Interleukin-22 promotes human hepatocellular carcinoma by activation of STAT3§

Authors

  • Runqiu Jiang,

    1. Liver Transplantation Center, First Affiliated Hospital of Nanjing Medical University, Nanjing, Jiangsu Province, P.R. China
    Search for more papers by this author
  • Zhongming Tan,

    1. Liver Transplantation Center, First Affiliated Hospital of Nanjing Medical University, Nanjing, Jiangsu Province, P.R. China
    Search for more papers by this author
  • Lei Deng,

    1. Liver Transplantation Center, First Affiliated Hospital of Nanjing Medical University, Nanjing, Jiangsu Province, P.R. China
    Search for more papers by this author
  • Yun Chen,

    1. Liver Transplantation Center, First Affiliated Hospital of Nanjing Medical University, Nanjing, Jiangsu Province, P.R. China
    Search for more papers by this author
  • Yongxiang Xia,

    1. Liver Transplantation Center, First Affiliated Hospital of Nanjing Medical University, Nanjing, Jiangsu Province, P.R. China
    Search for more papers by this author
  • Yun Gao,

    1. Liver Transplantation Center, First Affiliated Hospital of Nanjing Medical University, Nanjing, Jiangsu Province, P.R. China
    Search for more papers by this author
  • Xuehao Wang,

    1. Liver Transplantation Center, First Affiliated Hospital of Nanjing Medical University, Nanjing, Jiangsu Province, P.R. China
    Search for more papers by this author
  • Beicheng Sun

    Corresponding author
    1. Liver Transplantation Center, First Affiliated Hospital of Nanjing Medical University, Nanjing, Jiangsu Province, P.R. China
    • Liver Transplantation Center, First Affiliated Hospital of Nanjing Medical University, 300 Guangzhou Road, Nanjing, Jiangsu Province, P.R. China
    Search for more papers by this author
    • Fax: 86-25-86560946


  • Potential conflict of interest: Nothing to report.

  • Supported by grants from the National Natural Science Foundation of China (81072029 and 91029721 to B.S.) and the New Century Excellent Talents in University, Ministry of Education of China (NCET-09-0160 to B.S.). The work was also supported in part by the Program for Development of Innovative Research Teams in the First Affiliated Hospital of NJMU and the Priority Academic Program of Jiangsu Higher Education Institutions.

  • §

    These authors contributed equally to this study.

Abstract

Interleukin-22 (IL-22), one of the cytokines secreted by T helper 17 (Th17) cells, was recently reported to be a novel inflammation driver through STAT3 signaling activation. We aimed to investigate the role of IL-22 expression in hepatocellular carcinoma (HCC). We demonstrated significant up-regulation of IL-22 in human HCC tumor infiltrated leukocytes (TILs) compared to peripheral lymphocytes. Moreover, IL-22 expression was significantly higher in Edmondson Grade III-IV HCC patients versus Grade I-II, confirmed by both real-time polymerase chain reaction and immunohistochemistry. Both IL-22 receptor α and IL-23 were highly expressed in HCC and adjacent cirrhotic tissues compared to normal controls. Enhanced tumor growth and metastasis was found in mice that underwent subrenal transplantation of MHCC-97H cells cotransplanted with IL-22+ TILs cells. STAT3 phosphorylation and up-regulation of downstream genes Bcl-2, Bcl-XL, CyclinD1, and vascular endothelial growth factor (VEGF) promoted tumor growth and metastasis. In vitro studies confirmed the tumor-promoting and antiapoptotic effect of IL-22, as well as IL-6. In the mouse chronic hepatitis and HCC model, sustained and increased IL-22 expression and STAT3 activation were found in liver tissues. A linear correlation was demonstrated between IL-22 expression and hepatic complementary proliferation. An in vivo diethyl-nitrosamine-induced mouse HCC model verified that tumor formation was significantly decreased in IL-22 knockout mice. Conclusion: Excessive IL-22 can be found in the HCC microenvironment, leading to tumor growth, inhibition of apoptosis, and promotion of metastasis due to STAT3 activation. (HEPATOLOGY 2011;)

Hepatocellular carcinoma (HCC) is the fifth most common malignancy worldwide, with a 5-year survival rate of 9%.1 It is the final outcome of nonresolvable chronic hepatitis and is a major concern in clinical epidemiologic research. Tumor and inflammation microenvironments are considered the major battlefield between tumor promotion and antitumor immunity; malignant transformation occurs when tumor promotion dominates the microenvironment. Therefore, the determinants of pro- or antitumor effects in the microenvironment have become central in cancer research.

