Saito, Takeshi; Hishiki, Tomoro; Terui, Keita; Mitsunaga, Tetsuya; Terui, Elena; Nakata, Mitsuyuki; Yoshida, Hideo
See “Biliary Atresia: New Lessons Learned From the Past” by Muraji on page 586.
Biliary atresia (BA) is a devastating pediatric hepatobiliary disorder characterized by progressive inflammatory obliteration of the extra- and intrahepatic bile ducts, leading to fatal liver cirrhosis if untreated. Even though >50% of patients undergoing Kasai portoenterostomy can recover from jaundice, many long-term survivors experience postsurgical complications such as cholangitis and portal hypertension (1). Some patients have no choice but to undergo liver transplantation (LT) because of hepatic failure or poor quality of life exacerbated by long-term sequelae.
It is accepted that there are 2 major types of BA classified based on the clinical aspects (2): fetal, embryonic, or prenatal types, characterized by congenital anomalies such as polysplenia, preduodenal portal vein, or major heart anomalies; and acquired or perinatal type, characterized by the onset of hepatobiliary obliteration at or following birth with a jaundice-free interval. The perinatal type is believed to account for 75% to 90% of all BA (3,4). It is debatable that some reported BA cases prenatally diagnosed with hilar hepatobiliary cyst (5) forming a portion of biliary cystic malformations (6) should be classified as the fetal type or considered to be another entity (7).
The etiology of BA is largely unknown, and whether the fetal and perinatal types originated from the same insult is controversial. Abnormal immunological events (8), hepatobiliary developmental disorders (9), hepatotropic viral infections (10–12), genetic susceptibilities (13,14), and toxic agents have been implicated as possible causes for BA. Several authors (15–18) have suggested the maternal microchimerism theory in would-be patients with BA, whereby certain kinds of maternal cells migrate to the fetal liver and possibly initiate the destruction of the biliary trees through a series of unknown immunological events. This theory and others lack convincing and reproducible evidence, and details of the mechanisms leading to biliary obstruction have not been elucidated.
Innate immunity is the first-line immune defense against pathogenic microorganisms and it occurs before the mobilization of adaptive immunity. Binding of conserved motifs of microbial compounds to the pattern-recognition receptors on the host's innate immune cells activates a downstream intracellular cascade (19), resulting in the production of a variety of cytokines to protect the host from pathogens.
Toll-like receptors (TLRs) are the main pattern-recognition receptors expressed on immune cells and epithelial cells, such as cholangiocytes, which constitute innate immunity. Each TLR is activated by a specific pathogen-associated molecular pattern and has its own distinctive cellular expression profile (20). There are 10 members of the TLR family in humans. Their representative ligands include lipopolysaccharide (LPS; recognized by TLR4), lipoteichoic acid and bacterial lipopeptides (TLR2), pathogenic monomeric flagellin (TLR5), the unmethylated CpG DNA of bacteria and viruses (TLR9), double-stranded (ds) RNA (TLR3), and single-stranded (ss) RNA (TLR7 and TLR8) (21).
TLR signaling is tightly regulated, primarily to avoid potential damage to the host resulting from excessive and prolonged inflammation. In addition, TLR signaling also needs to be controlled to prevent it from interfering with resident flora. The liver, which receives large volumes of blood via the portal vein, is constantly exposed to bacterial products from enteric microflora, but healthy individuals do not develop liver inflammation, implying a degree of tolerance to TLR ligands. It has been hypothesized that a breakdown of this tolerance may result in inflammatory disorders of the liver (21,22). We consider that impaired immune reactions associated with deteriorated tolerance in the hepatobiliary system against certain pathogens may contribute to the pathogenesis of BA.
In the present study, we examined TLR mRNA expression levels in the liver tissues of patients with pediatric hepatobiliary diseases, including BA, that were obtained during surgery. TLR mRNA levels were compared between patients with BA and non-BA. Among the patients with BA, we also investigated the correlation between TLR mRNA expression and age at sampling and overall prognosis.
