Significance of immunoglobulin G4 (IgG4)-positive cells in extrahepatic cholangiocarcinoma: Molecular mechanism of IgG4 reaction in cancer tissue†
Article first published online: 5 JUN 2012
Copyright © 2012 American Association for the Study of Liver Diseases
Volume 56, Issue 1, pages 157–164, July 2012
How to Cite
Harada, K., Shimoda, S., Kimura, Y., Sato, Y., Ikeda, H., Igarashi, S., Ren, X.-S., Sato, H. and Nakanuma, Y. (2012), Significance of immunoglobulin G4 (IgG4)-positive cells in extrahepatic cholangiocarcinoma: Molecular mechanism of IgG4 reaction in cancer tissue. Hepatology, 56: 157–164. doi: 10.1002/hep.25627
Potential conflict of interest: Nothing to report.
- Issue published online: 3 JUL 2012
- Article first published online: 5 JUN 2012
- Accepted manuscript online: 30 JAN 2012 07:07AM EST
- Manuscript Accepted: 19 JAN 2012
- Manuscript Received: 6 NOV 2011
- Ministry of Education, Culture, Sports, Science and Technology of Japan. Grant Number: 23590393
- Health and Labour Sciences Research Grants for Research on Measures for Intractable Diseases
IgG4 reactions consisting of marked infiltration by immunoglobulin G4 (IgG4)-positive plasma cells in affected organs is found in cancer patients as well as patients with IgG4-related diseases. Notably, extrahepatic cholangiocarcinomas accompanying marked IgG4 reactions clinicopathologically mimic IgG4-related sclerosing cholangitis. The regulatory cytokine interleukin (IL)-10 is thought to induce the differentiation of IgG4-positive cells. In this study, to clarify the mechanism of the IgG4 reaction in extrahepatic cholangiocarcinoma, we investigated nonprofessional antigen-presenting cells (APCs) generating IL-10–producing regulatory T cells (anergy T cells) and Foxp3-positive regulatory cells producing IL-10. Immunohistochemistry targeting IgG4, HLA-DR, CD80, CD86, and Foxp3 was performed using 54 cholangiocarcinoma specimens from 24 patients with gallbladder cancer, 22 patients with common bile duct cancer, and eight patients with cancer of the Papilla of Vater. Moreover, a molecular analysis of Foxp3 and IL-10 was performed using a cultured human cholangiocarcinoma cell line. Consequently, 43% of the cholangiocarcinomas were found to be abundant in IgG4. Those expressing HLA-DR but lacking costimulatory molecules (CD80 and CD86) and those expressing Foxp3 detected by an antibody recognizing the N terminus accounted for 54% and 39% of cases, respectively. Moreover, the number of IgG4-positive cells was larger in these cases than in other groups. In cultured cells, the presence of a splicing variant of Foxp3 messenger RNA and the expression of IL-10 were demonstrated. Conclusion: Extrahepatic cholangiocarcinoma is often accompanied by significant infiltration of IgG4-positive cells. Cholangiocarcinoma cells could play the role of nonprofessional APCs and Foxp3-positive regulatory cells, inducing IgG4 reactions via the production of IL-10 indirectly and directly, respectively. (HEPATOLOGY 2012;56:157–164)
Biliary tract cancers can be anatomically divided into intrahepatic and extrahepatic cholangiocarcinomas, the latter including hepatic hilar cancer, common bile duct cancer, gallbladder cancer, and cancer of the Papilla of Vater. The biological behavior and carcinogenesis of each cancer differ, but the histology of most biliary tract cancers is the same as that of ordinary adenocarcinomas. In addition to neoplastic lesions, several types of cholangitis causing biliary stenosis are important in the differential diagnosis of biliary diseases. Particularly, primary sclerosing cholangitis and a complication of immunoglobulin G4 (IgG4)-related systemic diseases, IgG4-related sclerosing cholangitis, clinicopathologically mimic extrahepatic cholangiocarcinomas.