Tumor microenvironments that have a growth stimulation effect in malignancy include a large population of tumor cells, malignant transforming cells, immune cells, macrophages, etc. Moreover, much of the growth stimulating crosstalk between immune and malignant cells is mediated by cytokines that activate the oncogenic transcription factor STAT3,2 a major intrinsic activator in cancer inflammation3, 4 and a regulator of the tumor microenvironment. STAT3 modulates many immunosuppressive and tumor-promoting genes by targeting myeloid derived suppressor cells (mDSCs) and tumor-associated macrophages.5-8 STAT3 also mediates T-regulatory cell expansion in tumors and is necessary for Th17 cell development,9, 10 which promotes tumor growth.2, 11, 12 Moreover, cytokines, growth factors, and angiogenic factors induced by STAT3 still activate it as a freeforward loop, one of the important determinants in tumor-promoting inflammation and suppression of antitumor immunity in the face of persistent stimulation.4, 8, 13

A STAT3 activator such as interleukin-6 (IL-6) and its related signaling pathway play a central role in promoting HCC development through sustained STAT3 activation.14, 15 Kupffer cells persistently activate STAT3 signaling in tumor cells through IL-6 secretion. However, the activation of signaling pathways by way of nuclear factor kappa B (NF-κB) and STAT3 should be distinguished from stimulation by tumor and immune cells, such as T helper 17 (Th17). Th17 cells are a novel subset of T helper cells that markedly express RORγT and secrete cytokines, including IL-17A, IL-17F, IL-21, and IL-22. Th17 cells can develop as a result of cytokine release, including IL-1β, IL-6, IL-23, tumor necrosis factor alpha (TNF-α), and transforming growth factor beta (TGF-β).16-18 These cytokines are often present in the HCC microenvironment and are secreted mainly by Kupffer cells.19 Recently, Th17 cells have been shown to be activated by Kupffer cells,20 and are significantly associated with reduced survival and increased invasiveness in HCC.21 Therefore, another cellular freeforward loop exists in the HCC microenvironment, which is composed of tumor cells, Kupffer cells, and immune cells. This loop is central to persistent activation of STAT3, promoting tumor growth through secreting cytokines that create another loop, also by persistent activation of STAT3.

As one of the cytokines involved in the freeforward loop and secreted by Th17 cells, IL-22 also activates STAT3. Early in 2002, IL-22 was reported to be an activator for multiple signaling pathways such as JAK/STAT, ERK, JNK, and p38 MARPK in rat hepatoma cells.22 However, only a few reports of its role in tumors were published. Zhang et al.23 indicated that its antiapoptotic effect was enhanced through IL-22 in an autocrine manner in lung cancer, and knockdown research by small interfering RNA (siRNA) showed its therapeutic effect in a xenograft model. Another two studies focusing on ALK+ anaplastic large cell lymphoma revealed IL-22/IL-22R1 signaling to be an inflammation driver and tumor promoter.24, 25 The role of IL-22 in acute liver injury was also investigated by Zenewicz et al.,26 who claimed that IL-22, not IL-17, protected hepatocytes by STAT3 activation during conA-induced acute liver injury. In that study, the author mentioned that IL-22 might show a reverse effect during chronic hepatitis; in other words, long-term STAT3 activation by IL-22 might promote tumor growth by targeting damaged hepatocytes and tumor cells, similar to the HCC promotion by IL-6. Therefore, we sought to investigate IL-22 expression and related molecules in human HCC to further determine whether IL-22 was a tumor promoter and how it exerted its influence.

Abbreviations

EGFP, enhanced green fluorescence protein; HCC, hepatocellular carcinoma; IL, interleukin; mDSC, myeloid-derived suppressor cell; PBMC, peripheral blood mononuclear cell; Th, T helper; TIL, tumor infiltrated leukocyte; VEGF, vascular endothelial growth factor.

Patients and Methods

Patients.