PATIENTS AND METHODS
Patients and Clinical Data
From 1987 to 2005, 49 patients with pediatric hepatobiliary disease were enrolled after obtaining informed consent. The diseases and median age and range at time of operation were as follows: BA (n = 19; median age 83 days [range 24 days–3 years 9 months]), choledochal cyst (CC; n = 21; 3 years 6 months [69 days–13 years 7 months]), and other entities (other diseases; include neonatal hepatitis in 1 and hepatoblastoma in 4; n = 9; 1 year 6 months [57 days to 10 years 4 months]). Liver tissue was obtained during surgery, immediately snap-frozen in liquid nitrogen in the operating room, and stored in liquid nitrogen until nucleic acid extraction. The samples were classified into the following groups based on the type of surgery performed: Kasai portoenterostomy for BA (early BA group; n = 13; median age at sampling 66 days [range 24–128 days], subsequent reoperation or LT for post-Kasai BA (late BA group; n = 11; 244 days [54 days–3 years 9 months]), hepaticojejunostomy for CC (CC group), and hepatectomy for hepatoblastoma and open laparotomy for the other diseases (other diseases group). For age-matched control comparison with the early BA group, 7 cases consisting of 4 with CC and 3 with other diseases were selected and assorted into an age-matched non-BA group. Demographic data of each group are shown in Table 1. In addition, BA tissues obtained during the first surgery (early BA group) were classified based on whether the patient later underwent LT (LT group; n = 4; median age of 93 days [range 66–128 days]) or maintained native liver (non-LT group; n = 9; 49 days [range 24–83 days]).
RNA Extraction and Reverse Transcription
RNA was extracted from liver tissue using Trizol (Invitrogen, Carlsbad, CA), purified with the Qiagen RNeasy Mini kit (Qiagen, Hilden, Germany), and DNase treated using the RNase-free DNase Set (Qiagen). Samples were stored at −80°C until use. Reverse transcription (RT) was performed by applying 1 μg of total RNA to the Transcriptor First Strand cDNA Synthesis kit (Roche Diagnostics, Mannheim, Germany) following the manufacturer's instructions.
The LightCycler-primer sets for GAPDH and TLRs were purchased from Roche Diagnostics (Roche Diagnostics). TLR primers for quantitative polymerase chain reaction (PCR) were designed based on the custom-made ProbeFinder version 2.40 for human samples (Roche Diagnostics) and purchased from Nihon Gene Research Laboratories (Sendai, Japan). Among the TLR members in humans, TLRs 2, 3, 4, 7, and 8 were selected for the investigation based on availability of primers. The primer nucleotide sequences are shown in Table 2.
Real-time Quantitative RT-PCR and Comparative Analysis
Steady-state mRNA levels were measured by real-time quantitative RT-PCR (QRT-PCR) using the LightCycler 2.0 System (Roche Diagnostics). Each 20-μL PCR reaction mixture contained 5 μL of cDNA template (3.3 ng/μL), 100 nmol/L Universal Probe Library probe (Roche Diagnostics), 200 nmol/L forward and reverse primers, and 4.0 μL of LightCycler TaqMan Master (Roche Diagnostics). Amplification was performed with an initial denaturation at 95°C for 10 minutes, followed by 45 cycles of 95°C for 10 seconds, 68°C for 10 seconds, and 72°C for 16 seconds. Melting curve analysis was performed to verify the absence of nonspecific products such as dimers. The cDNA concentration of each gene was normalized to that of GAPDH, a housekeeping gene with a relatively constant mRNA expression level, in each sample.
Nonparametric analysis was used to compare individual TLR mRNA expression levels. That is, the Mann-Whitney U and Kruskal-Wallis tests were adopted for comparisons between 2 and 3 groups, respectively. The Spearman correlation coefficient was used to evaluate the age dependence of hepatic TLR mRNA expression. Data were expressed as the mean ± standard deviation or standard error and defined as significant at P < 0.05.
Comparison of Targeted TLR mRNA Expression Levels in Pediatric Patients With Early BA, CC, and Other Diseases
Figure 1 shows the comparison of hepatic TLR mRNA expression among disease groups: early BA, CC, and other diseases. TLR8 mRNA level was highest in BA (P = 0.008). Pairwise comparison of the expression levels between groups showed that TLR8 mRNA expression level was significantly higher in early BA than in CC (P = 0.004) or other diseases (P = 0.02). For TLRs 2, 3, 4, and 7, no significant differences were found among the 3 groups. TLR7 mRNA expression level was higher for BA than for CC and other diseases, but these differences were not significant.
