Oligoclonal Expansions of CD4⁺ and CD8⁺ T-Cells in the Target Organ of Patients With Biliary Atresia

Abstract

Background & Aims:

Biliary atresia is an inflammatory, fibrosclerosing neonatal cholangiopathy, characterized by a periductal infiltrate composed of CD4⁺ and CD8⁺ T cells. The pathogenesis of this disease has been proposed to involve a virus-induced, subsequent autoreactive T cell-mediated bile duct injury. Antigen-specific T-cell immunity involves clonal expansion of T cells expressing similar T-cell receptor (TCR) variable regions of the β-chain (Vβ). We hypothesized that the T cells in biliary atresia tissue expressed related TCRs, suggesting that the expansion was in direct response to antigenic stimulation.

Methods:

The TCR Vβ repertoire of T cells from the liver, extrahepatic bile duct remnants, and peripheral blood of biliary atresia and other cholestatic disease controls were characterized by fluorescent-activated cell sorter analysis, and TCR junctional region nucleotide sequencing was performed on expanded TCR Vβ regions to confirm oligoclonality.

Results:

FACS analysis revealed Vβ subset expansions of CD4⁺ and CD8⁺ T cells from the liver or bile duct remnant in all patients with biliary atresia and only 1 control. The CD4⁺ TCR expansions were limited to Vβ3, -5, -9, and -12 T-cell subsets and the CD8⁺ TCR Vβ expansions were predominantly Vβ20. Each Vβ subset expansion was composed of oligoclonal populations of T cells.

Conclusions:

Biliary atresia is associated with oligoclonal expansions of CD4⁺ and CD8⁺ T cells within liver and extrahepatic bile duct remnant tissues, indicating the presence of activated T cells reacting to specific antigenic stimulation. Future studies entail identifying the specific antigen(s) responsible for T-cell activation and bile duct injury.

Biliary atresia (BA) is a progressive, inflammatory cholangiopathy of infancy that leads to fibrosis and obstruction of extra- and intrahepatic bile ducts with eventual biliary cirrhosis. Due to the progression of disease, it is the most common indication for liver transplantation in children. The etiology of BA is not known, but one theory regarding its pathogenesis is that bile duct damage occurs secondary to a viral-induced, autoreactive T-cell-mediated process directed against bile duct epithelia.^1-7 In the rotavirus-induced murine model of BA, we have recently demonstrated that adoptive transfer of hepatic T cells leads to a bile duct-specific inflammatory reaction in recipient severe combined immunodeficiency disease mice, suggesting cellular autoimmunity.⁸ To date, little information exists regarding the characterization of the T-cell populations infiltrating the intrahepatic and extrahepatic bile ducts in human BA and forms the basis of this study.

T cells express the T-cell receptor (TCR) for antigen on their surface, which is composed of a heterodimer of disulfide-linked α- and β-chains.⁹ Functional TCR α-chain (TCRA) and β-chain (TCRB) genes result from somatic rearrangement of germline gene segments during T-cell development. For example, TCRB genes are generated from the rearrangement of the variable (BV), diversity (BD), and junctional (BJ) region gene segments as well as random nucleotide additions and/or deletions at the junctions (Figure 1). This process makes the potential TCR repertoire enormous with greater than 10⁷ sequence possibilities.¹⁰ Thus, it is highly improbable that any 2 T cells will express identical or similar TCRs, resulting in a heterogeneous TCR repertoire.^11,12

Figure 1.

The sequence of events in recombination and gene expression of the TCR β-chain. Expressed TCR β-chain (TCRB) genes are generated from the rearrangement of variable (V) to diversity (D) to junctional (J) region gene segments. Further diversification is generated from the presence of random nucleotide additions and deletions at the V-to-D and D-to-J joining points. Shown here is an example of the TCRB region encoded by the selection of exons Vβ1, Dβ1, the second exon in the Jβ1 cluster, and the constant region Cβ1. This process makes the potential TCR repertoire enormous with greater than 10⁷ sequence possibilities.