IgG4 is a minor immunoglobulin subtype composing 3%-6% of all the IgG circulating in adults,1 but is important for a systemic disorder, IgG4-related disease, that features elevated serum IgG4 levels and abundant infiltration with IgG4-positive plasma cells in affected organs.1-3 Moreover, IgG4-related cholangitis and pancreatitis (autoimmune pancreatitis, type 1) are characterized by sclerosing lesions (storiform fibrosis) and cholangiocarcinomas and pancreatic cancer usually accompany some degree of desmoplastic change and also, in some cases of pancreatic cancer, IgG4 reactions.4 Therefore, a pathological examination is necessary to differentiate IgG4-related diseases from tumors in pancreatico-biliary lesions. We have already observed that extrahepatic cholangiocarcinomas also accompany various degrees of IgG4 reactions assumed to be associated with the evasion of immunosurveillance (Kimura et al., unpublished data). However, the mechanisms inducing IgG4 reactions in cholangiocarcinoma tissue are still unknown.
Interleukin (IL)-10, a regulatory cytokine mainly produced by Foxp3+ regulatory T cells (Treg cells), T helper 2 cells, and IL-10–producing Treg cells, is thought to induce the differentiation of IgG4-positive plasma cells or favor B cell switching to IgG4 in the presence of IL-4.5, 6 The expression of Foxp3 and IL-10 has been demonstrated in several carcinoma tissues and cultured cancer cell lines, suggesting that cancer cells themselves induce the Treg cell–like immunoregulatory milieu to evade immunosurveillance.7-10.
Major histocompatibility complex class II (MHC-II)–positive cells lacking the costimulatory molecules CD80 (B7-1) and CD86 (B7-2) induce anergy to native T cells. Among T cell subsets, Treg type 1 cells characterized by the production of IL-10 are induced by immature dendritic cells (DCs).11 Moreover, costimulation-dependent T cell clones stimulated without provision of the costimulatory signal were demonstrated not to be proliferative, but to differentiate into IL-10–producing anergic T cells in primary biliary cirrhosis.12 In addition to immunocompetent cells such as DCs, nonimmunocompetent cells, including carcinoma and normal epithelial cells, have been demonstrated to express MHC-II, indicating an ability for antigen presentation, but these MHC-II–positive epithelial cells are usually called nonprofessional antigen-presenting cells (APCs), differing from professional APCs such as DCs. Several studies have suggested that antigen presentation by MHC-II–positive epithelial cells that lack costimulation signals, such as keratinocytes and pancreatic islet cells, would favor the generation of anergic T cells.13-15
It is clinicopathologically important, but practically difficult, to differentiate between IgG4-related sclerosing cholangitis and extrahepatic cholangiocarcinoma. In this study, we retrospectively evaluated IgG4-positive plasma cells in extrahepatic cholangiocarcinomas and mechanisms in terms of cholangiocarcinoma cells as nonprofessional APCs and regulatory cells. This study should help to clarify the pathological significance of IgG4 reactions in cholangiocarcinomas and also IgG4-related diseases.
Patients and Methods
Patients and Tissue Preparations.
Formalin-fixed and paraffin-embedded sections of 54 surgically resected specimens from 24 gallbladder cancers, 22 common bile duct cancers, and eight cancers of the Papilla of Vater (29 men, 25 women; average age, 74 years) were obtained from the registry of liver diseases in the Department of Pathology, Kanazawa University School of Medicine. Each cholangiocarcinoma was classified histologically as a well-differentiated (including papillary), moderately differentiated, or poorly differentiated adenocarcinoma based on the predominant histological grade. Special histological types such as adenosquamous carcinoma and mucinous carcinoma were not included in the present study. Serial sections (4 μm) were prepared from each formalin-fixed, paraffin-embedded block.