Tumor and corresponding adjacent tissues and sera were obtained from 109 patients with HCC at surgery (Table 1). Normal liver parenchymal tissues were obtained as controls from 48 Chinese patients with benign disease, such as hemangioma and hepaticolithiasis. Sera of healthy volunteers were obtained during routine physical examination. Study data were obtained from patients who presented between January 2009 and August 2010 at the First Affiliated Hospital of Nanjing Medical University (Nanjing, China). Informed consent for gene expression analysis was obtained from all patients prior to surgery and the study was approved by our Institutional Ethics Committee. Diagnosis and staging of liver cancer was according to the Edmondson Staging System for HCC. Adjacent tissue was considered to be within 1 cm of the tumor and was confirmed to be nontumor tissue on pathology. These tissues contained abnormal morphology, such as pseudolobuli, hepatic steatosis, and hepatic hydropsia.

Table 1. Clinical Characteristics of the 109 HCC patients
Patient CharacteristicsPatients
No. of patients109
Age, year (median, range)47, (21-76)
Gender (male/female)75/34
HbeAg (negative/positive)18/91
Cirrhosis (absent/present)12/97
ALT, U/L (≤45/>45)62/47
AFP, ng/ml (≤13.6/>13.6)23/86
Tumor size, cm (≤5/>5)43/66
Tumor multiplicity (solitary/multiple)82/27
Vascular invasion (absent/present)35/74
Edmondson grade (I+II/III+IV)63/46

Diethyl-Nitrosamine (DEN)-Induced HCC Mouse Model and Subrenal Capsule Cell Transplantation in Nude Mice.

All mouse procedures were performed as described in the Supporting Materials and Methods. Kidney capsule transplantation was performed as previously reported.27

Isolation and Culture of Human HCC Infiltrated Leukocytes.

Fresh tumor tissues were washed twice in RPMI 1640. Fatty, connective, and necrotic tissue was removed. Tissues were minced into 1-2 mm pieces in RPMI 1640, transferred into 15 or 50 mL conical tubes, and incubated with triple enzyme digestion medium containing DNase (30 U/mL), hyaluronidase (0.1 mg/mL), and collagenase (1 mg/mL) for 2 hours at room temperature with gentle shaking. Tissues were resuspended in 10 mL RPMI 1640 and filtered through a 70-μm cell strainer (BD Pharmingen). Tissue trapped by the strainer was placed into individual wells containing 1 mL of T-cell growth medium in a 24-well plate for isolation or detection by flow cytometry .

Enzyme-Linked Immunosorbent Assay (ELISA), Quantitative Real-Time Polymerase Chain Reaction (PCR), Western Blot, Immunohistochemistry (IHC), Flow Cytometry, Bromodeoxyuridine (BrdU) Cooperation, Immunofluorescence Assay, Peroxide-Induced Apoptosis and Flow Cytometry Analysis.

Detailed information is provided in the Supporting Materials and Methods.

Statistical Analysis.

Results are expressed as the mean ± standard deviation (SD). Comparisons between the two groups were performed using the Mann-Whitney U test. Correlations between parameters were ascertained by Pearson correlation and linear regression analysis, as appropriate. All statistical analyses were performed using SPSS statistical software v. 13.0 and two-tailed tests were applied to all data unless otherwise specified, considering a P-value of less than 0.05 (95% confidence interval [CI]) statistically significant.

Results

High Expression of IL-22 Investigated in Human HCC.

As the major cellular source of IL-22 is lymphocytes,28 we aimed to investigate the expression of IL-22 in human HCC; thus, HCC TILs were isolated from 109 HCC patients (63 Edmonson Grade I-II and 46 III-IV). TILs messenger RNAs (mRNAs) were extracted after isolation and their cDNA were analyzed by real-time PCR. IL-22 was significantly overexpressed in HCC TILs compared to human peripheral blood mononuclear cells (PBMCs) (P < 0.001 across all grades by Mann-Whitney U test; Fig. 1A). Serum IL-22 was significantly up-regulated both in chronic hepatitis and HCC patients compared with controls (0.523 ± 0.210 and 0.857 ± 0.296 versus 0.174 ± 0.181 ng/mL, P < 0.01; HCC versus CH, P < 0.05; Fig. 1B). High expression of IL-22 in HCC patients was also confirmed by flow cytometry (Fig. 1E). The distribution and location of IL-22 in human HCC tissues and adjacent cirrhotic tissues were also investigated by IHC. As reported in previous studies, IL-22 was located in nonparenchymal liver and IL-22-positive cells were distributed around cancer nests or fibrous tissues around pseudolobules in adjacent cirrhotic tissues (Fig. 1D1-D3). Almost no expression of IL-22 was detected in tumor cells. Further studies on IL-22 expression in different histologic tumor grades demonstrated that IL-22 expression was stronger in TILs of Grade III-IV versus Grade I-II tumors, both in mRNA transcription and protein secretion (Fig. 1A,C). Moreover, wider infiltration of IL-22 into the cancer nest was found in Grade III-IV compared with Grade I-II tissues (Fig. 1D1-D2). TILs involved in less differentiated and highly malignant HCC had stronger expression and wider distribution of IL-22, implying that IL-22 was associated with tumor growth and degree of malignancy.

Figure 1.

High expression of IL-22 was investigated in human HCC. (A) Expression of IL-22 in PBMCs (n = 48) and TILs respectively from Edmonson Grade I-II HCC patients (n = 63) and III-IV (n = 46), as detected by real-time PCR. (B) Detection of IL-22 in human sera of healthy controls (n = 30), patients with chronic hepatitis (n = 30) and patients of HCC (n = 30) by ELISA. (C) Average integrated optical density (IOD) was obtained by analyzing five fields for each slide evaluated by Image-Pro Plus software (v. 5.0) for IHC staining of IL-22. (D1-D4) Expression and distribution of IL-22 in normal liver tissues (n = 48) (D4), HCC Edmondson Grade I-II (n = 63) (D2) and III-IV (n = 46) (D1) tumor tissues (indicated by arrows) and adjacent tissues (D3), as analyzed by IHC (×100). (E) IL-22 expression in TILs from one of 30 patients by flow cytometry costained with antibody to CD3, CD4, CD8, and CD68. IL-22 expression in PBMCs costained with CD3 considered controls. **P < 0.01 and *P < 0.05.

IL-22 related molecules were also analyzed. IL-22 binding protein (IL-22BP), which served as one of the IL-22 receptors, was highly expressed in HCC and adjacent cirrhotic tissues compared to normal tissues (tumor versus normal, P < 0.001; tumor versus cirrhosis, P < 0.05; cirrhosis versus normal, P < 0.01) (Fig. 2A,C1,D1,D5,D9), which implied that the effect of IL-22 could extend to tumor, as well as cirrhotic tissues. As a major downstream effect of IL-22, phosphorylation of STAT3 signaling was evaluated. Phosphorylated STAT3 (S727 and Y705) were all detected by IHC. Phosphorylation of STAT3 was confirmed in human HCC tumor tissues and adjacent cirrhotic tissues (tumor versus normal, P < 0.001 for Y705 and S727; tumor versus cirrhosis, P < 0.05 for Y705 and S727; cirrhosis versus normal, P < 0.01 for Y705 and S727), which was identical to previous research (Fig. 2C3,C4,D3,D4,D7,D8,D11,D12). Finally, IL-23, a key cytokine stimulating IL-22 expression in Th17 cells, was detected by both real-time PCR and IHC. Thus, IL-23 was significantly overexpressed in tumor and adjacent cirrhotic tissues compared to normal liver (tumor versus normal, P < 0.001; tumor versus cirrhosis, P < 0.05; cirrhosis versus normal, P < 0.01) (Fig. 2B,C2,D2,D6,D10). Combined with other IL-22 expression stimuli, such as IL-1β, TNF-α, and IL-6 (which are highly expressed in human HCC),19 it could be concluded that HCC tumor microenvironments exist where high IL-22 expression is observed.

Figure 2.

(A,B) Expression of IL-22 BP and IL-23 in human HCC (n = 80), adjacent cirrhosis tissues (n = 80), and normal liver (n = 48), as detected by real-time PCR. Data represent the relative mRNA expression compared with GAPDH and are presented as the mean ± SD. (C1-C4) Average integrated optical density (IOD) obtained by analyzing five fields for each slide evaluated using Image-Pro Plus software (v. 5.0) for IHC staining of IL-22BP, IL-23, p-STAT3 (Y705), and p-STAT3 (S727). (D1-D4) Expression and distribution of IL-22BP, IL-23, p-STAT3 (Y705), and p-STAT3 (S727) in human normal liver tissues (n = 48), as analyzed by IHC (×100). (D5-D8) Expression and distribution of IL-22BP, IL-23, p-STAT3 (Y705), and p-STAT3 (S727) in tissues adjacent to human HCC (n = 109) as analyzed by IHC (×100). (D9-D12) Expression and distribution of IL-22BP, IL-23, p-STAT3 (Y705), and p-STAT3 (S727) in human HCC tissues (n = 109) as analyzed by IHC (×100). **P < 0.01, *P < 0.05.

To identify the phenotypic features of the cells expressing IL-22, IL-22 was costained with other cellular surface markers such as CD4, CD8, and CD68 in 30 cases of TILs isolated from human HCC. The results indicated that both CD4 and CD8-positive lymphocytes expressed IL-22 (8.15 ± 2.82% for CD4-positive cells and 5.68 ± 2.91% for CD8-positive cells). It was unexpected that a small population of Kupffer cells (CD68-positive cells, 4.12 ± 1.08%) also secreted IL-22. We isolated IL-22-expressing TILs for further functional studies. To avoid possible antitumor effects of CD8+ TILs, CD4+ and CD68+ TILs were isolated by flow cytometry. CD4+ lymphocytes were maintained in vitro after Th22 polarizing stimulation and were then mixed with CD68+ TILs. This cell mixture is denoted IL-22+ TILs.

IL-22 Promotes Tumor Growth and Metastasis In Vivo.

To investigate the effect of IL-22 on tumor growth and metastasis, an in vivo tumorigenesis assay was designed using the subrenal capsule cell transplantation model. MHCC-97H, an HCC cell line with high metastatic potential, was cotransplanted with IL-22+ TILs from two HCC patients. Cell suspensions consisting of a 1:1,000 ratio of cell mixture (IL-22+ TILs: 97H) were injected in the subrenal capsule of BALB/c nude mice. IL-22+ TILs from two HCC patients promoted tumor growth and metastasis. Significant increase of tumor volume was found when 97H cells were cotransplanted with IL-22+ TILs compared to 97H cells alone (TILs 1, 1.93 ± 0.30 cm3 versus 97H, 0.34 ± 0.19 cm3, P < 0.0001; TILs 2, 2.10 ± 0.43 cm3 versus 97H, P < 0.001, by unpaired t test; Fig. 3A1,A2). Tumor tissue was analyzed by IHC for IL-22 staining. IL-22+TILs from the two patients existed and proliferated in the subrenal capsule (Fig. 3B). Increased phosphorylation of STAT3 in tumor tissues were detected; up-regulation of CyclinD1 was also found as the effect of activation of STAT3 signaling, which can explain the enhanced tumor growth promoted by IL-22+ TILs. BCL-XL and BCL-2, two antiapoptosis genes, were also increased in the tumor tissues with two IL-22+ TILs, which were also downstream transcripts of STAT3 activation and might also explain tumor growth enhancement (Fig. 3D).

Figure 3.

IL-22 promotes tumor growth and metastasis in vivo. IL-22 (+)TILs were isolated from two HCC patients and cotransplanted into the subrenal capsule of nude mice with EGFP labeled MHCC97H cells. Tumor growth and metastasis was investigated. (A1) Tumor tissues obtained from nude mice from each group. (A2) Comparison of tumor volume of each group. (B) IL-22 positive cell distribution in tumor tissues of each group investigated by IHC. (C) Abdominal invasion and lung metastasis were investigated in each group respectively by an in vivo fluorescence imaging system and fluorescent microscopy (×100). (D) Western blot detection of expression of IL-22, p-STAT3 (Y705), p-STAT3 (S727), total STAT3, Bcl-XL, Bcl-2, CyclinD1, and VEGF, which all normalized to β-actin for every three samples in each group. **P < 0.01.

Metastasis was also investigated in the 97H cell by enhanced green fluorescence protein (EGFP) labeling and in vivo fluorescence imaging. Visible abdominal metastasis was detected in the group injected with one of the IL-22+ TILs group only (three out of total six mice), whereas no other abdominal invasion was visible when injected with another IL-22+ TILs or 97H cell alone. Fluorescent microscopy was used to detect lung metastasis on mice lung tablet slides. EGFP-labeled 97H could be detected, even in the 97H group, due to the nature of lung metastasis. However, the volume and distribution increased dramatically in the lung when IL-22+ TILs were combined, with the most severe metastasis in the TILs group with abdominal invasion. Green 97H cells were flaky and lumpish, and more 97H metastases were detected in each vision in mice lung (Fig. 3C). Based on this result, we analyzed VEGF expression in tumor tissues, which indicated that VEGF expression was increased with the effect of IL-22+ TILs (Fig. 3D); increased VEGF was also a downstream target of STAT3 activation.

IL-22 Enhances Tumor Proliferation and Antiapoptotic Ability by Activating STAT3 Signaling In Vitro.

To further investigate the mechanism of IL-22 on tumor growth, in vitro studies were designed (Fig. 4A). 97H cells were cocultured with TILs and IL-22+ TILs. In order to exclude the effect from IL-6, secreted by IL-22+ TILs, IL-22 and IL-6 were blocked. The IL-22 effect on tumor cell proliferation and antiapoptosis were investigated. The proliferation ability of each group was accessed by the BrdU cooperation assay. Both TILs and IL-22+ TILs significantly enhanced the proliferation ability of 97H cells (Fig. 4B1-B6). However, when IL-6 and IL-22 were blocked the percentage of dividing cells (BrdU-positive cells) decreased significantly (Fig. 4B8). Meanwhile, CyclinD1 expression also decreased, corresponding to fewer dividing cells (Fig. 4D). IL-22+ TILs and TILs also significantly enhanced the antiapoptosis ability of 97H cells, which decreased the percentage of apoptotic cells induced by peroxide compared to 97H cells. But when the effect of IL-6 or IL-22 was eliminated by their antibody, apoptosis significantly increased (Fig. 4C). To explain this, BCL-2 and BCL-XL, two antiapoptosis genes, were determined, which increased with the existence of TILs and IL-22+ TILs and decreased when IL-6 or IL-22 was blocked. Apparently, all these changes originated from activation or blocking of STAT3 signaling. IL-6 has already been proven to be essential in HCC promotion due to its capacity to activate STAT3 signaling and its downstream effects such as proliferation enhancement and inhibition of apoptosis. In this study we introduce another STAT3 activator, IL-22, with a similar tumor promotion mechanism. However, the role of IL-22 in liver injury and inflammation and malignant transformation to HCC needs further investigation.

Figure 4.

IL-22 enhances tumor proliferation and antiapoptosis ability by activating STAT3 signaling in vitro. (A) Schematic diagram of experimental design. (B1-B6) Immunofluorescence staining of BrdU in green and DAPI in blue, reflecting cell proliferation in each group detected by fluorescent microscopy (×200). (B7) Immunofluorescence staining with a nonspecific antibody control (NC) (×200). (B8) Comparison of cell proliferation in each group; the experiment was repeated in triplicate. (C1-C6) Analysis of apoptosis of MHCC97H cells induced by peroxide in each group by flow cytometry. (C7) Comparison of apoptosis in each group (experiments performed in triplicate). (D) Western blot detection of IL-22, p-STAT3 (Y705), p-STAT3 (S727), total STAT3, Bcl-XL, Bcl-2, and CyclinD1 expression, all normalized to β-actin. *P < 0.05; **P < 0.01.

IL-22 Promoted HCC Development in Mice.

To investigate the role of IL-22 during the entire process from liver injury to HCC, we used the DEN-induced mouse HCC model. Systemic treatment with DEN can induce acute inflammation, liver injury, and compensatory proliferation, which are observed during the development of human HCC. Dynamic IL-22 expression was investigated by IHC. We detected IL-22 expression 0.5 month after DEN administration, which gradually increased with the passage of time (Fig. 5A). An in vivo BrdU cooperation assay was performed, reflecting that liver complementary proliferation gradually enhanced over a 5-month period (Fig. 5B). Moreover, a linear correlation study between IL-22 expression and liver complementary proliferation was performed by counting positive cells per field of both IL-22 and BrdU (Fig. 5C). IL-22 expression significantly correlated with liver complementary proliferation (R2 = 0.898, P < 0.001). Meanwhile, STAT3 activation and its downstream cell proliferation related genes such as CyclinD1 were also investigated. All genes increased significantly 2 months after DEN injection until the fifth month, when tiny nodes were visible in the liver (Fig. 5E).

Figure 5.

IL-22 promoted HCC development from chronic hepatitis in mice. (A1-A6) Expression of IL-22 in mice chronic hepatitis tissues 0.5, 1, 2, 3, 4, and 5 minutes after DEN administration (×100). (B1-B6) Expression of BrdU in mice chronic hepatitis tissues 0.5, 1, 2, 3, 4, and 5 minutes after DEN administration (×100). (C) Linear correlation between expression of IL-22 and BrdU-positive cells in mice hepatitis tissues (R2 = 0.898, P < 0.0001). Number of positive cells of both indices was obtained by averaging positive cell number from five individual visions of each slide. (D1-D3) Average integrated optical density (IOD) was obtained by analyzing five individual visions for each slide by Image-Pro Plus software (v. 5.0) for IHC staining of IL-22, IL-22 BP, and IL-23 in mouse HCC (n = 10) and normal tissues (n = 10). (D4-D6) Expression of IL-22, IL-22 BP, and IL-23 in mouse HCC tissues as detected by IHC. (D7-D9) Expression of IL-22, IL-22 BP, and IL-23 in normal mouse liver tissues detected by IHC. (E) Western blot detection of p-STAT3 (Y705), p-STAT3 (S727), total STAT3, and CyclinD1, which were all normalized to β-actin over time after DEN administration. **P < 0.01.

We also investigated IL-22 expression and that of related genes in mouse HCC, which is similar to human HCC. IL-22 was highly expressed in liver tumors compared to normal livers (P < 0.01); IL-22-positive cells also accumulated around the tumor nest area (Fig. 5D1,D4,D7). IL-22Bp and IL-23 were also significantly up-regulated in mouse HCC tissues (P < 0.01 for IL-22Bp and P < 0.01 for IL-23), indicating that a similar tumor microenvironment existed in mouse HCC (Fig. 5D2-D3,D5-D6,D8-D9).

DEN-induced tumorigenesis was compared between IL-22 knockout and wildtype mice, which revealed that IL-22 deletion led to reduced tumorigenesis (Fig. 6A,B). No IL-22 expression was detected in knockout compared to wildtype mice (Fig. 6C). Thus, IL-22 significantly promotes the development of mouse HCC.

Figure 6.

(A) Comparison of tumor multiplicity in male wildtype mice (n = 10) and IL-22−/− mice (n = 10). (B) Liver from 8-month-old DEN-treated mice. Tumorigenesis was investigated in IL-22−/− mice. (C) Expression of IL-22 in murine HCC tissues analyzed by IHC. **P < 0.01.

In conclusion, IL-22 was highly expressed in chronic hepatitis and carcinogenesis, and appears to be essential in the process of malignant transformation from chronic hepatitis to HCC in the mouse model.

Discussion

IL-22 is a member of IL-10 cytokine family and was discovered in 2000.29 Different from other interleukins, IL-22 does not serve as a messenger between immune cells. The distribution of two IL-22 reporters, IL-22R1 (IL-22BP) and IL-10R2, suggested that the most important target cells of IL-22 are the skin, digestive tract (including pancreas and liver), lungs, and kidney, with the general effects of antimicrobial defense, regeneration, and protection against injury.30 So far, the most precisely characterized effects of IL-22 were obtained from keratinocytes. These results showed that three functional categories of genes were affected by IL-22: antimicrobial proteins, differentiation-associated proteins, and mobility-regulating proteins—these proteins are functionally related to dermis protection and keratinocyte alterations in psoriasis.30, 31 In lung tissue, IL-22 increased transepithelial resistance to injury by up-regulating the expression of lipocalin 2, and consequently resolving pulmonary Klebsiella pneumoniae infection.32 In C. rodentium intestinal infection, IL-22 was produced earlier than IL-17A and played a decisive role, whereas IL-17A did not.33 However, a pathogenic role of IL-22 was also found in a colitis mouse model, where massive IL-22 expression was found in the mesenteric lymph nodes and the inflamed intestine. This finding implied a potential systemic role of IL-22 in patients with Crohn's disease (CD); moreover, serum IL-22 in CD was highly elevated compared with healthy persons and correlated with disease severity.34

IL-22 possesses a directly protective role in acute liver injury. Radaeva et al.35 used the model of Con-A-induced hepatitis, which was associated with massive infiltration of activated T cells and IL-22 production. Zenewicz et al.26 confirmed the observation that IL-22 protected hepatocytes during acute liver inflammation using IL-22-deficient mice. Additionally, the authors demonstrated that the protective effect of IL-22 might be due to STAT3 activation and its antiapoptotic and regeneration promotion effect, and that IL-22 may also limit damage during chronic inflammation; recombinant IL-22 therapy may prevent development of HCCs.

However, IL-22 may have opposing short-term and long-term effects in the liver. Expression of IL-22 during chronic inflammation may allow survival of damaged hepatocytes that are precursors for HCCs, and may therefore promote cancer. Along the same lines as Zenewicz et al., we found excessive IL-22 secretion by TILs of HCCs, and up-regulated serum IL-22 in chronic hepatitis patients. Both results can explain a hypothesis of long-term and sustained STAT3 activation in a tumor and chronic hepatitis microenvironment. HCC Th22 cells may be the cellular source of IL-22, as reported.28 In natural killer (NK) cells, however, we did not expect to see that a small population of Kupffer cells expressed IL-22. Only Gu et al.36 showed that IL-10 null macrophage could express high levels of both IL-17 and IL-22. In addition, the amount of IL-22 varied in different degrees of HCC malignancy. IL-22 expression in Edmondson Stage III-IV HCC tissues was higher than Stage I-II, implying that IL-22 may contribute to stronger proliferation and a higher degree of malignancy. Prior studies focused on the association between phosphorylated STAT3 and degree of malignancy in glioma,37, 38 but the explicit mechanisms remain unknown.

The HCC microenvironment, composed of interacting tumor cells, Kupffer cells, immunocytes, and their cytokines, consist of a freeforward loop centralized by persistent activation of STAT3 to promote tumor growth. Consisting of IL-23 and IL-22 binding protein, an IL-22 freeforward loop existing in HCC or chronic hepatitis microenvironments has been shown. The IL-22 signal can be transmitted through a heterodimeric receptor complex consisting of IL-22R1 (IL-22 BP) and IL-10R2.39, 40 Unlike the ubiquitously expressed IL-10R2 chain, the IL-22BP chain was normally restricted to nonimmune cells such as epithelial cells and hepatocytes.41 Corresponding to IL-22 overexpression in Kupffer cells and immunocytes, IL-22 BP was also overexpressed in tumor cells that ensured the transmission IL-22 signal. Protumor cytokine IL-23, regulated by STAT3,7 was also overexpressed in human HCC. Excessive expression of IL-23 combined with TNF-α, IL-6, and IL-1β, which are already overexpressed in human HCC, comprise a milieu for the infiltration of naive lymphocytes also enrolled by STAT3 activation; these can differentiate into Th17 cells that express IL-22.2 We transplanted this IL-22 tumor microenvironment in the subrenal capsule of nude mice. The proliferation and metastasis promotion effect of this freeforward loop was confirmed: proliferation-associated CyclinD1, cell survival-associated Bcl-XL and Bcl-2, and VEGF-associated metastasis were all up-regulated, all mediated by STAT3 activation.2 A more precise and accurate mechanism was confirmed by in vitro studies, because IL-6 was well accepted as a STAT3 activator and had a similar effect to IL-22 in the HCC microenvironment. A similar tumor promotion effect was obtained whether IL-6 was present or not, implying that there were two cytokines with a similar effect on tumor promotion both in vivo and in vitro.

Although high expression of serum IL-22 in human chronic hepatitis was demonstrated, the role of IL-22 in the transformation of chronic hepatitis to HCC cannot be fully investigated due to legislative and ethical limitations. Therefore, we investigated a DEN-induced mice chronic hepatitis and tumorigenesis model. IL-22 gradually increased over a 5-month period after DEN administration, until tiny nodes were visible. Moreover, sustained IL-22 expression promoted cell proliferation by persistent activation of STAT3 and CyclinD1 enhancement. A strongly positive IL-22 area was found mostly around the central vein, meaning that IL-22 may originate from the blood. Taken together with the findings on reduced tumorigenesis in IL-22−/− mice, the results indicate that IL-22 may be another key promoter in mouse HCC development, similar to IL-6, which has already been confirmed as a essential protein in mouse HCC development.14, 15 Moreover, an analogous tumor microenvironment was also confirmed in mouse HCC.

This study demonstrated a previously unknown role of IL-22 in human chronic hepatitis and HCC, having a promotion effect in proliferation, cell survival, metastasis, and transformation from chronic hepatitis to HCC. Further studies are warranted to investigate the therapeutic potential of anti-IL-22 in the prevention and treatment of human HCC.

Ancillary