The highly variable TCRB junctional region forms the complementary determining region 3 (CDR3), which is critically involved in the TCR's interaction with the antigen/major histocompatibility complex (MHC). In response to the appropriate antigen/MHC combination, specific T cells become activated, are recruited to target organs, and undergo clonal expansion (ie, now many T cells express an identical TCR). These oligoclonal T-cell populations can often be identified through detection of an altered TCR Vβ repertoire on organ-specific T cells.¹² Oligoclonal expansions have been identified in infection, autoimmunity, and malignancy. Increased expression of a particular TCR Vβ family is a sign of clonal proliferation that can be confirmed through TCRB junctional region sequencing of the T-cell population.^9-12 In the present study, we hypothesized that T cells present in the portal tracts and extrahepatic bile duct remnant of BA patients are composed of oligoclonally expanded T-cell populations, suggesting their accumulation in response to conventional antigenic stimulation.

Materials and Methods

Study Population

Between 2004 and 2006, peripheral blood mononuclear cells (PBMCs), fresh liver wedge tissue, and extrahepatic bile duct remnant tissue were collected from 5 patients with BA at the time of the Kasai portoenterostomy. Extrahepatic bile duct remnant tissue only was obtained from 1 BA patient at the time of liver transplantation (BA 6). Liver wedge tissue was collected from 3 control patients with other cholestatic liver diseases, 2 subjects with TPN-related cholestasis and 1 with Alagille's syndrome. Extrahepatic bile duct tissue was collected from 3 control patients with bile duct lesions, 2 subjects with choledochal cysts at the time of surgical repair, and 1 with cystic fibrosis liver disease at time of liver transplant (Table 1). Exclusion criteria included known bacterial or viral infection at the time of surgery. This study was approved by the Colorado Multiple Institutional Review Board, The Children's Hospital, and informed consent was obtained from the parents.

Table 1.

Characteristics of Study Subjects

Patient	Age	Diagnosis	Bilirubin (mg/dL-tot./direct)
BA1	9.5 wk	Biliary atresia	11.5/8.3
BA2	10 wk	Biliary atresia	7.9/6.2
BA3	6 wk	Biliary atresia	8.3/5.9
BA4	9.5 wk	Biliary atresia	7.1/5.2
BA5	16 wk	Biliary atresia	11.1/6.4
BA6	11 mon	Biliary atresia	16.3/11.8
Control liver 1	10 wk	TPN-related cholestasis	14.9/12.3
Control liver 2	4 wk	TPN-related cholestasis	4.8/4
Control liver 3	24 mon	Alagille's syndrome	14.3/11.7
Control duct 1	3 yr	Choledochal cyst	1/0.2
Control duct 2	16 yr	Choledochal cyst	0.4/0.1
Control duct 3	12 mon	Cystic fibrosis	57.3/36

Isolation and Culture of PBMC and Mononuclear Cells from Tissue

PBMCs were isolated by density gradient separation (Amersham, Uppsala, Sweden) and resuspended in a fluorescence-activated cell sorter (FACS) buffer for analysis. Liver and extrahepatic bile duct tissue were finely minced, and each individual piece was placed into a well of a 48-well plate with RPMI 1640 media (Invitrogen, Carlsbad, CA) supplemented with 20 U/mL of r-interleukin-2 (IL-2) (R&D Systems, Minneapolis, MN). T cells activated through TCR engagement become more responsive to IL-2, leading to their preferential expansion in culture.¹³ Tissue remained in culture for 2 weeks and 2 × 10⁶ cultured immune cells were resuspended in FACS buffer while any remaining cells were cryopreserved for future T-cell subset isolation.