The deparaffinized and rehydrated sections were microwaved in citrate buffer (pH 6.0) for CD80 and CD86 or ethylene diamine tetraacetic acid buffer (pH 9.0) for Foxp3 for 20 minutes in a microwave oven. Following the blocking of endogenous peroxidase activity, these sections were incubated at 4°C overnight with antibodies against IgG4 (mouse monoclonal; diluted 1:200; Southern Biotech, Birmingham, AL), Foxp3 that reacts with the C terminus (mouse monoclonal; 5 μg/mL; Abcam, Tokyo, Japan), Foxp3 that reacts with the N terminus (rat monoclonal, 2.5 μg/mL, eBioscience, San Diego, CA), HLA-DR (mouse monoclonal, 0.5 μg/mL, Dako Japan, Tokyo), CD80 (rabbit monoclonal, 1:200, Epitomics, Burlingame, CA), and CD86 (rabbit monoclonal, 1:250, Abcam, Tokyo, Japan) and then at room temperature for 1 hour with anti-mouse, anti-rabbit, or anti-goat immunoglobulin conjugated to a peroxidase-labeled dextran polymer (Simple Staining Kit; Nichirei, Tokyo, Japan). After a benzidine reaction, sections were counterstained lightly with hematoxylin. No positive staining was obtained when the primary antibodies were replaced with an isotype-matched, nonimmunized immunoglobulin as a negative control of the staining procedures.
In addition to the histological observations by hematoxylin and eosin staining, the distribution of the immunopositive cells was examined. In a primary survey, we examined all tumorous areas in each specimen and, for counting IgG4-positive mononuclear cells, selected three representative areas containing IgG4-positive plasma cells, and expressed the results as the mean number of immunopositive cells in high-power fields (HPFs). Because ≥10 IgG4-positive cells/HPF is proposed according to HISORt (Histology, Imaging, Serology, Other organ involvement, Response to therapy) criteria published for autoimmune pancreatitis,16, 17 the cases with ≥10 and <10 IgG4-positive cells/HPF on average were evaluated as IgG4-rich and IgG4-poor cases, respectively. For the expression of Foxp3, HLA-DR, CD80, and CD86, positive carcinoma cells were evaluated as positive (distinct expression) or negative (no or faint expression) according to the staining intensity.
Two commercially available cell lines, HuCCTl and MCF7 (positive control of IL-10),10 were obtained from Health Science Research Resources Bank (Osaka, Japan). The cell lines were derived from cholangiocarcinoma and breast cancer cells, respectively.
Reverse-Transcription Polymerase Chain Reaction.
The cell lines were cultured in flasks with a standard medium for 48 hours. Cultured cells were collected from the flasks or plates with a cell scraper for determination of the baseline messenger RNA (mRNA) expression of Foxp3 and IL-10 by via reverse-transcription polymerase chain reaction (RT-PCR). Lymph node tissue was also used as a positive control for Foxp3 mRNA. Briefly, total RNA was isolated from each sample with the RNeasy Total RNA System (QIAGEN, Hilden, Germany) and treated with RNase-Free DNaseI. For RT-PCR, 1 μg of total RNA, M-MLV RTase (ReverTra Ace, Toyobo, Tokyo, Japan) and oligo-dT primers were used. Polymerase chain reaction (PCR) amplification was performed using DNA polymerase (Takara EX Taq, Takara, Tokyo, Japan) and specific primers for human mRNA sequences (Table 1). The glyceraldehyde 3-phosphate dehydrogenase mRNA was used as a housekeeping gene. Following After PCR, an annealing of primers for 1 minute, and an extension at 72°C for 2 minutes (the annealing temperature and cycle number are shown in Table 1), PCR products were subjected to agarose gel electrophoresis.
