superoxide (Ushio-Fukai, 2006).
The NADPH oxidase is a multisubunit enzyme that has catalytic subunits as well as regulator subunits called “organizer subunits” (Bedard and Krause, 2007). NADPH oxidase activity is controlled by regulatory subunits including Nox regulators p47phox, p67phox, and their homolog's Noxo1 (our gene of interest), NoxA1 or Duox1A1 or Duoax2. The impact that these regulatory subunits have on modulating NADPH oxidase activity has been minimally studied. One study showed mice that deficient for p47phox (p47phox−/−) are resistant to chronic EtOH-induced liver injury (Kono et al., 2000).
In this study, we show that increased expression of regulatory subunit, Noxo1, in BMDM isolated from 17C-6 resulted in greater cellular ROS. We propose that the SNPs in theA/J allele of Noxo1 increase gene expression of the regulator, Noxo1, thus increasing the activity of NADPH oxidase. This in turn would generate more oxidative stress in macrophages and activate HSCs resulting in greater fibrosis for the 17C-1 strain. There may also be cross-talk between the macrophages, HSCs, and hepatocytes that were not detected in our primary cultured cells. Future studies will analyze altered expression of Noxo1 in coculture experiments to determine the impact of altered macrophage expression of Noxo1 and NADPH oxidase on HSC function.
116
The second gene we identified was Nlrc4, which supported a role for the immune system
in development of liver fibrosis. Presently, the current hypothesis for EtOH-induced liver
injury proposes that EtOH consumption may result in leakage of bacterial products from the gut, altering the jejunal microflora and leading to proliferation of gram-negative bacteria. LPS, a component of the cell wall in gram-negative bacteria, is increased in circulation of alcoholics (Wang et al., 2010). The increased LPS induces TNFα production from Kupffer cells. In addition, recognition of conserved microbial structures known as pathogen-associated molecular patterns (PAMPs) is accomplished by membrane-bound Toll-like receptors and cytoplasmic nucleotide oligomerization domain-like receptors (NLRs; Schroder and Tschopp, 2010). Upon sensing of PAMPs,
NLRs interact with members of the inflammasome complex, and the assembly of the inflammasome complex leads to cleavage and activation of procaspase-1 (Schroder and
Tschopp, 2010). Once activated, caspase-1 promotes proteolytic maturation and activation of interleukin (IL)-1β, IL-18, and caspase-7 and deactivation of IL-33
(Lamkanfi et al., 2007) as well as mediating apoptotic cell death (Bergsbaken et al., 2009).
Nlrc4 mRNA was elevated in isolated macrophage, yet it was decreased in isolated HSCs
(day 0 and day 7 of culture) from 17C-6 mice (Fig. 8.10). This increased expression of Nlrc4 was associated with induced inflammation as measured by infiltration of neutrophils recognized with NIMP-R14 IHC. We propose that the increase in Nlrc4 in the
macrophage and decrease in Nlrc4 mRNA expression in the HSC are protective from
117
CCl4-induced fibrosis. The mechanism is not understood, but we hypothesize that increased Nlrc4 expression induces Nlrc4 inflammasome activity. This in turn would
increase activation of caspase-1 activity and release of cytokines, IL-1β and IL-18. IL-18
has been shown to induce expression of interferon γ (IFN-γ) in macrophages (Barbulescu
et al., 1998; Baroni et al., 1996; Nakahira et al., 2002). Previous studies found that IFN-γ
administration to HSCs inhibit in vitro activation of HSCs and caused a marked decrease
in collagen synthesis (Czaja et al., 1987). IFN-α has also been shown to inhibit
proliferation of HSCs and induce apoptosis of HSCs (Ogawa et al., 2009). Another
potential mechanism could be a direct effect on hepatocyte regeneration via IL-18 and
IL-6. Studies in mice reveal that IL-18 stimulation of peritoneal macrophages induces IL-
6 production (Netea et al., 2000). We suggest that the increase in Nlrc4expression in the
macrophage would increase IL-18, which in turn could increase IL-6 production. IL-6
can bind to its receptor in the hepatocyte and initiate a signaling cascade that would
promote liver regeneration (Taub, 2004). Future studies will look at the cross-talk
between macrophage and hepatocytes and will be used to understand the mechanism for
protection from liver injury by Nlrc4.
In summary, we found altered gene expression of the A/J-derived alleles
for Noxo1 and Nlrc4 compared to B6 alleles. We suggest that sequence differences in
regulatory or coding regions of these genes modulate development, maintenance, and
resolution of fibrosis. Future studies will address these issues using congenic strains and
knockout animals for each gene and analyze the role of these genes within Kupffer cells,
HSCs, and hepatocytes and the cross-talk between these cells. Elucidation of the
118 molecular mechanisms behind these genes will improve our fundamental knowledge of the mechanism(s) for progression to fibrosis and will promote the development of new therapeutic interventions to prevent ASH.
119
References
Anderegg U, Breitschwerdt K, Kohler MJ, Sticherling M, Haustein UF, Simon JC, Saalbach A (2005) MEL4B3, a novel mRNA is induced in skin tumors and regulated by TGF-beta and pro-inflammatory cytokines. Exp Dermatol 14:709–718. Babior BM, Lambeth JD, Nauseef W (2002) The neutrophil NADPH oxidase. Arch Biochem Biophys 397:342–344. Barbulescu K, Becker C, Schlaak JF, Schmitt E, Meyer zum Buschenfelde KH, Neurath MF (1998) IL-12 and IL-18 differentially regulate the transcriptional activity of the human IFN-gamma promoter in primary CD4 + T lymphocytes. J Immunol160:3642–3647. Baroni GS, D'Ambrosio L, Curto P, Casini A, Mancini R, Jezequel AM, Benedetti A (1996) Interferon gamma decreases hepatic stellate cell activation and extracellular matrix deposition in rat liver fibrosis. Hepatology 23:1189–1199. Bataller R, Schwabe RF, Choi YH, Yang L, Paik YH, Lindquist J, Qian T, Schoonhoven R, Hagedorn CH, Lemasters JJ, Brenner DA(2003) NADPH oxidase signal transduces angiotensin II in hepatic stellate cells and is critical in hepatic fibrosis. J Clin Invest112:1383–1394. Bedard K, Krause KH (2007) The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology. Physiol Rev87:245–313. Bergsbaken T, Fink SL, Cookson BT (2009) Pyroptosis: host cell death and inflammation. Nat Rev Microbiol 7:99–109. Bissette G, Seidler FJ, Nemeroff CB, Slotkin TA (1996) High affinity choline transporter status in Alzheimer's disease tissue from rapid autopsy. Ann N Y Acad Sci 777:197– 204. Blomhoff R, Berg T (1990) Isolation and cultivation of rat liver stellate cells. Methods Enzymol 190:58–71. Buchner DA, Burrage LC, Hill AE, Yazbek SN, O'Brien WE, Croniger CM, Nadeau JH (2008) Resistance to diet-induced obesity in mice with a single substituted chromosome. Physiol Genomics 35:116–122. Constandinou C, Henderson N, Iredale JP (2005) Modeling liver fibrosis in rodents. Methods Mol Med 117:237–250. Czaja MJ, Weiner FR, Eghbali M, Giambrone MA, Zern MA (1987) Differential effects of gamma-interferon on collagen and fibronectin gene expression. J Biol Chem 262:13348–13351. Diehl AM, Goodman Z, Ishak KG (1988) Alcohollike liver disease in nonalcoholics. A clinical and histologic comparison with alcohol-induced liver injury. Gastroenterology 95:1056–1062.
120
Dooley S, ten Dijke P (2012) TGF-beta in progression of liver disease. Cell Tissue Res 347:245–256. Edenberg HJ, Xuei X, Chen HJ, Tian H, Wetherill LF, Dick DM, Almasy L, Bierut L, Bucholz KK, Goate A, Hesselbrock V, Kuperman S,Nurnberger J, Porjesz B, Rice J, Schuckit M, Tischfield J, Begleiter H, Foroud T (2006) Association of alcohol dehydrogenase genes with alcohol dependence: a comprehensive analysis. Hum Mol Genet 15:1539–1549. Gines P, Cardenas A, Arroyo V, Rodes J (2004) Management of cirrhosis and ascites. N Engl J Med 350:1646–1654. Higashiyama R, Inagaki Y, Hong YY, Kushida M, Nakao S, Niioka M, Watanabe T, Okano H, Matsuzaki Y, Shiota G, Okazaki I (2007)Bone marrow-derived cells express matrix metalloproteinases and contribute to regression of liver fibrosis in mice. Hepatology45:213–222. Hong J, Stubbins RE, Smith RR, Harvey AE, Nunez NP (2009) Differential susceptibility to obesity between male, female and ovariectomized female mice. Nutr J 8:11. Itoga S, Harada S, Nomura F (2001) Polymorphism of the 5′-flanking region of the CYP2E1 gene: an association study with alcoholism. Alcohol Clin Exp Res 25:11S– 15S. Kimura M, Higuchi S (2011) Genetics of alcohol dependence. Psychiatry Clin Neurosci 65:213–225. Kono H, Rusyn I, Yin M, Gabele E, Yamashina S, Dikalova A, Kadiiska MB, Connor HD, Mason RP, Segal BH, Bradford BU, Holland SM, Thurman RG (2000) NADPH oxidase-derived free radicals are key oxidants in alcohol-induced liver disease. J Clin Invest106:867–872. Lamkanfi M, Kanneganti TD, Franchi L, Nunez G (2007) Caspase-1 inflammasomes in infection and inflammation. J Leukoc Biol82:220–225. Lee SH, Chae KS, Sohn DH (2004) Identification of expressed sequence tags of genes expressed highly in the activated hepatic stellate cell. Arch Pharm Res 27:422–428. Lin P, Hartz SM, Wang JC, Agrawal A, Zhang TX, McKenna N, Bucholz K, Brooks AI, Tischfield JA, Edenberg HJ, Hesselbrock VM,Kramer JR, Kuperman S, Schuckit MA, Goate AM, Bierut LJ, Rice JP (2012) Copy number variations in 6q14.1 and 5q13.2 are associated with alcohol dependence. Alcohol Clin Exp Res 36:1512– 1518. Millward CA, Burrage LC, Shao H, Sinasac DS, Kawasoe JH, Hill-Baskin AE, Ernest SR, Gornicka A, Hsieh CW, Pisano S, Nadeau JH, Croniger CM (2009) Genetic factors for resistance to diet-induced obesity and associated metabolic traits on mouse chromosome 17. Mamm Genome 20:71–82. Millward CA, Desantis D, Hsieh CW, Heaney JD, Pisano S, Olswang Y, Reshef L, Beidelschies M, Puchowicz M, Croniger CM (2010)Phosphoenolpyruvate
121
carboxykinase (Pck1) helps regulate the triglyceride/fatty acid cycle and development of insulin resistance in mice. J Lipid Res 51:1452–1463. Morimoto M, Hagbjork AL, Wan YJ, Fu PC, Clot P, Albano E, Ingelman-Sundberg M, French SW (1995) Modulation of experimental alcohol-induced liver disease by cytochrome P450 2E1 inhibitors. Hepatology 21:1610–1617. Nakahira M, Ahn HJ, Park WR, Gao P, Tomura M, Park CS, Hamaoka T, Ohta T, Kurimoto M, Fujiwara H (2002) Synergy of IL-12 and IL-18 for IFN-gamma gene expression: IL-12-induced STAT4 contributes to IFN-gamma promoter activation by up-regulating the binding activity of IL-18-induced activator protein 1. J Immunol 168:1146–1153. Netea MG, Kullberg BJ, Verschueren I, Van Der Meer JW (2000) Interleukin-18 induces production of proinflammatory cytokines in mice: no intermediate role for the cytokines of the tumor necrosis factor family and interleukin-1beta. Eur J Immunol30:3057–3060. Ogawa T, Kawada N, Ikeda K (2009) Effect of natural interferon alpha on proliferation and apoptosis of hepatic stellate cells. Hepatol Int 3:497–503. Parola M, Robino G (2001) Oxidative stress-related molecules and liver fibrosis. J Hepatol 35:297–306. Pritchard MT, Malinak RN, Nagy LE (2011) Early growth response (EGR)-1 is required for timely cell-cycle entry and progression in hepatocytes after acute carbon tetrachloride exposure in mice. Am J Physiol Gastrointest Liver Physiol 300:G1124– G1131. Pritchard MT, Nagy LE (2005) Ethanol-induced liver injury: potential roles for egr- 1. Alcohol Clin Exp Res 29(11 Suppl): 146S–150S. Pritchard MT, Nagy LE (2010) Hepatic fibrosis is enhanced and accompanied by robust oval cell activation after chronic carbon tetrachloride administration to Egr-1- deficient mice. Am J Pathol 176:2743–2752. Recknagel RO, Glende EA Jr, Dolak JA, Waller RL (1989) Mechanisms of carbon tetrachloride toxicity. Pharmacol Ther 43:139–154. Salmon DM, Flatt JP (1985) Effect of dietary fat content on the incidence of obesity among ad libitum fed mice. Int J Obes 9:443–449. Sato N, Lindros KO, Baraona E, Ikejima K, Mezey E, Jarvelainen HA, Ramchandani VA (2001) Sex difference in alcohol-related organ injury. Alcohol Clin Exp Res 25:40S–45S. Schroder K, Tschopp J (2010) The inflammasomes. Cell 140:821–832. Singer JB, Hill AE, Burrage LC, Olszens KR, Song J, Justice M, O'Brien WE, Conti DV, Witte JS, Lander ES, Nadeau JH (2004)Genetic dissection of complex traits with chromosome substitution strains of mice. Science 304:445–448.
122
Sudbery I, Stalker J, Simpson JT, Keane T, Rust AG, Hurles ME, Walter K, Lynch D, Teboul L, Brown SD, Li H, Ning Z, Nadeau JH,Croniger CM, Durbin R, Adams DJ (2009) Deep short-read sequencing of chromosome 17 from the mouse strains A/J and CAST/Ei identifies significant germline variation and candidate genes that regulate liver triglyceride levels. Genome Biol 10:R112. Taub R (2004) Liver regeneration: from myth to mechanism. Nat Rev Mol Cell Biol 5:836–847. Ushio-Fukai M (2006) Localizing NADPH oxidase-derived ROS. Sci STKE 2006:re8. Vasiliou V, Pappa A, Petersen DR (2000) Role of aldehyde dehydrogenases in endogenous and xenobiotic metabolism. Chem Biol Interact 129:1–19. Wang HJ, Zakhari S, Jung MK (2010) Alcohol, inflammation, and gut-liver-brain interactions in tissue damage and disease development. World J Gastroenterol 16:1304–1313. Webb A, Lind PA, Kalmijn J, Feiler HS, Smith TL, Schuckit MA, Wilhelmsen K (2011) The investigation into CYP2E1 in relation to the level of response to alcohol through a combination of linkage and association analysis. Alcohol Clin Exp Res 35:10–18. West MA, Billiar TR, Curran RD, Hyland BJ, Simmons RL (1989) Evidence that rat Kupffer cells stimulate and inhibit hepatocyte protein synthesis in vitro by different mechanisms. Gastroenterology 96:1572–1582. Wheeler MD, Kono H, Yin M, Nakagami M, Uesugi T, Arteel GE, Gabele E, Rusyn I, Yamashina S, Froh M, Adachi Y, Iimuro Y,Bradford BU, Smutney OM, Connor HD, Mason RP, Goyert SM, Peters JM, Gonzalez FJ, Samulski RJ, Thurman RG (2001) The role of Kupffer cell oxidant production in early ethanol-induced liver disease. Free Radic Biol Med 31:1544–1549. Wong FW, Chan WY, Lee SS (1998) Resistance to carbon tetrachloride-induced hepatotoxicity in mice which lack CYP2E1 expression. Toxicol Appl Pharmacol 153:109–118. Xuei X, Dick D, Flury-Wetherill L, Tian HJ, Agrawal A, Bierut L, Goate A, Bucholz K, Schuckit M, Nurnberger J Jr, Tischfield J,Kuperman S, Porjesz B, Begleiter H, Foroud T, Edenberg HJ (2006) Association of the kappa-opioid system with alcohol dependence. Mol Psychiatry 11:1016–1024.
123
Figure Legends
Figure 8.1. Congenic strains from chromosome substitution strain-17 (CSS-17). (A)
Construction of congenic strains derived from CSS-17. A/J-derived segments are
represented with shaded squares and the letter “A,” and C57BL/6J (B6)-derived segments are represented with white squares and the letter “B.” Genetic markers (m1 to m11 as described) are listed in order from centromere (left) to telomere (right; Millward et al.,
2009) and are represented to scale showing their approximate location of DNA sequence obtained from the Mouse Genome Informatics website (www.informatics.jax.org). The number of mice tested in each group is in the column labeled “Num.” Chromosomal regions inferred to contain a quantitative trait loci, based on strong phenotypic
differences between strains, are indicated.
Figure 8.2. CYP2E1 induction with alcohol consumption. (A) Western blot analysis of proteins from mice fed ethanol-containing diet (EtOH) or pair-fed diet (PF). Western blots were normalized with heat shock cognate-70 (HSC-70) as a loading control. (B)
Densitometric scan of Western blots. Values represent mean ± SEM for a total of 4 to 6
female mice per group. Values with different superscripts are significantly different p <
0.05 as determined with ANOVA and Bonferroni correction for multiple testing.
Figure 8.3. Measurements of liver injury. (A) Hepatic triglycerides were measured biochemically from mice fed the ethanol-containing diet (+EtOH) or pair-fed diet (pair- fed). Plasma (B) alanine aminotransferases (ALTs) were measured with enzymatic assays
124
from mice at the end of the +EtOH feeding trial. Values represent mean ± SEM for a total
of 4 to 6 female mice per group. Values with different superscripts are significantly
different from each other. p < 0.05 as determined with ANOVA and Bonferroni
correction for multiple testing.
Figure 8.4. Response to chronic CCl4 administration in congenic strains. (A) Western blot analysis of hepatic proteins from mice treated with CCl4 as described in Materials
and Methods. Western blots were normalized with heat shock cognate-70 (HSC-70) as a loading control. Densitometric scan of Western blots is shown. (B) Plasma alanine aminotransferase (ALT) measurement of mice 3 days after the final CCl4 dose. Values
represent mean ± SEM for a total of 6 to 10 male mice per group. Values with different superscripts are significantly different from each other. p < 0.05 as determined with
ANOVA and Bonferroni correction for multiple testing.
Figure 8.5. Reduced liver fibrosis in 17C-6 congenic strain. At 3 days after the final
CCl4 dose, mice were euthanized and livers were stained with Sirius red. Images were
taken at 5× magnification. The images were scanned and quantified as described in
Materials and Methods. Values are means ± SEM, n = 6 to 10 male mice per group.
Superscripts without a common letter differ from each other. p < 0.05 as determined with
ANOVA and Bonferroni correction for multiple testing.
125
Figure 8.6. Immunohistochemistry with α-SMA. At 3 days after final CCl4 injection,
mice were euthanized and portions of individual livers preserved in formalin for
assessment of α-SMA protein expression by IHC. Images were taken at
10× magnification. The images were scanned and quantified as described in Materials
and Methods. Values are means ± SEM for n = 4 to 10 male mice per group. Superscripts
without a common letter differ from each other. p < 0.05 as determined with ANOVA
and Bonferroni correction for multiple testing.
Figure 8.7. Detection of α-SMA. Livers from B6, chromosome substitution strain-17
(CSS-17), 17C-1, and 17C-6 were treated with CCl4 and analyzed for expression of α-
SMA by (A) Western blot normalized with heat shock cognate-70 (HSC-70) for loading
control. Densitometric scans of a representative Western blot are shown for n = 4 to 6
male mice per group. Values with different superscripts are significantly different from
each other. p < 0.05 as determined with ANOVA and Bonferroni correction for multiple
testing.
Figure 8.8. Expression of candidate genes in congenic strains. RNA was isolated from
mice fed Lieber–DeCarli EtOH-containing diet (EtOH-fed) or pair-fed controls (pair-fed)
or treated with either olive oil or CCl4 injections 2×/wk for 3 weeks. Expression of Noxo1 mRNA in (A) EtOH-fed and pair-fed mice and (B) CCl4-treated mice.
Expression of Nlrc4 mRNA in (C) EtOH-fed and pair-fed mice and (D) CCl4-treated
mice. The values are the means ± SEM and normalized with 18S rRNA for n = 4 to 6
126
mice per group. Superscripts without a common letter differ from each other. p < 0.05 as
determined with ANOVA and Bonferroni correction for multiple testing.
Figure 8.9. Expression of Noxo1 in cells isolated from the liver. Hepatocytes,
macrophages, and hepatic stellate cells were isolated from livers of B6 and 17C-1 mice as
described in Materials and Methods. RNA was isolated from (A) hepatocytes, (B)
macrophages with and without lipopolysaccharide stimulation, and (C) hepatic stellate
cells on day 0 or after 7 days of culture. Expression of Noxo1 was measured by RT-PCR.
The values are the means ± SEM and normalized with 18S rRNA for n = 4 to 6 mice per
group. (D) Bone marrow-derived macrophages were isolated from B6 and 17C-1 mice
and cellular reactive oxygen species as described in Materials and Methods. Superscripts
without a common letter differ from each other. p < 0.05 as determined with ANOVA
and Bonferroni correction for multiple testing.
Figure 8.10. Expression of Nlrc4 in cells isolated from the liver. Hepatocytes,
macrophages, and hepatic stellate cells (HSCs) were isolated from livers of B6 and 17C-6
mice as described in Materials and Methods. RNA was isolated from (A) hepatocytes,
(B) macrophages with and without lipopolysaccharide (LPS) stimulation, and (C) HSCs on day 0 or after 7 days of culture. Expression of Nlrc4 was measured by RT-PCR. The values are the means ± SEM and normalized with 18S rRNA for n = 4 to 6 mice per group. Superscripts without a common letter differ from each other. p < 0.05 as determined with ANOVA and Bonferroni correction for multiple testing. (D) Three days
127 after the final CCl4 injection, mice were euthanized and portions of individual livers preserved in formalin for assessment of neutrophil infiltration by immunohistochemistry.
Control (olive oil–injected mice)-treated livers did not exhibit positive NIMP-R14 staining (C57BL/6 used as representative for each strain), but NIMP-R14-positive neutrophils were detected in C57BL/6, A17, and 17-C6 mice after CCl4 exposure
(red/brown staining, black arrows). Images were taken at 200× magnification.
Quantification of NIMP-R14 as described in Materials and Methods. CV, central vein. N = 2 (olive oil) or 3 (CCl4) for each strain. Superscripts without a common letter differ from each other. p < 0.05 as determined with ANOVA and Bonferroni correction for multiple testing.
Figure 8.S1. Expression levels for candidate genes after ethanol feeding study.
Hepatic RNA was isolated from mice fed Lieber DeCarli ethanol-containing diet or pair fed controls. Expression of A) TRIP 10, B) Cdkl4, C) Plg, D) Fndc1, E) Ddx11, F)
Slc5a7 was analyzed with RT-PCR. The values are the means ±SEM and normalized with 18S rRNA for n=4-6 mice per group. Superscripts without a common letter differ from each other, P<0.05 as determined with ANOVA with Bonferroni’s correction for multiple testing.
Figure 8.S2. Noxo1 mRNA expression in bone marrow derived macrophages
(BMDM). BMDM were isolated from B6, CSS-17 and 17C-1RNA and cultured as
128 described in methods. Expression of Noxo1 mRNA was analyzed with RT-PCR. The values are the means ±SEM and normalized with 36b4 for n=4 per group. Superscripts without a common letter differ P<0.05 as determined with ANOVA with Bonferroni’s correction for multiple testing.
129
Table Legends
Table 8.1. Lieber–DeCarli diet study. Values are the mean ± SEM for n = 4 to 6 female mice per group. The means in a row with superscripts without a common letter differ from each other. p < 0.05 as determined with ANOVA and Bonferroni correction for multiple testing. PF, pair-fed diet; +E, EtOH-containing diet.
Table 8.2. Summary of candidate genes. By analyzing deep sequencing data for A/J compared to B6 chromosome 17, we identified candidate genes on chromosome 17 that may regulate susceptibility to liver injury in alcoholic steatohepatitis (ASH). We selected candidate genes in Obrq13 and Obrq15 that contained nonconsensus SNPs in 5′ untranslated region or within the gene. For ASH, mice were determined to have steatosis if fed Lieber–DeCarli EtOH-containing diet and had statistically significant increase of hepatic triglyceride values over pair-fed littermates. Mice were considered to have increased fibrosis if α-smooth muscle actin measurements in CCl4-treated mice were significantly increased over olive oil–treated mice. Detailed results for these experiments can be found in Table 1 for EtOH feeding trial and Figs 4-7 for CCl4 treatment. The data for the high-fat simple carbohydrate are summarized from our previously published study of congenics derived from chromosome substitution strain-17 (CSS-17; Millward et al.,
2009). These mice were characterized as obese if the weight gain was similar to B6 mice.
The mice were considered either insulin sensitive (IS) or insulin resistant (IR) as measured by glucose tolerance tests and homeostasis model of insulin resistance as previously published (Millward et al., 2009). ALT, alanine aminotransferase; AST, aspartate aminotransferase; HFSC, high-fat, simple-carbohydrate.
130
Figure 8.1.
131
Figure 8.2.
132
Figure 8.3.
133
Figure 8.4.
134
Figure 8.5.
135
Figure 8.6.
136
Figure 8.7.
137
Figure 8.8.
138
Figure 8.9.
139
Figure 8.10.
140
Figure 8.S1.
141
Figure 8.S2.
142
Table 8.1.
143
Table 8.2.
144
CHAPTER 9
PUBLICATION
9.1. DeSantis DA, Ko CW, Wang L, Lee PL, Croniger CM. Constitutive activation of the
Nlrc4 inflammasome prevents hepatic fibrosis and promotes hepatic regeneration
after partial hepatectomy. (Submitted to Fibrogenesis & Tissue Repair, April 15th,
2015).
145
9.1.
CONSTITUTIVE ACTIVATION OF THE NLRC4 INFLAMMASOME
PREVENTS HEPATIC FIBROSIS AND PROMOTES HEPATIC
REGENERATION AFTER PARTIAL HEPATECTOMY
David A. DeSantis, Chih-wei Ko, Lan Wang, Peter Lee and Colleen M. Croniger
Department of Nutrition, Case Western Reserve University School of Medicine;
Cleveland, Ohio 44106 USA
Address Correspondence to: Colleen M. Croniger, Department of Nutrition, School of
Medicine, Case Western Reserve University, 10900 Euclid Ave., Cleveland, OH 44106,
USA. Tel: (216)368-4967; Fax: (216)368-6560; E-mail: [email protected]
146
Abbreviations
SNP: single nucleotide polymorphism; Cdx-1: caudal-related homeobox-1; CCl4; carbon tetrachloride; 2/3PH: partial hepatectomy; NAFLD: Non-alcoholic fatty liver disease;
ECM: extracellular matrix proteins; HSC: hepatic stellate cells; KC: Kupffer cells; TNF-
α: tumor necrosis factor α; IL-1β: interleukin-1β; IL-6: interleukin-6; DAMP: Damage- associated molecular patterns; PAMP: pathogen-associated molecular patterns; Naip5:
NLR family, apoptosis inhibitory protein 5; Nlrc4: NLR family, CARD domain containing 4; CARD: caspase activating and recruitment domain; ASC: apoptosis- associated speck-like protein containing a caspase recruitment domain; GWAS: Genome- wide association studies; B6: C57BL/6J; ChIP: Chromatin Immunoprecipitation;
STAT3: signal transducer and activator of transcription 3; IL-18: interleukin-18; IκBα:
NF-Kappa-B Inhibitor-α; c-Fos: FBJ osteosarcoma oncogene; c-Myc: myelocytomatosis oncogene; Crp: C-reactive protein; Ccnd1: Cyclin D1; Nlrc4-MAS: Nlrc4-macrophage activation syndrome; NOMID: neonatal onset multisystem inflammatory disease; PCR: polymerase chain reaction; CMV: cytomegalovirus; M-CSF: macrophage colony- stimulating factor; LPS: lipopolysaccharide; ALT: alanine aminotransferase; BrdU: bromodeoxyuridine; SEM: standard error of the mean; qPCR: Real-Time Quantitative
PCR.
Acknowledgements
The authors would like to acknowledge the CWRU MMMPC for performing partial
hepatectomy studies (DK76174) and the CWRU Genomics Core in the department of
Genetics and Genome Sciences for performing and processing DNA Sequencing data
used in this manuscript. Images were generated at the Imaging Core Facility, Department
147 of Genetics and Genome Sciences, SOM, Case Western Reserve University, and supported by the NIH Office of Research Infrastructure Programs under award number
(S10RR021228). DD was supported by Metabolism Training Grant T32DK007319.
148
Abstract
Background
The molecular mechanisms responsible for the development of hepatic fibrosis are not fully understood. The Nlrc4 inflammasome detects cytosolic presence of bacterial components, activating inflammatory cytokines to facilitate clearance of pathogens and infected cells. Inflammatory cytokines have been shown to stimulate hepatocyte proliferation. We hypothesized that low-grade constitutive activation of the Nlrc4 inflammasome may lead to induced hepatocyte proliferation and prevent the development of hepatic fibrosis.
Results
The promoter of Nlrc4 contains a single nucleotide polymorphism (SNP) (rs74459439-T) when comparing inbred genetic mouse strains C57BL/6J and A/J. Transcription factor caudal-related homeobox-1 (Cdx-1) binds the mouse Nlrc4 promoter and regulates Nlrc4 transcription in the murine macrophage cell line RAW264.7. Cdx-1 stimulates Nlrc4 transcription at higher levels when the Nlrc4 promoter consists of the A/J allele for SNP rs74459439-T. A second SNP (rs29502769) exists within exon 5 of Nlrc4 between
C57BL/6J and A/J mice resulting in an amino acid substitution. Bone marrow derived macrophages from 17C-6 mice, containing the A/J allele of Nlrc4, have increased cellular
IL-1β. The 17C-6 mice have increased IL-1β in plasma after chronic carbon tetrachloride
(CCl4) administration compared to B6 mice. At 36 hours after two-thirds partial hepatectomy (2/3PH) 17C-6 mice restore 87% original liver mass while B6 mice restore
57%. The livers from 17C-6 mice have greater Cyclin D1 protein compared to B6 at
149
sham, 4, 12, and 36 hour time points post 2/3PH. The livers from 17C-6 mice have
increased BrdU incorporation compared to B6 at the 8, 12, and 36 hour time points post
2/3PH. After sham surgery, the livers from 17C-6 mice have increased levels of mRNA
for genes associated with inflammatory cytokine stimulation of hepatocyte proliferation compared to B6 sham operated mice.
Conclusion
These data reveal mild constitutive activation of the Nlrc4 inflammasome as the results of two SNPs, which leads to the stimulation of hepatocyte proliferation. The increased liver regeneration induces rapid liver mass recovery after hepatectomy and may prevent the development of hepatotoxin-induced liver fibrosis. By characterizing the Nlrc4 inflammasome and its contribution to liver regeneration we have potentially provided
new insights for treatment of liver disease.
150
Introduction
Non-alcoholic fatty liver disease (NAFLD) has become a significant cause of chronic liver disease with staggering occurrence in the U.S. and worldwide [1]. NALFD disease prevalence studies estimate anywhere between 2.8% to 38.5% of the U.S. general population has NAFLD [2-6]. Epidemiological studies indicate NAFLD is pathophysiologically linked to metabolic syndrome as it is associated with obesity, hypertension, dyslipidemia, and insulin resistance [7, 8]. Current approximations have greater than 75% of obese individuals afflicted with NAFLD [9].
NALFD is a term used to label a spectrum of liver disease ranging from early stage fatty liver (steatosis) to advanced cirrhosis of the liver and hepatocellular carcinoma [10]. This disease is aptly named as it occurs in individuals who consume little to no alcohol and the
NAFLD pathology closely resembles that of a diseased liver attributable to alcohol abuse
[11]. NAFLD is believed to originate with the accumulation of fatty acids within the liver as a result of insulin resistance [12]. Consequently increased membrane lipid peroxidation and oxidative stress within the cells of the liver result in inflammation and increased deposition of extracellular matrix proteins (ECM), i.e. fibrosis [13]. Excessive fibrosis leads to atypical hepatic arrangement and subsequent scarring. As tissue scarring develops there is decreased hepatic blood flow which initiates hepatocellular dysfunction
[14]. Genetic factors contribute to an individual’s predisposition to the development and progression of NAFLD disease [15].
151
The development of hepatic fibrosis is regulated primarily by hepatic stellate cells (HSC), which synthesize ECM proteins. In a quiescent state, HSC store vitamin A as retinol esters and make up roughly one third of non-parenchymal cells of the liver. When HSC are activated they become fibrogenic. HSCs are activated by injured hepatocytes and stimulated resident macrophages known as Kupffer cells (KC) [16, 17]. KC play a significant role in immune surveillance and production of cytokines such as tumor necrosis factor α (TNF-α), interleukin-1β (IL-1β), and interleukin-6 (IL-6) [18].
Interestingly KC have been shown to be essential for liver regeneration after partial hepatectomy of the liver [19] via cytokine production [20, 21]. KC and their immune response to various toxic insults play a role in disease progression and hepatic homeostasis [22].
The inflammasome is a multimeric protein complex that is assembled and activated upon the detection of cellular infection or cell stress [23]. Damage-associated molecular patterns (DAMP) and pathogen-associated molecular patterns (PAMP) bind to pattern- recognition receptors inducing an intracellular signaling cascade. These events lead to oligomerization of inflammasome components, activation of intracellular cysteine protease caspase-1, and the subsequent cleavage and maturation of pro-inflammatory cytokines IL-1β and IL-18 [24-27].
The Nlrc4 inflammasome detects the cytosolic presence of bacterial flagellin during infection using NLR family, apoptosis inhibitory protein 5 (Naip5) and NLR family,
CARD domain containing 4 (Nlrc4) to form a hetero-oligomeric inflammasome structure
152
[28, 29]. Several activated Nlrc4s then form an inflammasome complex by which Nlrc4 uses an N-terminal caspase activating and recruitment domain (CARD) to interact with the CARD of pro-caspase-1, leading to cleavage and activation of caspase-1. The Nlrc4 inflammasome may associate with caspase-1 independent of adaptor protein apoptosis- associated speck-like protein containing a caspase recruitment domain (ASC) which itself contains a CARD [30].
Identification of genetic factors contributing to the pathogenesis of NAFLD facilitates the potential to target susceptible individuals for interventional strategies to ameliorate
NAFLD progression. Genome-wide association studies (GWAS) are an unbiased tool for the identification of gene variants associated with genetic traits. Unfortunately, GWAS examining phenotypes relevant to NAFLD is lacking. A 2010 GWAS by Chalasani et al. linked genes involved in lipid metabolism and collagen deposition with NAFLD characterized by histology [31]. A 2015 GWAS found the Nlrc4 inflammasome was involved in IL-18 production in patients with acute coronary syndromes [32]. Association studies such as these should be used in conjunction with hypothesis-driven investigative studies to identify key genetic regulators of NAFLD. We have previously shown allelic differences of the Nlrc4 gene between inbred genetic mouse strains C57BL/6J (B6) and
A/J modulate the development and/or resolution of hepatic fibrosis, a critical stage of
NALFD development [33]. Identifying the molecular mechanisms by which Nlrc4 modulates liver fibrosis is essential to fully understand the complex dynamics of liver disease origin and development. Using models of hepatotoxin-induced liver injury and
153 liver regeneration after partial hepatectomy, we established the role of Nlrc4 in governing hepatic fibrosis development.
154
Experimental Procedures
Animal Husbandry
All procedures involving animals were approved by the Case Western Reserve University
Institutional Animal Care and Use Committee. B6 and A/J mice were obtained from
Jackson Laboratories and maintained at Case Western Reserve University for over 10 generations [59]. Congenic strain 17C-6 was derived from CSS-17 as previously described [60]. Mice were housed in a microisolator environment on a 12hr:12hr light/dark cycle. All mice were weaned at 3-4 weeks of age and maintained on LabDiet
#5010 autoclavable rodent chow (LabDiet, Richmond, IN) with food and water provided ad libitum.
Carbon tetrachloride (CCl4) Administration and Tissue Collection
For both acute and chronic CCl4 studies, CCl4 (Sigma-Aldrich, St. Louis, MO) was performed as previously described [61]. Briefly, CCl4 was prepared fresh by pre-diluting
1 part CCl4 to 3 parts olive oil. For acute CCl4 administration, male mice, ages 10-12 weeks, were given a single injection (1µL/g body weight) of diluted CCl4 to the intraperitoneal cavity. Control mice received a single injection of olive oil (1µL/g body weight). Mice were sacrificed at 24 hours and 48 hours after injection. All control mice receiving olive oil were sacrificed 48 hours after injection. For chronic CCl4 administration, mice were given two injections weekly (Tuesday/Friday) for 5 weeks.
Mice were gradually increased in dose of CCl4 over the first three injections (first, 0.25
μL/g body weight; second, 0.5 μL/g body weight; third and subsequent doses, 1µL/g body weight). Both CCl4 and olive oil treated mice were sacrificed 72 hours after the
155
final injection. Mice were allowed access to food and water ad libitum for the duration of the acute and chronic studies. Blood plasma and liver tissue were collected and frozen until further processing.
DNA Constructs, Cell Transfections, and Reporter Assays
DNA was isolated from B6 and A/J mice. Using custom designed primers that span the
1000bp region preceding the start site of transcription for Nlrc4, we isolated the Nlrc4 promoter by polymerase chain reaction (PCR). Each 1kb promoter was cloned into a pSC-A-amp/kan cloning vector using StrataClone PCR cloning kit (Aligent
Technologies, Santa Clara, CA) and then subcloned into the luciferase plasmid vector
pGL4.10-luc2 using QuickLink DNA Ligation Kit (Sigma Aldrich, St. Louis, MO) and
sequenced by the Case Western Reserve University Genomics Core. The expression
constructs were co-transfected with either an overexpression vector for Cdx-1 driven by
cytomegalovirus (CMV) promoter or a control vector containing only the CMV promoter
(Open Biosystems Products, Huntsville AK) into murine macrophage cell line RAW
264.7 using FuGENE HD Transfection Reagent (Promega, Madison, WI). Cells were
cultured using ATCC-formulated Dulbecco's Modified Eagle's Medium (Life
Technologies, Grand Island, NY) supplemented with 10% fetal bovine serum (Life
Technologies, Grand Island, NY). Dual-Glo Luciferase assay system (Promega, Madison,
WI) activating firefly (Photinus pyralis) and Renilla (Renilla reniformis) luciferases were
used to control for transfection efficiency. Luminescent signal was measured by plate-
reading illuminometer (Molecular Devices, Sunnyvale, CA) and activity was calculated
according manufacturer’s protocol.
156
Chromatin Immunoprecipitation
ChIP was conducted on a cultured murine macrophage cell line RAW 264.7 as previously described [62]. The cells were cross-linked with 11% buffered formaldehyde solution and sonicated into DNA fragments. DNA fragments were selected using a Cdx-1 antibody
(Abcam, Cambridge, MA) conjugated with magnetic beads and purified. The fragments
were analyzed by PCR (Life Technologies, Grand Island, NY). Custom designed primers
(IDT, Coralville, IA) were used to detect potential transcription factor Cdx-1 binding sites (Additional File 2). Anti-GFP antibody (Abcam, Cambridge, MA) served as an antibody negative control. Primers for a negative control locus and a non-template
control were used during RT-PCR. The amplified products were visualized on 2%
agarose gels.
Bone Marrow Derived Macrophages (BMDM)
12-16 week old male B6 and congenic 17C-6 mice were anesthetized and their hind legs
harvested from the pelvis keeping the femur bone intact. The surrounding skin and muscle tissues were removed revealing femur, tibia, and fibula bones. In an aseptic
environment the femur and tibia bones were cut at each end and the lumen flushed with
culture media. The marrow cells obtained were strained, purified, and plated. Cells were
cultured in macrophage differentiation medium containing 10 ng/mL macrophage
colony-stimulating factor (M-CSF) at 37°C in an air atmosphere with 5% CO2 for 7-10
days to differentiate macrophage progenitor cells into mature macrophages as described
[63, 64].
157
LPS Exposure to BMDM
Once BMDM were fully differentiated they were exposed to bacterial LPS (100ng/mL)
(Sigma-Aldrich, St. Louis, MO) for 2, 4, and 8 hours at 37°C in an air atmosphere with
5% CO2. Culture media was collected and cells were then washed in 1X PBS and collected. Subcellular fractionation of BMDM cells was prepared as previously described and either RNA or protein was isolated [65].
2/3 Partial Hepatectomy
12-16 week old male B6 and congenic 17C-6 mice underwent two-thirds partial hepatectomy surgery performed by Case Western Reserve University Mouse Metabolic
Phenotyping Center (U24-DK76174) as previously described [66]. The weight of resected liver for each mouse was recorded at the time of surgery. We found that resection of the median and left lateral lobes resulted in an average removal of 55% of total liver tissue. The animals recovered for 2, 4, 8, 12, 36, and 168 hour (7 days) time periods and then were euthanized. Blood plasma was taken and the remaining liver tissue was removed, weighed, and frozen. The animal survival rate was >95% at all time points with no statistical difference between genotypes. Percent liver regeneration was calculated by dividing the weight of the liver at the time of sacrifice by the initial liver weight of animal and multiplying by 100. The initial liver weight was obtained by assuming the resected liver weight was 55% of original liver mass [67, 68].
% / X 100
158
A = (Liver weight at sacrifice)
B = (Estimated liver weight before PH)