Immunofluorescence Analysis of TCR Variable Regions of the β-Chain (Vβ) Expression

PBMCs and cultured liver and duct remnant T cells were analyzed by 3-color immunofluorescence using monoclonal antibodies (mAbs) specific for CD4, CD8 (BD Biosciences, San Jose, CA), and TCR V region expression using mAbs directed against 16 different TCR Vβ receptor subfamilies:¹⁴ Vβ2 (MPB2D5), Vβ5.1 (IMMU157), Vβ12 (VER2.32.1), Vβ13.1 (IMMU 222), Vβ14 (CAS1.1.3), Vβ17 (E17.5), Vβ20 (ELL1.4), Vβ22 (Immu546), Coulter-Immunotech, Fullerton, CA; Vβ3.1 (8F10), Vβ5.2 (F3.15.13), Vβ6.7 (OT145), Vβ8.1 (MX-6) (Endogen, Rockford, IL); Vβ7 (3G5.5), Vβ13.2 (H132.8), gifts from P. Marrack, Denver, CO; Vβ9 (FIN9), Vβ23 (AHU17), Pharmingen, San Jose, CA. The lymphocyte population was analyzed using a FACSCalibur cytometer (BD Biosciences, San Jose, CA) as previously described.¹² Tissue-specific TCR Vβ expansions were defined as a ≥3-fold increase in the level of expression of a particular Vβ compared to its expression in blood.¹⁴

Analysis of TCRB Gene Expression in Tissue T Cells

Highly purified T-cell subsets (CD3⁺/CD4⁺, CD3⁺/CD8⁺) from previously cultured immune cells were obtained using a MoFlo FACS (Dako Cytomation, Carpinteria, CA) with a purity yield of 95%–99% of gated cells. Cellular RNA was extracted using Trizol as previously described⁶ and cDNA synthesized using the Advantage RT for polymerase chain reaction (PCR) kit (Clontech, Mountain View, CA). PCR products were generated utilizing HotStarTaq-polymerase (Qiagen, Valencia, CA), 3′ primer CB (5′-TTCTGATGGCTCAAACAC-3′) and the specified Vβ 5′ primer (see Supplemental Table 1 for primer sequences [see supplemental material online at www.gastrojournal.org]). Each PCR product was ligated into a pCR 2.1 cloning vector and transformed into competent Escherichia coli (TA Cloning Kit, Invitrogen, Carlsbad, CA). Twenty-four to 48 colonies containing insert were randomly selected for nucleotide sequencing. Cycle sequencing was performed using M13 reverse (5′ -CAGGAAACAGCTATGAC-3′) sequencing primer, dRhodamine sequencing solution, and an automated ABI 377 sequencer (Applied Biosystems Inc, Foster City, CA).

HLA Genotyping

High-resolution HLA genotyping (HLA-A, -B, -C, and HLA-DR, -DQ) was performed by standard molecular techniques at ClinImmune Labs (Aurora, CO).

Statistical Analysis

All quantification of T-cell subsets is shown as the mean ± SEM. The following statistical methods were utilized to determine the significance of differences between the various study groups: Mann–Whitney test for cell surface marker expression, Student's t test for total T-cell populations and Fisher exact test for the presence or absence of TCR Vβ expansions (In Stat Software, San Diego, CA). A P value of < .05 was considered statistically significant.

Results

Subtype and Distribution of Tissue T Cells in BA

Previous studies of BA have shown that both the extrahepatic and intrahepatic bile ducts are surrounded by a mixed inflammatory cellular infiltrate, composed primarily of CD4⁺ and CD8⁺ T cells and macrophages.^6,15-17 In this study, the presence of periductal inflammation was documented by H&E staining of a portion of each tissue specimen collected, as shown in Figure 2. Extensive periductal inflammation was observed within the portal tracts and extrahepatic bile duct remnants of BA patients. Liver tissue controls likewise had portal tract infiltrates; however, minimal periductal inflammation was observed in the extrahepatic bile duct controls.

Figure 2.

Representative histology of liver and extrahepatic bile duct remnant from BA and control tissue. Marked inflammation was observed within the portal tracts (A) and extra-hepatic bile duct remnant (B) of BA patients. Intense portal tract inflammation was also seen within the liver of control subjects (C, TPN-related cholestasis liver), while mild inflammation was present in extrahepatic bile duct tissue from control subjects (D, choledochal cyst).