|Exon 1||Forward: 5′-ACCGTACAGCGTGGTTTTTC-3′||111 bp|
|Exon 3||Forward: 5′-TGCCTCCTCTTCTTCCTTGA-3′||125 bp|
|Exons 10-12||Forward: 5′- CACAACATGCGACCCCCTTTCACC -3′||167 bp|
|Reverse: 5′- AGGTTGTGGCGGATGGCGTTCTTC-3′|
|Exon 12||Forward: 5′- CAGCTGCTCGCACAGATTAC -3′||91 bp|
|Reverse: 5′- TTGGGGTTTGTGTTGAGTGA-3′|
|IL-10||Forward: 5′- TGCAAAACCAAACCACAAGA -3′||325 bp|
|Reverse: 5′- GCATCACCTCCTCCAGGTAA-3′|
|GAPDH||Forward: 5′-GCACCGTCAAGGCTGAGAAC-3′||142 bp|
Enzyme-Linked Immunosorbent Assay.
Approximately 1 × 104 HuCCT1 cells per well in 96-well plates were cultured for 24 hours. Supernatants were then tested for human IL-10 via enzyme-linked immunosorbent assay (ELISA) (R&D Systems).
Data were analyzed using the Welch t test; P < 0.05 was considered statistically significant.
Infiltration of IgG4-Positive Cells in Extrahepatic Cholangiocarcinoma.
Immunohistochemistry revealed that IgG4-positive plasma cells were scattered within and around cancerous nests to various degrees in most cases (Fig. 1). In the cases with marked infiltration, the IgG4-positive cells were prominent with intermingling of other inflammatory cells. Figure 1C shows the number of IgG4-positive cells/HPF in extrahepatic cholangiocarcinomas from common bile ducts, gallbladder, and the Papilla of Vater, but there was no significant difference in IgG4-positive cell counts among anatomical locations of extrahepatic cholangiocarcinomas. Therefore, they were integrated as shown in Fig. 1D. Consequently, the combined quantitative evaluation revealed that 23 (43%) of 54 cholangiocarcinoma patients had ≥10 IgG4-positive cells/HPF. There was no correlation between the density of IgG4-positive cells and any clinicopathological factor including age, sex, anatomical location (common bile ducts, gallbladder, and the Papilla of Vater), or the histological differentiation (well, moderate, and poor) of extrahepatic cholangiocarcinoma.
Cholangiocarcinoma Cells as Nonprofessional APCs and Their Association with IgG4 Reactions.
Representative images of immunostaining are shown in Fig. 2. Expression of HLA-DR was found in some infiltrating immunocompetent cells. Moreover, HLA-DR–positive cholangiocarcinoma cells were also found in 33 of 54 cases. HLA-DR expression in tumor cells showed uniformity and metastatic foci in lymph nodes as well as main tumors expressing HLA-DR. In contrast, the expression of costimulatory molecules (CD80 and CD86) was mostly faint or absent. Only four cases were clearly positive for CD86 in cholangiocarcinoma cells, and all of them were positive for HLA-DR. No cases evidently expressed CD80. Cholangiocarcinoma cells expressing HLA-DR but lacking costimulatory molecules (CD80 and CD86) were found in 29 of 54 cases (54%) and suggested to act as nonprofessional APCs inducing IL-10–producing anergy T cells. The relation between IgG4 reactions and HLA-DR and costimulatory molecules in cancer cells is shown in Fig. 3. In cases of positivity for HLA-DR and negativity for costimulatory molecules, the number of IgG4-positive cells was significantly higher than in cases of negativity for HLA-DR and of positivity for both HLA-DR and costimulatory molecules.
Cholangiocarcinoma Cells as Regulatory Cells.
Immunohistochemistry using the antibody reacting with the C terminus of Foxp3 detected only mononuclear cells (Treg cells), but the antibody reacting with the N terminus highlighted cholangiocarcinoma cells as well as Treg cells (Fig. 4A). The cytoplasm as well as nucleus of tumor cells was positive in several cases. However, because Foxp3 is a transcription factor, the nuclear pattern was evaluated as functional expression. Consequently, 21 of 54 (39%) cholangiocarcinomas tested positive for Foxp3 by the antibody reacting with the N terminus. The relation between the IgG4 reaction and Foxp3 expression in cholangiocarcinoma cells is shown in Fig. 5. In cases of positivity for Fopx3, the number of IgG4-positive cells was significantly higher than in cases of negativity for Foxp3.