Comparison of TLR mRNA Expression Levels Between Early BA And Age-matched Non-BA Groups
To minimize the effect of age on the difference of hepatic TLR mRNA levels among groups as much as possible, we selected patients younger than 1 year at surgery from among those with CC and other diseases, which were classified to the age-matched non-BA group. Table 1 shows that between early BA and age-matched non-BA groups, significant differences were not found concerning age at operation and laboratory data, except for bilirubin level. Comparison of TLR mRNA expression between the groups shows that only differences in TLR8 were significant (Table 3).
Comparison of TLR mRNA Expression Levels Between Early and Late BA Groups
We investigated the relation between TLR mRNA levels and age at sampling among the patients with BA (Fig. 2) and found that the early BA group showed significantly higher TLR2 and -8 mRNA expression levels (P = 0.02 and 0.006, respectively). The early BA group also demonstrated a tendency of higher TLR3 mRNA levels (P = 0.19). Concerning mRNA expression of TLRs 4 and 7, no significant differences were found. In addition, whether a history of cholangitis had any effect on hepatic TLR mRNA levels in the late BA group was examined. Three of 11 cases (27%) in the late BA group experienced cholangitis after the first Kasai operation. Despite a tendency for higher TLR4 and -7 mRNA levels in patients with episodes of cholangitis (P = 0.07 and 0.06, respectively), no significant difference with any TLR was found for history of cholangitis (data not shown).
Correlation Between TLR mRNA Levels in Liver Samples Taken During the First Operation and Age at Sampling
To investigate the effect of age at the first operation (age at sampling) on TLR mRNA levels in BA, the correlation coefficients were calculated and compared. The data in Table 4 demonstrate a significant correlation for TLR7 mRNA (r = 0.77, P = 0.001) and age at sampling, whereas no significant correlations were observed for TLRs 2, 3, 4, or 8. In detail, clinical data and TLR mRNA levels of each case among early BA group were depicted in Table 5.
Comparison of TLR mRNA Levels in BA Samples Taken During the First Operation and Patient Outcome
Of the BA hepatic tissues obtained during the initial operation, TLR mRNA levels were compared between patients with good outcome (non-LT group) and those with poor outcome (LT group) (Fig. 3, Table 5). The LT group had significantly higher mRNA expression levels of both TLRs 3 and 7 (P = 0.02 and 0.01, respectively), whereas no significant differences were found for TLRs 2, 4, or 8.
The onset of hepatobiliary disorders seen in BA has not been specified, despite numerous attempts. Ductal plate malformation theory attributes the cause of BA to developmental abnormalities of the hepatobiliary system during the early stage of organogenesis (9), whereas viral infection theory focuses on perinatal events associated with the host's immunological susceptibility. Other researchers have suggested different pathogeneses for fetal and perinatal types of BA (4).
Infants undergo a dramatic transition from the sheltered intrauterine environment to the radically distinct environment of the outside world, and BA presents within the first 3 months of life. Within this discrete period of time, it is plausible to infer that a pathogen could stimulate the immature immune defense system, inducing an imbalance in proinflammatory cytokines and causing intractable inflammation in the hepatobiliary duct or exacerbating inflammatory destruction that started before birth. We believe that by identifying the series of immunological events around birth in infants that comprise BA, the pathogenesis of this disease will be clarified.
Recent research has suggested that failure to regulate TLR signaling may result in the instability of cytokine networks governing both systemic and local environments. This may lead to physical disorders, including certain gastrointestinal and liver diseases. For instance, upregulated expression of TLRs 2 and 4 in intestinal epithelial cells and defective TLR responses have been reported in patients with inflammatory bowel disease (21). Additionally, TLR-mediated signals have been implicated in the progression of hepatitis B virus infection (23), hepatitis C virus (24), and primary biliary cirrhosis (25). In other words, inappropriate immunological reactions against unknown ligands via TLR cascades may trigger progressive inflammatory biliary destruction that manifests as BA. Thus, we analyzed TLR mRNA expressed in liver tissues of pediatric patients with hepatobiliary disorders, including BA.