In BA, the infiltrating portal tract T cells are highly activated¹⁵; however, the relative distribution and phenotype of these activated T cells was not known. Therefore, we next determined the yield and subtype (CD4⁺ or CD8⁺) of activated T cells in BA and control specimens. The liver and extrahepatic bile duct remnant tissues were placed in culture for 2 weeks in the presence of IL-2, leading to the preferential expansion of previously activated T cells. The amount of BA or control liver tissue placed in culture was approximately 10–20 mg for each specimen. Greater numbers of immune cells were isolated from BA liver tissue (mean ± SEM: 5.1 ± 1.9 × 10⁶) compared with controls (3.0 ± 0.6 × 10⁶) (Figure 3A). Approximately 20 mg of extrahepatic bile duct remnant tissue from BA patients was cultured compared to 200 mg of extrahepatic bile duct tissue from controls. Despite this 10-fold greater amount of tissue available from the controls, the BA duct tissue yielded almost twice as many immune cells in culture (BA: 3.6 ± 0.7 × 10⁶; controls: 1.9 ± 0.6 × 10⁶).

Figure 3.

High yield of immune cells from BA tissue with a predominance of CD8⁺ T cells in liver tissue and CD4⁺ T cells in the extrahepatic duct remnant. (A) After 2 weeks in culture with IL-2, total cell yield was determined with a hemocytometer. Greater numbers of immune cells were obtained from BA livers and extrahepatic duct remnants compared to controls. (B) Percentage of total cells isolated in culture that were positive for the T-cell marker CD3. Significantly more CD3⁺ T cells were present in the liver from BA patients compared to controls (*P < .05). (C) Representative FACS analysis density plots of isolated T cells from BA specimens and controls, gated on all cells. Shown is a representative percentage of CD3⁺ T cells that were CD4 or CD8 positive in each group. (D) Summary of cell surface expression of CD3, CD4, and CD8. Significantly greater percentages of CD3⁺CD8⁺ T cells and smaller percentages of CD3⁺CD4⁺ T cells were detected from BA livers compared with liver controls (*P < .05).

Analysis of the types of immune cells cultured from the tissue revealed that the majority of the cells were CD3⁺ T cells. BA livers contained significantly more CD3⁺ T cells compared to control livers (CD3⁺: BA, 79.5 ± 3.5%; controls, 53.9 ± 3.6%, P = .003). However, similar percentages of T cells were found in the BA bile duct remnant and controls (Figure 3B). The normal liver also harbors a generous amount of natural killer (NK) cells, and thus, further analysis of the cultured liver cells for NK cells and NKT cells was performed. Significantly less NK cells (CD56⁺CD3⁻) were present in BA livers compared to controls (BA, 8.8 ± 3.9%; controls, 32 ± 4.8%, P = .007). Regarding NKT cells, the majority of liver NKT cells have a restricted TCR-α chain expressing Vα24 and this cell surface marker is used routinely to identify classic NKT cells.¹⁸ No expansions of TCR Vα24⁺ cells within cultured lymphocytes were identified in either group (BA, 0.50 ± 0.14%; controls, 0.47 ± 0.12%).

Quantification of T-cell subtypes was determined by FACS analysis, and the percentage of CD3⁺ T cells expressing CD4 or CD8 was determined. Representative FACS density plots of cultured cells from BA and controls are shown in Figure 3C. Significantly less CD4⁺ T cells (24.4 ± 3.1%) and more CD8⁺ T cells (38.1 ± 5.6%) were present from BA liver tissue compared to controls (CD4: 60.4 ± 8.1%; CD8: 17.9 ± 1.9%) (P < .05). In contrast, cells from the BA extrahepatic bile duct remnants were mainly composed of CD4⁺ T cells (50.9 ± 7.9%), the predominant phenotype also present in control duct tissues (55.8 ± 10.2%) (Figure 3D).

Limited TCR Vβ Repertoire Expression on CD4⁺ T Cells from BA Livers and Extrahepatic Bile Duct Remnants

To identify which TCR Vβ subsets may harbor populations of oligoclonally expanded T cells, PBMCs and cultured liver and bile duct remnant T cells were analyzed for CD4 and TCR Vβ region expression by FACS analysis using mAbs directed against 16 different TCR Vβ subsets. For all studies, TCR Vβ expression of organ-specific T cells was considered significantly increased when the percentage of a specific TCR Vβ was greater than 3-fold over baseline TCR Vβ expression from the patient's PBMCs.^12,14 As a control, to determine if culturing cells in the presence of IL-2 would skew TCR Vβ cell surface expression, perihepatic lymph nodes isolated from 2 patients were cultured for 2 weeks in media alone or media supplemented with IL-2, followed by TCR Vβ analysis of both CD4⁺ and CD8⁺ T cells. No skewing of the TCR Vβ repertoire was observed in lymph node T cells cultured for 2 weeks in the presence of IL-2 compared with media alone or PBMCs (data not shown).

TCR Vβ repertoire analysis indicated that all BA tissue specimens (livers or bile duct remnants) contained 1 to 2 significant CD4⁺ TCR Vβ expansions (Figure 4A). The predominant Vβ subsets expanded included Vβs 3, 5, 9, and 12. Patient BA6 had the greatest expansion of a single Vβ, with over 50% of all CD4⁺ T cells from the bile duct remnant expressing Vβ12, compared with 2% of the PBMCs.

Figure 4.

Expansion of CD4⁺ T cells with limited Vβ repertoire in BA livers and extrahepatic duct remnants. TCR Vβ repertoire of cultured CD4⁺ T cells from liver (shaded black), bile duct tissue (shaded gray), and PBMCs (white) in BA patients (A) and controls (B). Asterisks denote expanded TCR Vβ regions (>3-fold increase over PBMCs) in individual patients. All BA tissue contained 1 or 2 CD4⁺ TCR Vβ expansions. Shown in B are the mean ± SEM percent of CD4⁺ T cells expressing each of the Vβ subtypes from control tissues. Only 1 control tissue (choledochal cyst bile duct) had a significant CD4⁺ TCR Vβ expansion (Vβ17).

In contrast to BA, only 1 of the 6 control tissues demonstrated a CD4⁺ TCR Vβ expansion. Figure 4B shows the percentage of CD4⁺ T cells expressing each of the Vβ subsets from control tissues. A single choledochal cyst remnant had a significant expansion of Vβ17-expressing CD4⁺ T cells, with 26% of these cells from the bile duct expressing Vβ17 compared with 5.1% of PBMCs. To summarize, all 6 BA patients demonstrated at least 1 CD4⁺ T cell expansion compared with only 1 expansion in the 6 control patients (P = .015). The CD4⁺ T cells found within diseased liver and bile duct remnants of BA subjects utilized a limited TCR Vβ repertoire, suggesting that the distribution of Vβs found was not random but in response to an antigen driven expansion.

TCR CDR3 Sequence Analysis Confirms Oligoclonality of Expanded CD4⁺ TCR Vβ Subsets in BA

To determine if the TCR Vβ expansions identified in BA contained oligoclonal populations, the TCRB junctional regions were sequenced. The prevalent, deduced amino acid sequences of the different clones are summarized in Figure 5, with the complete sequences of all identified clones in Supplemental Figure 1 (see supplemental material online at www.gastrojournal.org). All TCR Vβ expansions identified by FACS analysis were confirmed to be oligoclonal by nucleotide sequence determination. A minimum of 3 identical sequences from different clones defined a clonal expansion.^12,14 Dominant clones that constituted ≥40% of a particular CD4⁺ TCR Vβ subset were present in 11 of 13 (84%) BA samples tested. In several instances, almost the entire Vβ subset was dominated by a single clone with ≥90% of clones expressing identical TCRs (specimens BA1, BA3, and BA6).

Figure 5.

TCRB junctional region amino acid sequences expressed in CD4⁺ T cells from liver and extrahepatic duct remnants of BA patients. Dominant TCRB sequences are shown here. For each T-cell clone the entire TCRB junctional region is shown, extending from the 5′ end of the selected TCRBV family gene, including the highly rearranged nBDn gene segment, and ending at the selected BJ gene segment. The column TCRBJ denotes for each clone for which a BJ gene family member was selected during genetic rearrangement. The number of identical sequences is shown over the total number of sequences analyzed for a given sample and is depicted as percentage of frequency.

Striking Expansions of CD8⁺ T Cells Expressing TCR Vβ20 in BA Liver and Extrahepatic Bile Duct Remnants

TCR Vβ subsets expressed on CD8⁺ T cells were also analyzed by FACS analysis to determine the presence of Vβ expansions in this T cell subset. As shown in Figure 6A, 4 of the 6 BA specimens revealed striking expansions of CD8⁺ T cells expressing Vβ20 from the liver and/or bile duct remnant, compared to autologous PBMCs. In specimens BA1, BA3, and BA5, both the liver and bile duct remnant contained CD8⁺ TCR Vβ20 expansions, while it was present only in the liver in BA 2. Furthermore, although not classified as an expansion by our definition, a 2.2-fold increase in TCR Vβ20 expression on CD8⁺ T cells from the extrahepatic bile duct remnant tissue of patient BA6 was also observed. In contrast, control livers and extrahepatic bile ducts did not contain CD8⁺ TCR Vβ expansions compared to autologous PBMCs (Figure 6B).

Figure 6.

Striking expansion of CD8⁺ TCR Vβ20 in BA liver and extra-hepatic duct remnants. TCR Vβ repertoire of cultured CD8⁺ T cells from livers (shaded black) and bile duct tissue (shaded gray) compared with autologous PBMCs (white) in BA patients (A) and controls (B). Asterisks denote the expanded TCR Vβ region (>3-fold increase over PBMCs) in individual patients. Four of 6 BA specimens revealed striking expansions of CD8⁺ TCR Vβ20 within the liver and ductal remnants compared with autologous PBMCs. Shown in B are the mean ± SEM percent of CD8⁺ T cells expressing each of the Vβ subsets from control tissues. None of the controls contained CD8⁺ TCR Vβ expansions.

TCR CDR3 Sequence Analysis Confirms Oligoclonality of the Expanded CD8⁺ TCR Vβ20 in BA

Oligoclonality of the CD8⁺ TCR Vβ20 expansions found in BA samples was confirmed by sequence analysis of the TCRB junctional regions (Figure 7). CD8⁺ TCR Vβ20 expansions were verified to be oligoclonal in nature by demonstrating at least 3 identical sequences from different clones. Dominant clones that constituted ≥40% of a specific CD8⁺ TCR Vβ were present in 4 of 8 (50%) BA samples tested, with a frequency of identical clones ranging from 13%–81% within a specimen.

Figure 7.

TCRB junctional region amino acid sequences expressed in CD8⁺ T cells expressing TCR Vβ20 from liver and extrahepatic duct remnants of BA patients. CD8⁺ TCR Vβ20 TCRB sequences found 3 times or more in any sample are shown here. For each T-cell clone the entire TCRB junctional region is shown. The column TCRBJ denotes for each clone for which a BJ gene family member was selected during genetic rearrangement. The number of identical sequences is shown over the total number of sequences analyzed for a given sample and is depicted as percentage frequency.

Taken together, our findings demonstrate that the lymphocytes in the livers and extrahepatic bile duct remnants of BA patients include a unique subset of previously activated oligoclonal CD4⁺ and CD8⁺ T cells, signifying expansion in response to conventional antigenic stimulation.

HLA Genotypes of Patients with Vβ Subset Expansions

The particular TCR variable region expressed by expanded T cells is a consequence of either the stimulating antigen and/or the MHC-presenting molecule. It was possible, therefore, that the predominant CD8⁺ TCR Vβ20 expression or the limited CD4⁺ TCR Vβ repertoire present in BA patients was related to a common HLA genotype. Table 2 summarizes the biliary atresia CD8⁺ and CD4⁺ TCR Vβ expansions and compares them with the MHC class I and II genotypes of these patients. No obvious association between MHC class I genotype and CD8⁺ TCR Vβ20 was observed. Interestingly, for MHC class II genotypes, 2 associations were observed. Patients BA1 and BA6 both had significant expansions of CD4⁺ T cells expressing Vβ12 and shared the HLA-DRB1*0701/DQB1*0202 alleles. Patients BA3 and BA4 both had significant expansion of TCR Vβ3 and shared HLADQB1*0301. However, the limited number of specimens available for HLA analysis in this study group prohibits any conclusion regarding a definite association of HLA genotypes and BA.

Table 2.

HLA Genotyping of Biliary Atresia Patients

Patient	Dominant CD8+ Vβ		HLA-A		HLA-B	HLA-Cw
BA1	20		0301, 1101		0801, 3501	0401, 0701
BA2	13.2, 20		2402, 3101		3901, 5102	0801, 1203
BA3	20		0201, 0301		1501, 3502	0303, 0602
BA4	—		0101, 0201		3501, 5701	0401, 0602
BA5	20		0101, 0201		0801, 5201	0303, 0701
BA6	7, 20		0201, 3101		1515, 5101	0102, 1402
Patient		Dominant CD4+ Vβ		HLA-DRB1		HLA-DQB1
BA1		9, 12		0701, 1301		0202, 0603
BA2		6, 7		1201, 1402		0301
BA3		3, 9		0103, 1104		0301, 0501
BA4		3, 5, 17		0701, 1101		0301, 0303
BA5		5		0301, 1401		0201, 0503
BA6		12		0701, 0802		0202, 0402

Discussion

We have postulated that the pathogenesis of BA involves a viral-induced, progressive autoimmune-mediated injury of bile duct epithelia.¹ In this scenario, a perinatal infection with a virus that is tropic for the bile duct epithelia would cause initial bile duct epithelial injury. Despite viral clearance, persistent inflammation and injury to the bile duct epithelia would ensue. The damaged bile duct epithelial cells may express either altered or previously sequestered “self”-antigens that are subsequently recognized as foreign, eliciting autoreactive T-cell-mediated inflammation (bystander activation pathway of autoimmunity). Alternatively, viral proteins may be structurally similar to bile duct epithelial proteins and autoimmunity could be elicited based on the molecular mimicry pathway. In support of this autoimmune theory, we have recently shown evidence for both cellular and humoral autoimmunity in a rotavirus-induced murine model of BA.⁸

Oligoclonal T-cell expansions are found in antigen-driven immune responses, including autoimmune diseases, infections, and malignancies.¹¹ In this context, our aim was to determine if the organ-specific T cells in BA were oligoclonal in nature, lending support to the pathogenesis theory described above. We found that the activated T cells from BA tissue were composed of both CD4⁺ and CD8⁺ T cell subsets within the liver and predominantly CD4⁺ T cells within extrahepatic bile duct remnants. All BA patients demonstrated CD4⁺ T-cell expansions and 5 of the 6 patients had CD8⁺ T cell expansions. Both T-cell subsets were found to be oligoclonal in nature with the CD4⁺ T cells displaying a limited TCR Vβ repertoire and the expanded CD8⁺ T cells expressing mainly TCR Vβ20. Given the enormous diversity of the human TCR repertoire, the identification of oligoclonal T-cell populations in BA strongly suggests that the T cells are being actively recruited to the liver and bile duct in response to conventional antigen recognition. In addition, the limited number of Vβ-expressing subsets in these BA patients suggests that the dominant immune response is directed against a finite number of antigens.

The nature of the antigen(s) being recognized by the oligoclonal T cells in BA is not known. Many researchers have demonstrated oligoclonality with a limited TCR Vβ repertoire in CD4⁺ T-cell populations from CD4⁺ T-cell-driven autoimmune diseases including autoimmune hepatitis,^19,20 Crohn's disease,²¹ and multiple sclerosis.^22-24 Interestingly, in our study 3 of the 5 BA patients had different CD4 TCR Vβ subset expansions within their liver compared with their own extrahepatic bile duct remnant. The different usage of Vβ subsets in the context of the same MHC may reflect different peptides being recognized at the site of pathology. This would not be unexpected, as many autoimmune diseases such as multiple sclerosis and type 1 diabetes mellitus have more than 1 potential stimulating autoantigen.¹¹ Furthermore, concurrent CD8⁺ T-cell oligoclonality has been shown for multiple sclerosis^25,26 and other autoimmune processes including aplastic anemia²⁷ and juvenile dermatomyositis syndrome.²⁸ The predominance of CD8⁺ TCR Vβ20 in our BA population also points to a possible role of a viral antigen activating the CD8⁺ T cells. Future studies aimed at identifying the inciting antigen(s) responsible for T-cell activation will entail analyzing candidate viral antigens including reovirus,^3,4,29 rotavirus,⁵ and cytomegalovirus,³⁰ as well as bile duct epithelial autoantigens.

Potential limitations to this study include both the inability to determine the ex vivo TCR Vβ repertoire of target organ T cells due to the small size of tissue samples and the methods used to enhance lymphocyte growth from tissue specimens. Antigen-independent IL-2 stimulated in vitro expansion of T cells leads to the preferential release and growth of activated T cells from diseased tissue. In other diseases characterized by the absence of a known antigenic stimulus such as rheumatoid arthritis³¹ or autoimmune hepatitis,¹⁹ IL-2-induced T-cell expansion has been used to study the TCR repertoire. Several observations support the validity of using this experimental technique to expand disease-relevant T cells in BA. First, elegant studies by Arenz et al³² have shown that antigen-independent IL-2 stimulation does not alter the TCR repertoire. In those studies, analysis of TCR Vβ subsets on PBMCs activated in culture with PHA, anti-CD3, or IL-2 failed to skew the Vβ repertoire from un-stimulated controls. Second, in our own studies, no significant differences in the TCR repertoire were identified from lymph node cells cultured in the absence or presence of IL-2. Finally, if culturing of T cells in the presence of IL-2 were to result in nonspecific TCR Vβ expansions, one would have expected to see expansions in all of the control tissues, which was not observed in this study.

Another potential limitation to this study is that we were not able to analyze all of the TCR Vβ subfamilies, and theoretically may have failed to identify an important T-cell expansion with the TCR Vβ antibodies used. The human TCRB repertoire consists of approximately 47 functional V gene segments (Vβs) distributed in a total of 21–23 subfamilies.³³ Many of these gene segments are infrequently expressed. The 16 different TCR Vβ subfamilies analyzed in our study cover about 2/3 of the expressed TCR repertoire. Despite this theoretic limitation, we identified at least 1 expansion in all BA patients. Furthermore, if our analysis had missed large expansions of TCR Vβ subfamilies not covered by our panel of antibodies, then one would expect a reciprocal decrease in all other Vβ subfamilies within the tissue compared with PBMCs. However, this phenomenon was not observed in BA or control tissues.

In summary, BA is associated with oligoclonal expansions of CD4⁺ and CD8⁺ T cells within liver and extra-hepatic bile duct remnant tissues, indicating the presence of activated T cells reacting to specific antigenic stimulation. Identifying the specific antigen(s) responsible for T-cell activation will form the basis of future studies and lead to an enriched understanding of the role of T cells in the bile duct injury prevalent in this disease.

Abbreviations used in this paper

BA

biliary atresia

FACS

fluorescent-activated cell sorter

IL-2

interleukin-2

mAbs

monoclonal antibodies

MHC

major histocompatibility complex

NK

natural killer

PBMCs

peripheral blood mononuclear cells

PCR

polymerase chain reaction

TCR

T-cell receptor

Vβ

variable region of the β-chain

Appendix

Supplementary Data

Supplementary data associated with this article can be found, in the online version, at doi:10.1053/j.gastro.2007.04.032.