RT-PCR analysis demonstrated that a cholangiocarcinoma cell line, HuCCT1, expressed the mRNA of Foxp3, but close examination using four sets of primers corresponding to exons 1, 3, 10-12, and 12 revealed a lack of exon 3 (Fig. 6), suggesting the presence of a splicing variant of Foxp3 in cholangiocarcinoma cells. Moreover, RT-PCR and ELISA revealed that HuCCT1 cells expressed IL-10 mRNA (Fig. 6) and protein in the culture medium at 7.8-15.6 pg/mL.
IgG4 is important to the pathogenesis of IgG4-related diseases. However, patients with pancreatic adenocarcinomas accompanying IgG4 reactions and/or elevated serum IgG4 levels4, 18-20 and with pancreatic and biliary cancers arising from IgG4-related diseases20-22 have been reported, though a cause-and-effect relationship between IgG4 reactions and cancers has yet to be demonstrated. Moreover, in IgG4-nonrelated diseases, including primary sclerosing cholangitis, IgG4 reactions were found to various degrees.23, 24 Therefore, the presence of IgG4-positive cells is not a histological hallmark of IgG4-related diseases, and IgG4 reactions are speculated to be nonspecific in several pathological conditions, including cholangiocarcinomas. The present study also demonstrated the presence of extrahepatic cholangiocarcinoma cases with abundant IgG4 reaction, though there was no significant difference in IgG4-positive cell counts among anatomical locations of extrahepatic cholangiocarcinomas (common bile ducts, gallbladder, and the Papilla of Vater). The significance and mechanisms of IgG4 reactions in cancers as well as IgG4-related diseases are still unknown, but we speculated that cancer cells directly participate in the histogenesis of IgG4 reactions. Because the regulatory cytokine IL-10 is known to induce the differentiation of IgG4-positive plasma cells or favor B cell switching to IgG4 in the presence of IL-4,5, 6 we noted the IL-10–related regulatory cytokine network around cholangiocarcinoma tissue in this study.
Immunohistochemistry for MHC-II (HLA-DR) and costimulatory molecules (CD80 and CD86) revealed that cholangiocarcinoma cells as well as professional APCs such as B cells and DCs expressed HLA-DR and CD86. The expression of CD80 was limited in some APCs and not found in cholangiocarcinoma cells. Consequently, cholangiocarcinoma cells expressing HLA-DR, but lacking costimulatory molecules (CD80 and CD86) were found in 54% of cases. These cancer cells could act as nonprofessional APCs, possibly generating IL-10–producing Treg cells (anergy T cells), and then an IL-10–predominant cytokine milieu could cause the induction of IgG4-positive cells.5, 6 In these phenotypic cases, the number of IgG4-positive cells infiltrating carcinoma tissue was higher than in HLA-DR–negative cases and both HLA-DR– and CD86-positive cases, confirming this speculation. Cells positive for both HLA-DR and CD86 are suggested to play the role of professional APCs, as it was reported that MHC-II–positive thyroid epithelial cells could present antigens to T cells and activate autoreactive T cells.25, 26 Although further study is needed to clarify the functional mechanism of these cholangiocarcinoma cells as APCs, this study demonstrated that HLA-DR– and CD86-positive cancer cells were not associated with IgG4 reactions in cholangiocarcinoma tissue.