We demonstrated that hepatic TLR8 mRNA expression was significantly higher in early BA than in (age-matched) non-BA patients. With the structure and function of TLR8 being similar to TLR7, both of these TLRs recognize ssRNA oligonucleotides from ssRNA viruses (26). Interestingly, Huang et al (27) suggested the involvement of TLR7 and its downstream type 1–specific interferon signaling in the pathogenesis of BA, based on real-time QRT-PCR analyses of fresh liver tissues. This is partly consistent with the present study. Taken together, it may be reasonable to identify viral infection as the cause of BA. It has also been suggested that not only exogenous representative pathogens such as LPS, dsRNA, and ssRNA, but also endogenous ligands derived from apoptotic or necrotic cells can activate TLR signaling, thus hinting at the possible contribution of endogenous ligands to the development of inflammation (27,28) and autoimmunity (29–31). Thus, the present study suggests the involvement of innate immunity in the development of BA, but the specific cause remains undetermined.
Some reports have focused on the involvement of TLR3, which is expressed on cholangiocytes, in the etiology of BA (32). The underlying concept of this is based on the implication of dsRNA reoviruses as the causative agents of BA (33). Harada et al (32) demonstrated that by stimulating cultured human biliary epithelial cells with a synthetic analog of viral dsRNA, TLR3 downstream transcriptional factors are activated, with 1 of these factors upregulating the expression of tumor necrosis factor–related apoptosis-inducing ligand that leads to biliary apoptosis. The authors suggested that reoviral infection and subsequent biliary epithelial apoptosis were the contributing factors. This hypothesis is intriguing; however, the findings of the present study do not necessarily support this hypothesis because TLR3 mRNA levels were similar in patients with BA and non-BA.
Several points should be addressed to prove the theory that reoviral infection and biliary epithelial apoptosis play central roles in the development of BA. First, not only the behavior of cholangiocytes but also the coordination and dynamics of other immune-related cells must be scrutinized. This is because innate immune cells such as dendritic, Kupffer, and natural killer cells typically play more decisive roles than cholangiocytes in determining the hepatic defense system. Second, the complicated interactions of concomitant TLR signaling pathways in BA should be evaluated. When specific TLR signaling is activated in vivo, a variety of inflammatory or apoptotic cytokines are produced, which may affect other TLR routes. Last, and most important, the major stumbling block for proving this hypothesis has been to collect a sufficient body of evidence of reoviral infection in patients with BA (34) to differentiate it from non-BA.
Among the possible causative hepatotropic viruses, rotavirus (11) and cytomegalovirus (12) have been examined. Despite examination of the role of TLR3 and TLR2 in rotaviral and cytomegaloviral infection, respectively, in some organs (35–37), there have been no reports investigating the effect of each viral infection in hepatocytes or cholangiocytes on intracellular TLR cascades in patients with BA. Our results showing that TLR2 and -3 mRNA expression levels in the early BA group were not significantly upregulated compared with non-BA groups do not necessarily support the involvement of these viruses in the initiation of hepatobiliary disorder in BA, and again, insufficient supporting data have been gathered for both viruses to determine any significant differences of viral infection frequencies between BA and non-BA in hepatic tissues.
Higher mRNA expression of TLRs 2, 3, 7, and 8 correlated with earlier age at sampling (Fig. 2). This finding suggests that the innate immune response culminates around the timing of the Kasai portoenterostomy and gradually abates during the postoperative period. In contrast, TLR4 expression is nearly unchanged between the early and late BA groups, which may explain the implication of TLR4 in liver fibrosis. Because TLR4 has been reported to play a pivotal role in the pathogenesis of hepatic fibrosis (38), comparatively higher TLR4 expressions in the late group may explain the progressive liver fibrosis seen in most patients with BA after Kasai portoenterostomy. In addition, the experience of cholangitis after Kasai portoenterostomy did not have any effect on TLR mRNA expression in the late BA group. Because this may be the result of a relatively small number of patients with late BA experiencing postoperative cholangitis, no conclusion can be made about the effect of past cholangitis on hepatic TLR cascades.
There is still no conclusive report identifying a relation between age and hepatic TLR mRNA levels among pediatric populations. Thus, age of the patient may strongly contribute to the comparatively higher TLR mRNA expression in the early BA group, and the correlation coefficient between TLR levels and age at operation (sampling) was examined. In our study, no significant correlation was found between mRNA expression levels of TLRs 2, 3, 4, and 8 and age at Kasai portoenterostomy, and only mRNA expression of TLR7 was significantly correlated (Tables 4 and 5). Because the expression of TLR7 is suggested to be induced during LPS-induced maturation of dendritic cells (39), greater pathogen load originating from enteral bacteria in BA at older age facilitated enhanced recognition of specific ligands by TLR7 in hepatic immune cells, which in turn activates downstream cascades, exacerbates biliary inflammation, and results in poorer prognosis. To draw more definitive conclusions about the precise relation, more extended experiments using greater numbers of samples from both pediatric patients with hepatobiliary diseases and normal controls will be required.
The present study suggests a possible relation between TLR mRNA levels in the liver obtained at the time of the Kasai portoenterostomy and the patient prognosis (Fig. 3, Table 5). For example, mRNA expression levels of TLRs 3 and 7 were significantly higher in patients who later required LT, and levels of TLR2 and TLR8 tended to be higher in patients with a subsequent poor prognosis. This indicates that quantitative measurement of the extent of inflammation as TLR mRNA levels at the time of the initial operation may be an accurate predictor of fate of the native liver in BA and may be useful for the establishment of suitable post-Kasai portoenterostomy treatment regimens.
There are some limitations with the present study. First, our findings do not necessarily reflect TLR expression in the portal tract, which is the area most affected by BA, because RNA was extracted from liver samples including both portal and mesenchymal areas. Second, our results reflect overall TLR mRNA expression levels in the diverse cell populations that constitute the liver, rather than only in the immune cells. Last, the effect of bile congestion on TLR signaling has not been taken into account. That is, some TLR profiles demonstrated in BA of the present study may be attributable to severe bile stasis rather than to causative pathogens and the subsequent immunological response.
In conclusion, we used greatly sensitive real-time QRT-PCR on liver tissue samples obtained from patients with hepatobiliary disease, including BA, to demonstrate that innate immunity may contribute to the initiation and development of BA. Furthermore, TLR mRNA levels in liver samples taken during the first surgical procedure may be useful for predicting the prognosis of the native liver in patients with BA.
1. Nio M, Ohi R, Shimaoka S, et al. The outcome of surgery for biliary atresia and the current status of long-term survivors. Tohoku J Exp Med
2. Balistreri WF, Grand R, Hoofnagle JH, et al. Biliary atresia: current concepts and research directions. Hepatology
3. Silveira TR, Salzano FM, Howard ER, et al. Congenital structural abnormalities in biliary atresia: evidence for etiopathogenic heterogeneity and therapeutic implications. Acta Paed Scand
4. Kahn E. Biliary atresia revisited. Pediatr Dev Pathol
5. Saito T, Horie H, Yoshida H, et al. The perinatal transition of the hepatobiliary cyst size provides information about the condition of bile flow in biliary cystic malformation cases. J Pediatr Surg
6. Casaccia G, Bilancioni E, Nahom A, et al. Cystic anomalies of biliary tree in the fetus: is it possible to make a more specific prenatal diagnosis? J Pediatr Surg
7. De Matos V, Erlichman J, Russo PA, et al. Does “cystic” biliary atresia represent a distinct clinical and etiological subgroup? A series of three cases. Pediatr Dev Pathol
8. Broome U, Nemeth A, Hultcrantz R, et al. Different expression of HLA-DR and ICAM-1 in livers from patients with biliary atresia and Byler's disease. J Hepatol
9. Desmet VJ. Congenital diseases of intrahepatic bile ducts: variation on the theme ‘ductal plate malformation’. Hepatology
10. Tyler KL, Sokol RJ, Oberhaus SM, et al. Detection of reovirus RNA in hepatobiliary tissues from patients with extrahepatic biliary atresia and choledochal cyst. Hepatology
11. Riepenhoff-Talty M, Gouvea V, Evans MJ, et al. Detection of group C rotavirus in infants with extrahepatic biliary atresia. J Infect Dis
12. Fischler B, Woxenius S, Nemeth A, et al. Immunoglobulin deposits in liver tissue from infants with biliary atresia and the correlation to cytomegalovirus infection. J Pediatr Surg
13. Garcia-Barcelo MM, Yeung MY, Miao XP, et al. Genome-wide association study identifies a susceptibility locus for biliary atresia on 10q24.2. Hum Mol Genet
14. Leyva-Vega M, Gerfen J, Thiei BD, et al. Genomic alterations in biliary atresia suggest region of potential disease susceptibility in 2q37.3. Am J Med Genet A
15. Suskind DL, Rosenthal P, Heyman MB, et al. Maternal microchimerism in the livers of patients with biliary atresia. BMC Gastroenterol
16. Muraji T, Hosaka N, Irie N, et al. Maternal microchimerism in underlying pathogenesis of biliary atresia: quantification and phenotypes of maternal cells in the liver. Pediatrics
17. Hayashida H, Nishimoto Y, Matsuura T, et al. The evidence of maternal microchimerism in biliary atresia using fluorescent in situ hybridization. J Pediatr Surg
18. Kobayashi H, Tamatani T, Tamura T, et al. Maternal microchimerism in biliary atresia. J Pediatr Surg
19. Medzhitov R, Janeway C Jr. Innate immune recognition: mechanisms and pathways. Immunol Rev
20. Akira S. Toll-like receptors and innate immunity. Adv Immunol
21. Testro AG, Visvanathan K. Toll-like receptors and their role in gastrointestinal disease. J Gastroenterol Hepatol
22. Seki E, Brenner DA. Toll-like receptors and adaptor molecules in liver disease: update. Hepatology
23. Visvanathan K, Skinner NA, Thompson AJ, et al. Regulation of Toll-like receptor-2 expression in chronic hepatitis B by the precore protein. Hepatology
24. Dolganiuc A, Norkina O, Kodys K, et al. Viral and host factors induce macrophage activation and loss of Toll-like receptor tolerance in chronic HCV infection. Gastroenterology
25. Mao TK, Lian ZX, Selmi C, et al. Altered monocyte responses to defined TLR ligands in patients with primary biliary cirrhosis. Hepatology
26. Kawai T, Akira S. Innate immune recognition of viral infection. Nat Immunol
27. Huang YH, Chou MH, Du YY, et al. Expression of Toll-like receptors and type 1 interferon specific protein MxA in biliary atresia. Lab Invest
28. Kariko K, Ni H, Capodici J, et al. mRNA is an endogenous ligand for Toll-like receptor 3. J Biol Chem
29. Takii Y, Nakamura M, Ito M, et al. Enhanced expression of type I interferon and Toll-like receptor-3 in primary biliary cirrhosis. Lab Invest
30. Lovgren T, Eloranta ML, Bave U, et al. Induction of interferon-alpha production in plasmacytoid dendritic cells by immune complexes containing nucleic acid released by necrotic or late apoptotic cells and lupus IgG. Arthritis Rheum
31. Ronnblom L, Eloranta ML, Alm GV. Role of natural interferon-alpha producing cells (plasmacytoid dendritic cells) in autoimmunity. Autoimmunity
32. Harada K, Sato Y, Itatsu K, et al. Innate immune response to double-stranded RNA in biliary epithelial cells is associated with the pathogenesis of biliary atresia. Hepatology
33. Morecki R, Glaser JH, Cho S, et al. Biliary atresia and reovirus type 3 infection. N Engl J Med
34. Saito T, Shinozaki K, Matsunaga T, et al. Lack of evidence for reovirus infection in tissues from patients with biliary atresia and congenital dilatation of the bile duct. J Hepatol
35. Sato A, Iizuka M, Nakagomi O, et al. Rotavirus double-stranded RNA induces apoptosis and diminishes wound repair in rat intestinal epithelial cells. J Gastroenterol Hepatol
36. Brown RA, Gralewski JH, Razonable RR. The R753Q polymorphism abrogates Toll-like receptor 2 signaling in response to human cytomegalovirus. Clin Infect Dis
37. Compton T, Kurt-Jones EA, Boehme KW, et al. Human cytomegalovirus activates inflammatory cytokine responses via CD14 and Toll-like receptor 2. J Virol
38. Seki E, De Minicis S, Osterreicher CH, et al. TLR4 enhances TGF-beta signaling and hepatic fibrosis. Nat Med
39. Severa M, Remoli ME, Giacomini E, et al. Sensitization to TLR7 agonist in IFN-beta-preactivated dendritic cells. J Immunol