Potential conflict of interest: Nothing to report.
Alpha-1 antitrypsin (α1-AT) deficiency is the most common genetic cause of liver disease in children. The homozygous α1-ATZ mutation (PiZZ) results in significant liver disease in 10% of all affected patients. The α1-ATZ mutation also may lead to worse liver injury in the setting of other liver diseases such as cystic fibrosis, nonalcoholic fatty liver disease, and hepatitis C. Although cholestatic injury is common to many forms of liver disease, its effect on the PiZZ phenotype is unknown. To elucidate the interplay of cholestasis and the PiZZ phenotype, we performed bile duct ligation (BDL) on C57BL/6 mice possessing a transgenic α1-ATZ mutation and littermate controls. PiZ transgenic mice undergoing BDL developed more liver fibrosis by quantification of Sirius red staining (P = 0.0003) and hydroxyproline (P = 0.007) than wild-type mice after BDL. More activated hepatic stellate cells (HSCs) and apoptotic cells also were observed in the PiZ BDL model. Quantitative real time polymerase chain reaction (PCR) of the endoplasmic reticulum (ER) stress markers CHOP and GRP78 were 4-fold and 2-fold more up-regulated, respectively, in PiZ BDL mice when compared with wild-type BDL mice (P = 0.02, P = 0.02). Increased apoptosis was also noted in PiZ BDL mice by terminal deoxynucleotidyl transferase-mediated nick-end labeling (TUNEL) and cleaved caspase-3 histological staining. Conclusion: PiZ transgenic mice are more susceptible to liver fibrosis induced by cholestasis from BDL. Cholestasis therefore may lead to increased fibrosis in α1-AT deficiency, and the α1-ATZ mutation may act as a modifier gene in patients with concurrent cholestatic liver diseases such as cystic fibrosis. (HEPATOLOGY 2007.)
Alpha-1 antitrypsin (α1-AT) deficiency is the most frequent cause of genetic liver disease in children, leading to liver disease in 10% of patients affected by the most common form of the disease (α1-ATZ).1, 2
Homozygosity for the α1-ATZ mutation (also referred to as PiZZ) is generally required for the development of liver disease, although several studies report an increased risk to heterozygous individuals.3–7 A single α1-ATZ mutation may also predispose patients with cystic fibrosis to significant liver disease.8 Therefore, the α1-ATZ mutation not only may cause liver injury but it may act as a modifier gene that exacerbates other forms of liver disease. The phenotypical variability of liver disease in PiZZ individuals ranges from mild elevation in liver function tests to end-stage cirrhosis. In infants, α1-AT deficiency can present with neonatal hepatitis and jaundice, whereas later in life patients can develop cirrhosis and hepatocellular carcinoma.9 The variability and incidence of liver disease in α1-AT deficiency may relate to environmental and genetic factors. Because cholestasis is a common aspect of many insults to the liver, including cystic fibrosis, understanding the role of cholestasis in the progression of liver disease in the α1-ATZ phenotype is essential.
α1-AT belongs to the serine protease inhibitor (serpin) family and is produced primarily in the liver. The α1-ATZ variant is caused by a single amino acid substitution at position 342, producing a mutant protein that misfolds and accumulates within the endoplasmic reticulum (ER). Build-up of α1-ATZ protein within the ER of the hepatocyte appears to be the primary cause of liver disease by a gain of toxic mechanism.1 Globules of α1-ATZ proteins are classically observed at low-power magnification, staining positively with periodic acid-Schiff while remaining diastase resistant (DPAS). Although progress has been made in understanding α1-AT deficiency, the mechanism of liver disease has yet to be determined. The role of cholestasis as a modifier of liver disease in α1-AT deficiency has not been previously studied.
This study demonstrated that α1-ATZ transgenic mice undergoing bile duct ligation (BDL) have increased apoptosis and cytokine release, which leads to the activation of hepatic myofibroblasts and the generation of increased hepatic fibrosis.
α-SMA, alpha smooth muscle actin; α1-AT, alpha-1 antitrypsin; α1-ATZ, homozygous alpha 1 antitrypsin mutation; ALT, alanine aminotransferase; BDL, bile duct ligation; CHOP, CCAAT/enhancer-binding protein homologous protein or GADD153; DPAS, Diastase-resistant periodic acid-Schiff stain; ER, endoplasmic reticulum; GRP78, glucose regulated protein 78; HSC, hepatic stellate cell; mRNA, messenger ribonucleic acid; PCR, polymerase chain reaction; PiZ, protease inhibitor genotype; PiZZ, protease inhibitor genotype homozygous for the Z allele; RNA, ribonucleic acid; RT, reverse transcription; TGF-β1, transforming growth factor beta1; TIMP-1, tissue inhibitor of metallo proteinase 1; TUNEL, terminal deoxy-nucleotidyl transferase-mediated nick end-labeling.
Materials and Methods
Alpha-1 antitrypsin Z protein(PiZ) Mice were maintained on a C57Bl/6J background as described.10 Wild-type littermates served as controls. All mice were maintained in a specific pathogen-free facility on a 12-hour dark–light cycle and were fed ad libitum standard mouse chow and water. The experiment was approved by the Columbia University Institutional Animal Use and Care Committee and followed the guidelines outlined in “Guide for the Care and Use of Laboratory Animals” (NIH publication 86–23).
Mice underwent BDL at 12 weeks of age for a 3-week period before sacrifice. Mice were anesthetized with an intraperitoneal injection of ketamine (80 mg/kg) and xylazine (10 mg/kg). After midline laparotomy, the common bile duct was ligated 3 times with 6–0 silk and transected between the 2 most distal ligations. Sham operation was performed similarly except that the bile duct was not ligated or transected. A total of 41 mice were used for this experiment. Four wild-type and 4 PiZ mice were used as sham controls. Seventeen PiZ mice underwent BDL along with 16 wild-type mice. Two PiZ mice and 6 wild-type mice died before their scheduled sacrifice. At the time of sacrifice, liver specimens were flash frozen in liquid nitrogen or fixed in 10% formalin solution before embedment in paraffin wax for histological analysis. Serum was also collected for measurement of hepatic enzymes
Measurement of Hepatic Enzymes.
Measurement of serum alanine aminotransferase (ALT) and alkaline phosphatase was performed by Antech Diagnostic (Lake Success, NY).
Liver specimens were embedded in paraffin and cut into 5-μm sections. To histologically quantify α1-ATZ, cytokeratin, and Sirius red staining, Image J Software (NIH) was used to calculate the percent area staining positively in 5 random low-power views. Hematoxylin-eosin as well as DPAS staining were performed on all specimens. Immunohistochemical staining with polyclonal rabbit anti-bovine cytokeratin antibody (Dako, Carpenteria, CA) was done to measure biliary duct proliferation.11 Routine Sirius red (saturated picric acid containing 0.1% Direct Red 80 and 0.1% Fast Green FCF) staining was performed to assess the degree of fibrosis. Terminal deoxy-nucleotidyl transferase-mediated nick end-labeling (TUNEL) staining (Chemicon International, Temecula, CA) and immunohistochemical staining with polyclonal rabbit antibody against cleaved caspase-3 (Cell Signaling Technology, Danvers, MA) were performed on specimens to assess apoptosis. Apoptosis was quantified by counting positively staining cells in 5 random fields at 200× magnification. Apoptosis was measured for each specimen as a percentage of total cells per field. Staining for activated hepatic myofibroblasts was done using a monoclonal mouse anti-human α-smooth muscle actin (α-SMA) antibody (clone 1-A4 from DakoCytomation).
Measurement of Hydroxyproline Content.
Hydroxyproline content was measured by a spectrophotometric assay as an assessment of liver collagen content. Liver tissues were homogenized in ice-cold distilled water (1 mL) using a polytron homogenizer. Subsequently, 125 μl 50% trichloroacetic acid was added and homogenates were further incubated on ice for 30 minutes. Precipitated pellets were hydrolyzed for 24 hours at 110°C in 6 N HCL. After hydrolysis, samples were filtered and neutralized with 10 N NaOH. Hydrolysates were then oxidized with Chloramine-T (Sigma, St. Louis, MO) for 25 minutes at room temperature. The reaction mixture was incubated in Ehrich's perchloric acid solution at 65°C for 20 minutes, and cooled down at room temperature. Sample absorbance was measured at 560 nm. Purified hydroxyproline (Sigma) was used to set a standard. Hydroxyproline content was expressed as micrograms of hydroxyproline per gram of liver.
Quantitative Real Time Polymerase Chain Reaction.
Total ribonucleic acid (RNA) was extracted from liver tissue using TRIzol reagent (Invitrogen, Carlsbad, CA) and was treated with ribonuclease-free deoxyribonuclease (Promega, Madison, WI) for 30 minutes at 37°C. After deoxyribonuclease treatment, the RNA was cleaned using an RNeasy kit (Qiagen, Valencia, CA). RNA was reverse transcribed using a first-strand complementary deoxyribonucleic acid kit with random hexamers (Amersham Pharmacia Biotechnology, Buckinghamshire, UK) according to the manufacturer's protocol. Quantitative real time polymerase chain reaction (PCR) was performed using probe-primers sets for the following: tissue inhibitor of metalloproteinase-1 (TIMP-1), CCAAT/enhancer-binding protein homologous protein (CHOP), glucose regulated protein 78 (GRP78), transforming growth factor (TGF)-β1, α-SMA, collagen-α1(I), and 18s ribosomal RNA (Applied Biosystems, Foster City, CA). All assays were performed using TaqMan PCR and normalized with 18S RNA as an internal control. The PCR reaction and analysis were carried out using the ABI-Prism 7000 Sequence Detector and software (Applied Biosystems). The relative abundance of the target genes was obtained by calculating against a standard curve and normalized to an internal control. All PCR primers and probes were purchased from ABI TaqMan Gene Expression Assays (Applied Biosystems).
Liver tissue was homogenized in radioimmunoprecipitation buffer (10 mM Tris HCl, 150 mM NaCl, 1% Nonidet P-40, 0.1% sodium dodecyl sulfate) containing protease inhibitor cocktail (Roche) and phosphatase inhibitors. Forty micrograms of liver protein were loaded for each specimen into a sodium dodecyl sulfate polyacrylamide gel. Blotting was performed using polyclonal rabbit anti-human α-1 antitrypsin (1:2000, DAKO), monoclonal α-smooth muscle actin antibody (1:3000 dilution, Sigma, St. Louis, MO), and a polyclonal caspase-12 antibody (1:750 dilution, Cell Signaling Technology). Cleaved caspase-12 was measured at 42 kDa as per the manufacturer's directions. Equal loading of protein was determined by Western blot using a mouse β-actin antibody (1:5,000 dilution, Sigma, clone AC-15). Band densitometry was performed using Scion Image Software (Scion Corp., Frederick, MD).
Primary Hepatocyte Cultures and Treatments.
PiZ mice and littermate controls were anesthetized with ketamine and xylazine administered by intraperitoneal injection. Hepatocytes were then isolated by a nonrecirculating in situ collagenase perfusion of the liver through cannulation of the inferior vena cava as previously described.12, 13 After isolation, hepatocytes were washed 3 times and centrifuged at 50g for 1 minute. Cell viability was consistently 90% as determined by trypan blue exclusion. Cells (1 × 105 and 1 × 106) were plated on 6-well dishes coated with rat collagen type I in Waymouth's medium containing 10% fetal bovine serum and antibiotics (plating medium). After 4 hours, the culture was washed with phosphate-buffered saline and changed to serum-free Roswell Park Memorial Institute 1640 (RPMI 1640) medium. The RPMI 1640 was removed after 12 hours and replaced with serum-free RPMI 1640 containing either 0 μM H2O2 (control) or 250 μM H2O2. After 4 hours, cell death was assessed by double staining cell cultures with propidium iodide and Hoechst 33258. Propidium iodide is a dye used as a marker of cell death that penetrates compromised plasma membranes and stains nucleic acids. Hoeschst 33258 stains all cells. Briefly, cells were stained with both dyes (100 μM) for 10 minutes and examined under a fluorescent microscope (Olympus IX72, Tokyo, Japan). Quantification of necrotic cells was performed by counting at least 1,500 cells per well and expressing necrotic cells as a percentage of total cells counted. Three PiZ mice were used for cell isolations, and necrotic cells were quantified in 6 wells for each mouse (3 wells with 250 μM H202and 3 control wells). Cell death was quantified in 2 wild-type littermates in the same manner.
Hepatic Stellate Cell Isolation, Purification, and Culture.
Mouse hepatic stellate cells (HSCs) were isolated from PiZ and wild-type littermate controls by collagenase-pronase perfusion and subsequent density centrifugation on Nycodenz gradients as described previously.14, 15 For each isolation, HSCs from 3 mice were placed onto a single Nycodenz gradient. HSCs were cultured on plastic dishes for 5 days (“culture-activated HSC”) in Dulbecco's minimum essential medium containing 10% fetal bovine serum, glutamine, 4-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid buffer, and antibiotics.
Data between groups were analyzed using a 2- tailed t test and are reported as mean ± standard deviation. A P value of less than 0.05 was considered statistically significant.
PiZ Mice Accumulate α1-ATZ Protein Within Hepatocytes.
The PiZ transgenic mouse used for this experiment has been genetically engineered to express a human α1-ATZ gene that includes the human promoter region and has been frequently used to study α1-ATZ liver disease.10 In this mouse model, α1-ATZ protein accumulates within the ER of hepatocytes in a manner nearly identical to that seen in humans homozygous for the α1-ATZ mutation.9 DPAS staining was performed on all specimens to ascertain the baseline accumulation of α1-ATZ protein within PiZ mice and controls. Wild-type BDL and sham operated mice demonstrated no staining with DPAS, whereas PiZ mice demonstrated large areas of DPAS staining with and without BDL (Fig. 1A). However, the percent area that stained positively for DPAS was not significantly different between PiZ BDL and PiZ Sham mice (Fig. 1B). Both quantitative reverse transcription (RT)-PCR as well as Western blot analysis did not identify any significant difference in α1-ATZ mRNA or protein expression in PiZ mice with or without BDL (Fig. 1C,D). Therefore, the hepatic content of α1-ATZ protein was not significantly affected in PiZ mice after BDL.
Markers of ER Stress Are Up-Regulated in PiZ Mice.
The accumulation of α1-ATZ within the ER of hepatocytes may predispose PiZ mice to ER stress. ER stress is a response to abnormal ER function that can result in apoptosis. To investigate whether ER stress is increased in PiZ mice, we measured mRNA of CHOP, a transcription factor activated by ER stress that induces apoptosis.16, 17 Quantitative real-time PCR demonstrated a 4-fold up-regulation in CHOP messenger ribonucleic acid (mRNA) in PiZ BDL mice compared with wild-type BDL mice (P = 0.02; Fig. 2A). No statistical difference was noted when comparing levels of CHOP between PiZ shams and WT shams (P = 0.43), or PiZ shams and PiZ BDL mice (P = 0.3). We also measured another marker of ER stress, GRP78. The mRNA level of GRP78 in the liver was increased in PiZ mice before and after BDL (P 0.02) 2-fold compared with wild-type shams (Fig. 2B). In addition, cleaved caspase-12, which is involved in ER stress–induced apoptosis,18 was significantly elevated in PiZ BDL mice when compared with PiZ sham mice (Fig. 2C). In contrast, cleaved caspase-12 was not increased in wild-type BDL mice when compared with wild-type sham mice. These results indicate that a higher level of ER stress in PiZ mice may lead to apoptosis after BDL.
Increased Apoptosis After BDL in PiZ Mice.
Higher ER stress in PiZ mice suggested an increased susceptibility to ER stress–related hepatocyte apoptosis from BDL. To evaluate apoptosis in PiZ mice, both TUNEL staining and cleaved caspase-3 immunostaining were performed. Increased TUNEL staining of apoptotic hepatocytes was noted in PiZ BDL mice when compared with wild-type BDL mice, PiZ shams, and wild-type sham mice (Fig. 3A). To corroborate these results, immunohistochemical staining was also performed using an antibody against cleaved caspase-3, which is the final effector caspase in the apoptosis pathway.19 The number of hepatocytes that positively stained for cleaved caspase-3 was also increased in PiZ BDL mice when compared with wild-type BDL, PiZ shams, and wild-type sham controls (Fig. 3C). Quantification of apoptotic cells demonstrated more apoptosis in the PiZ BDL versus wild-type BDL by both TUNEL (P = 0.042, 2-fold increase) and cleaved caspase-3 staining (P = 0.014, 4-fold increase) (Fig. 3B, D). These results indicate that hepatic apoptosis is increased in the PiZ mouse model after BDL.
Increased Hepatocyte Cell Death After Exposure to H2O2in PiZ Hepatocytes.
To further characterize the mechanism behind increased apoptosis after liver injury in the PiZ BDL model, hepatocytes were isolated from non-BDL PiZ and wild-type mice to test their susceptibility to cell death after exposure to H2O2. Previous studies have shown that reactive oxygen species are increased after BDL, which then play a critical role the development of liver injury and fibrosis.14, 20, 21 To test the effect of ROS, primary hepatocytes from PiZ and wild-type mice were isolated, serum starved overnight, and then exposed to cell media containing either 250 μM or 0 μM H2O2 for 4 hours. Cell death was measured as a percentage of total cells counted per experimental well. After exposure to 250 μM of H2O2, 22% of PiZ hepatocytes were necrotic compared with only 4.5% of wild-type hepatocytes (P < 0.001 Fig. 4). Interestingly, even without exposure to H2O2,PiZ hepatocytes demonstrated more cell death than wild-type hepatocytes (P < 0.05). Therefore, primary isolated PiZ hepatocytes are more susceptible to cell death compared with hepatocytes isolated from wild-type mice.
Activated HSCs Are Increased in PiZ BDL Mice.
Hepatocyte apoptosis has been shown to correlate with hepatic fibrosis.22 Engulfment of apoptotic bodies by phagocytic cells leads to the production of transforming growth factor beta1 (TGF-β1), which in turn stimulates activation of HSCs.23, 24 Once activated, HSCs transdifferentiate into myofibroblasts, which act as a key cell population in hepatic fibrogenesis.25, 26 α-SMA expression is seen in cell types such as portal fibroblasts and HSCs when they have been activated and begin to produce collagens.27 To evaluate the status of myofibroblast activation in the liver, we performed both immunohistochemical staining and Western blot analysis for α-SMA. α-SMA expression was markedly increased after BDL in PiZ mice compared with wild-type mice by both of these methods (Fig. 5A, B). Next, we measured the mRNA level of TGF-β1, which is a potent profibrogenic mediator that activates HSCs.28, 29 TGF-β1 was increased 2-fold in PiZ BDL mice when compared with wild-type BDL mice (P = 0.03; Fig. 5C).
To assess whether the increased expression of α-SMA and TGF-β1 resulted from an intrinsic difference between the HSCs of PiZ and wild-type mice, we isolated HSCs from both PiZ and wild-type littermate controls. When the HSCs were activated after 5 days of culture, they were harvested for analysis. Although expression of α1-ATZ is thought to primarily occur in hepatocytes, α1-ATZ mRNA has never been measured in HSCs. HSCs from PiZ mice demonstrated barely detectable levels of α1-ATZ mRNA by quantitative RT-PCR analysis when compared with hepatocytes isolated from PiZ mice (Table 1A). In addition, hepatocytes and HSCs isolated from wild-type mice demonstrated no mRNA expression of α1-ATZ. Further quantitative RT-PCR analysis was performed to examine the relative activation of HSCs from PiZ and wild-type mice by measuring TGF-β, α-SMA, and collagen-α1 (I). These markers were similarly up-regulated by quantitative RT-PCR in culture activated HSCs in both PiZ and wild-type mice (Table 1B, C, D). Considering that PiZ HSCs cannot produce α1-ATZ mRNA and are similarly activated in vitro compared with wild-type HSCs, variations in HSC activation noted after BDL cannot be attributed to any inherent difference between HSCs in PiZ and wild-type mice.
Table 1. Activated HSCs isolated from PiZ and wild-type (WT) mice demonstrate similar activation patterns. HSCs were isolated from 3 PiZ and 3 WT mice and were allowed to activate in culture for 5 days.*
(A) The fold increase in α1-ATZ mRNA, as measured by quantitative RT-PCR, was dramatically expressed in PiZ hepatocytes (PiZ Hep) compared to PiZ HSCs. No expression of α1-ATZ mRNA was noted in hepatocytes or HSCs isolated from WT mice. (B,C,D) The fold increase in TGF-β1, α-SMA and collagen-α1 (I) mRNA was measured by quantitative RT-PCR in activated PiZ HSCs compared to WT HSCs. No significant difference in up-regulation was noted between PiZ HSC and WT HSCs by measuring any of these three markers.
A. α-1 ATZ mRNA
B. TGF-β1 mRNA
C. α-SMA mRNA
D. Collagen mRNA
Increased Fibrosis in PiZ Mice After BDL.
More apoptotic hepatocytes and more activated HSCs in the PiZ BDL model may lead to the development of increased hepatic fibrosis. To confirm this hypothesis, collagen deposition was measured by quantifying both Sirius red staining and liver hydroxyproline content. When liver specimens were stained for collagen using Sirius red, PiZ BDL mice were found to have a 2-fold increase in fibrosis compared wth wild-type BDL mice (P = 0.0003; Fig. 6A, B). No significant fibrosis was noted by Sirius red staining on sham operated PiZ and wild-type mice. Hydroxyproline was elevated by 1.5-fold in PiZ BDL mice when compared with wild-type BDL mice (P = 0.007; Fig. 6C). In addition, the mRNA level of TIMP-1, a marker of fibrosis,26 was 4-fold higher in PiZ BDL versus wild-type BDL mice (P = 0.002) as measured by quantitative real-time PCR (Fig. 6D).
Bile Duct Proliferation After BDL Is Similar in PiZ Versus Wild-type Mice.
After BDL, there is a proliferation of bile ducts within the liver in response to cholestasis. Liver fibrosis is more prominent in these areas of ductular proliferation because of the deposition of collagen by peribiliary myofibroblasts derived from activated HSCs and portal fibroblasts.27, 30, 31 To demonstrate whether biliary ductal proliferation in response to BDL is altered in PiZ mice, we performed immunohistochemical staining to quantify ductal proliferation. No significant differences were seen in bile duct proliferation between PiZ BDL mice and wild-type BDL mice, suggesting that PiZ expression does not have a significant effect on bile duct proliferation (Fig. 7A, B). Therefore, the increased fibrosis noted in PiZ BDL mice cannot be attributed to increased biliary ductal proliferation.
Hepatic Enzymes Are Similar in PiZ BDL and WT BDL Mice.
Serum ALT and alkaline phosphatase were measured to assess the degree of liver injury. No statistically significant differences were seen between PiZ BDL and WT BDL mice (Table 2).
Table 2. Serum hepatic enzymes after BDL. Serum alanine aminotransferase (ALT) levels were not significantly elevated in PiZ BDL versus wild-type (WT) BDL mice. Alkaline phosphatase levels were not significantly elevated in PiZ BDL versus WT BDL mice. Data are expressed as means ± SD
148 ± 71
113 ± 39
32 ± 16
15 ± 7
Mean Alkaline Phosphatase
769 ± 502
497 ± 313
26 ± 21
64 ± 32
α1-AT deficiency is one of the most common inheritable forms of liver disease and is the most frequent genetic indication for liver transplantation in children. Although homozygous individuals are more likely to develop liver disease, even a single α1-ATZ mutation may predispose patients to hepatic injury3, 4, 7, 32 and act as a modifier gene in the progression of liver disease in patients suffering from cystic fibrosis,8 non-alcoholic fatty liver disease, and hepatitis C.5 However, most patients homozygous for the Z mutation never develop clinically relevant liver disease. The development of liver injury in this model therefore must require an environmental or genetic factor in addition to the Z mutation. Genetic modifiers, including those that modulate degradation of misfolded proteins or those that could be directly injurious to the liver, could represent the key to understanding the development of liver disease in α1-AT deficiency. Because cholestasis is common to many forms of liver disease, studying the role of cholestasis on the phenotype of the PiZ genotype is critical. In our study, we hypothesized that a cholestatic insult would provide a “second hit” that in combination with the intrinsic Z mutation would lead to worse liver injury as measured by hepatic fibrosis.
A possible mechanism for increased fibrosis in the PiZ BDL model is increased ER stress. ER stress refers to a series of cellular processes that occur in response to abnormal ER function that eventually can lead to cell death.17 Accumulation of α1-ATZ protein within the ER of hepatocytes would presumably up-regulate this system. A previous investigator demonstrated increased caspase-12 activation in PiZ mice at baseline, whereas other markers such as GRP78 were not up-regulated.33 Our study shows increased ER stress in the PiZ mouse model after induced cholestatic liver injury. We chose CHOP as a marker for ER stress both because its transcription is up-regulated by many molecules in the ER stress pathway (ATF6, ATF4, XBP-1) and it has been positively correlated with apoptosis.17, 34 CHOP mRNA was significantly up-regulated in PiZ BDL mice when compared with wild-type BDL controls. However, CHOP was not significantly up-regulated in PiZ Sham operated mice, which have no noticeable liver disease. Another marker of ER stress, GRP78, was significantly increased when comparing PiZ BDL or PiZ Sham mice with wild-type BDL mice. The comparable up-regulation of GRP78 in both PiZ BDL and PiZ shams suggests a baseline elevation in ER stress. Nevertheless, caspase-12, which is released in response to ER stress and is a proposed inducer of apoptosis, was significantly up-regulated only in the PiZ BDL mice. This suggests that ER stress may play a significant role in the hepatocyte apoptosis noted in PiZ mice after the cholestatic liver injury of BDL.
In a state of increased ER stress, it would follow that the PiZ BDL mice would also have more apoptosis when compared with controls. Indeed, quantitative histological analysis with TUNEL and cleaved caspase-3 demonstrated that there was more apoptosis in PiZ BDL mice than in wild-type controls or sham operated mice. This is a key observation, because hepatocyte apoptosis has been previously correlated with cytokine and chemokine production and the activation of HSCs.22, 35 Primary hepatocyte isolation confirmed that hepatocytes from PiZ mice are more susceptible to cell death. Interestingly, ALT levels and alkaline phosphatase levels were not significantly different in the PiZ and wild-type BDL mice. The relatively modest elevation in ALT levels in both PiZ BDL and wild-type BDL (<150 U/L) are consistent with the chronic nature of the injury. However, TGF-β1, a cytokine that activates HSCs, was significantly up-regulated along with markers for activated HSCs in the PiZ BDL mice. In vitro analysis of isolated HSCs after activation in culture indicated no significant differences in markers of activation between PiZ and wild-type HSCs. Increased activation of PiZ HSCs after BDL therefore must be the result of the increased levels of profibrogenic molecules such as TGF-β1 after hepatocyte apoptosis. Finally, liver fibrosis, as quantified by collagen staining and hydroxyproline measurements, was increased in PiZ BDL mice. Increased fibrosis in the PiZ BDL model could not be attributed to increased biliary duct proliferation, because immunohistochemical staining of bile ducts was equivalent in both PiZ BDL and wild-type BDL mice.
Our study therefore suggests that the PiZ mouse model is more susceptible to apoptosis after BDL when compared with wild-type controls and PiZ shams. ER stress may play a key role in the higher number of apoptotic hepatocytes noted in PiZ BDL mice. Hepatocyte apoptosis would then lead to the production of profibrogenic molecules, such as TGF-β1, which would activate HSCs and lead to the development of fibrosis (Fig. 8). Our results indicate that patients with α1-AT deficiency may be susceptible to the development of more severe liver disease if suffering from concomitant cholestatic liver injury. We also can speculate that the α1-ATZ mutation also may act as a modifier gene that exacerbates liver injury in the setting of cholestatic liver diseases such as cystic fibrosis.