As to pathogenesis of IgG4 reactions in IgG4-related diseases, the participation of CD4+CD25+Foxp3+ Treg cells, which are capable of producing IL-10, has been speculated.27 Foxp3 is thought to be the master transcription factor of Treg cells and, until recently, Foxp3 expression was thought to be restricted to the T cell lineage. However, immunohistochemistry and flow cytometric analysis demonstrated that some carcinoma tissues and cultured cancer cell lines expressed Foxp3.7-10 Immunohistochemistry using the antibody recognizing the N terminus, but not the C terminus, of Foxp3-highlighted cholangiocarcinoma tissue in 39% of cases as well as Treg cell morphology, suggesting the presence of the splicing variant of Foxp3 in cholangiocarcinoma cells. Molecular analysis using a cholangiocarcinoma cell line demonstrated that the cells expressed mRNA of Foxp3, but lack Exon 3. This type of splicing variant has already been reported in a melanoma cell line and created a novel amino acid caused by a frame shift at the C terminus.9 This is why the antibody recognizing the C terminus of Foxp3 could not detect the variant of Foxp3 found in cholangiocarcinoma tissue. Although a functional analysis of this variant as a transcription factor is necessary, it has already been reported that Foxp3 expression is closely correlated with the expression of IL-10 in all Foxp3-positive cell lines.10 The present study, using a cholangiocarcinoma cell line, also demonstrated that cells express mRNA of IL-10 as well as Foxp3. Moreover, the IL-10 protein was detected in the culture medium by ELISA at a concentration of 7.8-15.6 pg/mL, suggesting that the production of IL-10 was preserved with this splicing variant. This finding suggests that cholangiocarcinoma cells themselves function in immunosuppression similar to Treg cells via IL-10 production. This was supported by the present data that in Foxp3-positive cases, the number of IgG4-positive cells infiltrating cholangiocarcinoma tissues was higher than that in Foxp3-negative cases, though several negative cases still accompanied a significant IgG4 reaction (≥10 IgG4+ cells/HPF).
In this study, we demonstrated two different types of IgG4 reactions in cholangiocarcinoma tissues. Although statistical significance could be obtained in terms of cholangiocarcinoma as both nonprofessional APCs and IL-10–producing regulatory cells, some cases deviated from each mechanism. Therefore, as shown in Fig. 7, we divided all cases into a non–IL-10–inducing group and an IL-10–inducing group and re-evaluated the present results accordingly. The former (n = 24) consisted of MHC-II–negative and Foxp3-negative cases and MHC-II–positive, costimulatory molecule (CD86)-positive, and Foxp3-negative cases; the latter (n = 30) included MHC-II–positive, costimulatory molecule–negative, and Foxp3-positive cases. This combined analysis demonstrated that all but two cases in the non–IL-10–inducing group were poor in IgG4 (<10 IgG4+ cells/HPF) and that the difference in IgG4 reactions between the IL-10–inducing group and the non–IL-10–inducing group was significant compared with that of the individual analysis in terms of nonprofessional APCs and IL-10–producing regulatory cells. This finding indicates that cholangiocarcinoma cells directly participate in the induction of IgG4 reactions via an IL-10–predominent cytokine milieu as nonprofessional APCs and/or regulatory cells. However, the presence of IgG4-rich cases belonging to the non–IL-10–inducing group suggests another possible mechanism inducing IgG4 reactions in cholangiocarcinomas. Further studies are needed to clarify the mechanism of IgG4 reactions.
In conclusion, the marked infiltration of IgG-positive cells is found in several cases of cholangiocarcinoma, indicating that we should consider the differentiation of IgG4-related diseases and cholangiocarcinoma. The IgG4 reactions in cholangiocarcinomas, moreover, are closely associated with the IL-10–predominant regulatory cytokine milieu caused by cancer cells themselves directly and indirectly. Because IL-10 plays a primary role in suppressing immune responses, IgG4 reactions in cholangiocarcinoma might reflect evasion from immunosurveillance.
- 19Value of serum IgG4 in the diagnosis of autoimmune pancreatitis and in distinguishing it from pancreatic cancer. Am J Gastroenterol 2007; 102: 1646-1653., , , , , , et al.Direct Link: