Methionine aminopeptidase 2 over-expressed in cholangiocarcinoma: Potential for drug target

Abstract

Methionine aminopeptidases (MetAP) are proteases which remove the N-terminal methionine from newly synthesized proteins. Associations of MetAP2 with tumor progression of different cancers have been repeatedly reported. We aim to determine if MetAP2 is expressed in cholangiocarcinomas (CCA) and investigate to see if it would be a useful therapeutic target. We evaluated MetAP2 expression by immunohistochemistry in 82 patients of intrahepatic CCA. MetAP2 was expressed in bile ducts to various degrees. It was occasionally expressed with weak staining in normal bile duct epithelium but was strikingly over-expressed in dysplastic bile duct epithelia, primary and metastatic CCA tissues (p < 0.001). The increased expression of MetAP2 in proliferating bile duct was evident. All metastatic tumors had stronger expression of MetAP2 than the corresponding primary tumors. Fumagillin, a MetAP2 specific inhibitor, significantly inhibited cell proliferation in dose dependent manner and the degree of growth inhibition was dependent on the amount of cellular enzyme. The present study highlights the involvement of MetAP2 in an early event of carcinogenesis of CCA. The findings represent the first description of increased MetAP2 expression in CCA. The inhibition of enzyme activity using MetAP2 inhibitors may be a potential strategy for long-term control of tumor development and progression in CCA patients.

Cholangiocarcinoma (CCA) is an aggressive and lethal cancer arising from biliary epithelium either within the intrahepatic or extrahepatic biliary tract. Early diagnosis of CCA is difficult and the majority of patients with CCA present with advanced incurable disease and are not good candidates for curative surgery. In general, the prognosis for patients with advanced CCA at any site is dismal. Even in those that have undergone complete surgical resection, the recurrence rate remains quite high and consequently the 5-year survival rates range only from 0–40% [1], [2]. Unfortunately, conventional adjuvant treatments such as radiation therapy and chemotherapy in various combinations have not improved long-term survival after resection. The majority of CCA patients still die of metastatic cancer recurrence. Thus novel treatment strategies directed against this malignancy are being pursued aggressively. Although CCA is relatively infrequent compared to hepatoma, attention to CCA is now growing because both incidence and mortality rates of CCA are increasing in the United States, the United Kingdom, Australia and overall worldwide [3], [4].

Methionine aminopeptidases (MetAP) are metallopeptidases that selectively catalyze the removal of the N-terminal methionine from newly synthesized proteins, a central step in protein maturation [5]. The process is essential for further amino terminal modifications (e.g. acetylation by N-α acetyltransferase and myristoylation of glycine by N-myristoyltransferase). The structural alterations resulting from these modifications is the most often occurring protein modification and is essential in the regulation of a number of cellular processes such as cell proliferation, protein turnover, and protein targeting.

In eukaryotes, two main MetAP isoforms, MetAP1 and MetAP2, have been identified [6]. MetAP activity is essential for cellular growth and viability. It has been shown in yeast that knockout either MetAP1 or MetAP2 causes a decrease in growth rates while elimination of both genes is lethal [7]. The identification of MetAP2 as the cellular target of angiostatic agents, fumagillin and ovalicin, and the significant growth inhibition observed in cells sensitive to MetAP2 inhibition, have attracted more attention of MetAP2 than MetAP1 [8]. Fumagillin and its analogs have been shown to selectively and covalently bind MetAP2 and block its aminopeptidase activity [9], [10].

As with all solid tumors, CCA depends on the process of neoangiogenesis, the formation of tumor blood vessels, for both local and metastatic growth. Inhibition of neoangiogenesis is a new and attractive target for tumor therapy, since it theoretically offers the hope of long-term control of tumor progression. The clinical trials of using MetAP2 inhibitor, fumagillin and its analog, TNP-470, to suppress tumor metastasis and angiogenesis have been demonstrated [11]. Here we show for the first time that MetAP2 is aberrantly expressed during development of CCA and inactivation of MetAP2 activity resulted in reduction of proliferation of CCA cells. This finding suggests that inhibition of MetAP2 by fumagillin or its analogs may be a potential strategy for long-term control of tumor development and progression in CCA patients.

Materials and methods

Materials

The paraffin embedded liver tissues of 82 CCA patients including 29 invasive intraductal papillary CCA and 53 mass forming CCA were obtained from the specimen bank of the Liver Fluke and Cholangiocarcinoma Research Center, Faculty of Medicine, Khon Kaen University, Thailand. Informed consent was obtained from each subject before surgery and the Human Research Ethics Committee, Khon Kaen University approved the research protocol (#HE471214). Normal bile duct epithelia were examined from non-cancerous portions of CCA liver tissues. The hyperplastic/dysplastic biliary epithelia were examined from liver tissues of primary CCA.

The age, gender, tumor location, histological grading and pTNM stage [12] were evaluated by reviewing the medical charts and pathologic records. Tumor size was evaluated using the greatest perpendicular diameter of each liver lesion from 37 cases of mass forming CCA.

Methods

Cell lines and culture condition

CCA cell lines were established as previously described [13]. Four distinct CCA cell lines with different phenotype: M213A, M055 and M214, and KKU100 which were established from primary tumors with well differentiated, moderately differentiated and poorly differentiated CCA, respectively, were used in proliferation study. All CCA cells were maintained in Ham F-12 supplemented with 10% heat-inactivated fetal calf serum, 1% L-glutamine, 1% penicillin-streptomycin at 37°C and 5% CO₂.

Cell proliferation assay

Cells (5×10³ cells/100 µl) prepared as described above were seeded into a 96-well plate and grown for 24 h. On the next day, the medium was changed and cells were grown in the presence of fumagillin (0.25 or 5 µg/ml) or vehicle (0.25% ethanol) for 72 h. The number of cells was determined using sulforhodamine (SRB) assay [14]. Briefly, the culture medium was removed and 10% cold trichloroacetic acid was added for 30 min at 4°C and subsequently washed five times with deionized water. Then the plates were air-dried and 0.4% SRB in 1% acetic acid was added for 30 min. Unbound dye was washed out with 1% acetic acid five times. After air-drying, SRB dye within cells were solubilized with 200 µl solution of 10 mM unbuffered Tris-base solution and plates were left on a plate shaker for at least 10 min. Absorbance was measured at 540 nm using a microplate reader (Tecan Austria GmbH, Salburg, Austria).

Immunohistochemistry

Surgical pathology specimens from CCA patients who had undergone resection for CCA were collected directly from the surgical operating room. Two tissues were selected from each case: one from the tumor and the other from normal appearing-tumor free area. Immunohistochemistry of 11 paired individual liver tissues from primary and metastatic CCA tissues from five lymph nodes, three intrahepatic nodules, two gallbladders and one omentum were performed on the same slide. All specimens were fixed in 10% neutral formalin buffer, embedded in paraffin, and cut into 4 µm thick sections. Tissue sections were deparaffinized and subjected to microwave treatment in 10 mM citrate buffer, pH 6. Endogenous peroxidase activity was blocked in a 0.6% hydrogenperoxide solution. Immunohistochemical staining was performed by an immunoperoxidase method using 1:100 rabbit anti-MetAP2 antibody (Zymed Laboratories Inc., San Francisco, CA) as primary antibody and 1:100 anti-rabbit IgG-peroxidase (Zymed Laboratories Inc.) as the secondary antibody. Section incubated with diluent only served as negative controls. The sections were reacted with 0.05% 3, 3′-diaminobenzindine tetrahydrochloride (DAB. Sigma Chemical Co., St Louis, MO) and 0.1% H₂O₂ in 50 mol/l Tris-HCl pH 7.8 and counterstained with hematoxylin.

Since MetAP2 was occasionally present in normal bile duct epithelium with weak staining and almost all tumor cells expressed MetAP2. The results of the MetAP2 staining were graded by the intensity of positively stained cells as follows: 0, negative; 1 + weak staining, 2 + moderate staining, and 3 + strong staining. A section of CCA tumor tissue with moderate staining was used as reference for grading in each run.

Electrophoresis and immunoblotting

Cells were extracted with lysis buffer (8 M Urea, 4% CHAPS containing complete protease inhibitor cocktail (Roche Molecular Biochemicals, Germany) and subjected to 12.5% SDS-polyacrylamide gel electrophoresis [15]. Protein expression of MetAP2 in CCA cells was determined by the immunoblotting method of Towbin et al. [16] and was probed with 1:1000 anti-MetAP2 antibody (Zymed Laboratories Inc., San Francisco, CA) or 1: 5000 anti-β actin antibody (Sigma-Aldrich, St. Louis, MO). The immunoreactive proteins were visualized by Western Lightning Chemiluminescence Reagents (PerkinElmer Ltd., Boston, MA). Quantitative analysis of MetAP2/β-actin expression was performed using ImageMaster Platinum 5.0 software.

Statistical analysis

Statistical analyses for comparisons between clinico-pathologic findings were performed using STATA software (Stata Corp., College Station, TX). Association among a variety of variables, including age, gender, tumor histology, tumor stage, tumor size and invasion, was evaluated using the χ² test for heterogeneity or the Fisher's exact test. A p-value < 0.05 was considered to be statistically significant.

Results

Eighty two samples of CCA tissues were studied, 53 of which were mass forming type. The median age of patients with CCA was 56 years (range, 32 to 73 years); 60 were male, and 22 were female (male: female = 2.7:1). More than 73% of the patients were in stage IVB with tumor invasion. Sixty patients (73%) were determined for liver fluke (Opisthorchis viverrini) infection. Of these, 24 patients (40%) were positive detection of Opisthorchis viverrini eggs in bile or faeces.

MetAP2 expression in non-neoplastic intrahepatic bile ducts

Protein expression of MetAP2 was evaluated with immunohistochemistry in large and small bile ducts from 59, 52 and 82 liver specimens with nonneoplasia, dysplasia and CCA, respectively. Fifty percent of the normal bile duct examined was negative for MetAP2 and the left was expressed with weak staining. The site of MetAP2 expression was diffuse in the cytoplasm of positive cells. The hyperplastic/dysplastic (62%) and primary tumor (73%) of biliary epithelial cells were uniformly expressed with moderate to strong intensity (Figure 1B, C). The expression level of MetAP2 was significantly increased in dysplastic and tumor bile duct than in normal bile duct (p < 0.001) (Table I).

Figure 1.  Immunohistochemical staining patterns of MetAP2 in human bile duct epithelium. A: normal biliary cells with negative staining, 40×magnification; B: hyperplastic and dysplastic bile duct epithelia with moderate staining, 20×magnification; C: well differentiated tubular CCA, 20×magnification and D: lymph node with metastatic CCA, with strongly positive staining, 10×magnification.

Table I.  Pathological features of CCA patients and expressions of MetAP2 in primary tumor tissues.

	MetAP2 expression level
	n	negative-weak	Moderate-strong	p
Age (yrs)	82
•  ≤ 56	41	11	30	1.000
•  > 56	41	11	30
Gender	82
• Male	60	18	42	0.285
• Female	22	4	18
Macroscopic appearance	82
• Mass forming	53	14	39	0.909
• Invasive Intraductal papillary	29	8	21
Tumor stage	68
• I–III	7	3	4	0.757
• IVA	11	3	8
• IVB	50	15	35
Tumor grade	82
• Invasive Intraductal papillary	29	8	21	0.339
• Well-differentiated	32	6	26
• Moderately differentiated	13	4	9
• Poorly differentiated	8	4	4
Vascular invasion	69
• Present	42	12	30	0.925
• Absent	27	8	19
Lymphatic invasion	69
• Present	52	16	36	0.568
• Absent	17	4	13
Neural invasion	69
• Present	27	11	16	0.084
• Absent	42	9	33
Tumor size	37
•  ≤ 5 cm	7	2	5	0.809
•  > 5 cm	30	10	20

There was no difference in expression patterns of MetAP2 between mass forming and invasive intraductal papillary types CCA.

MetAP2 expression in CCA and clinico-pathological findings

Similar to the dysplastic and CCA bile duct epithelia, MetAP2 was constantly expressed in metastatic tumors with moderate to high intensity (Figure 1D). However, higher expression level of MetAP2 in metastatic tumors than primary CCA tissues was obviously observed (Table II). In 11 individual matched pairs of primary and metastatic CCA tissues, the metastatic cells showed a similar or higher profile of MetAP2 expression than the primary tumor of individual cases (data not shown). Comparing to 24% of primary tumor, 90% of metastatic CCA cells had high expression level of MetAP2 (p < 0.001).

Table II.  Comparison of MetAP2 expressions in bile duct epithelia.

	MetAP2 expression level
	n	0	1 +	2 +	3 +	P
Invasive intraductal papillary CCA
Non-dysplastic bile ducts	23	8	15	0	0
Hyperplasia/dysplasia	19	1	2	11	5	0.001^a
Primary CCA	29	3	5	13	8	0.001^a
Metastatic CCA	4	0	0	0	4
Mass forming CCA
Non-dysplastic bile ducts	36	19	13	4	0
Hyperplasia/dysplasia	33	7	5	15	6	0.001^a
Primary CCA	53	7	7	27	12	0.001^a
Metastatic CCA	7	0	0	1	6

^a=significant different from the non-dysplastic bile duct.

We studied relations between pathological and clinical features (tumor grade, tumor stage, tumor size and tumor invasion), and MetAP2 expression level in mass forming and invasive intraductal papillary CCA. There is no significant association of MetAP2 expression and the pathological findings (Table I). Cumulative survival was compared in the patients with primary CCA – those who had a low (0 and 1 + ) and high (2+ and 3 + ) MetAP2 expression. The median survival time of these two groups is not significant different (data not shown).

MetAP2 activity mediated cell growth

Inhibitory effect of fumagillin on cell growth was demonstrated in four distinct CCA cell lines in a dose dependent manner (Figure 2A). Fumagillin (5 µg/ml) could decrease proliferation of only approximately 40% in CCA cell lines M213A and KKU100. However, similar concentration of fumagillin could inhibit growth of CCA cell lines M055 and M214 to 80%. Quantification of protein expression of MetAP2 in each cell line using immunoblotting with β-actin as internal standard, revealed that MetAP2 expressions of cell lines M055 and M214 were lower than those of KKU100 and M213A (Figure 2B). CCA cell lines responded differently to fumagillin treatment according to its MetAP2 appearance. Cells with high MetAP2 expression showed lower response to fumagillin than those with low MetAP2 expression.

Figure 2.  Effect of Fumagillin on cell proliferation in CCA cells. A: Various CCA cell lines were treated 0.25 and 5 µg/ml of fumagillin for 72 h. Number of cells was determined using SRB assay as described in Materials and Methods. One hundred percent corresponded to the optical density of the control. Results are mean±SD from triplicate assays; B: Expression of MetAP2 of each CCA cell lines were determined by immunoblotting with anti-MetAP2, β-actin was used as internal standard. The numbers represent the quantitative analysis of MetAP2/β-actin.

Discussion

Accumulated reports on the antiangiogenic activity of TNP-470, fumagillin and the structurally related natural product ovalicin, have called attention to the use of MetAP2 as a molecular target of the antiangiogenic agents [17], [18]. These compounds irreversibly inhibit MetAP2 catalytic activity responsible for the removal of N-terminal methionine [9], [10], [19], [20]. Down-regulation of human MetAP2 expression by an anti-sense oligonucleotide or small interfering RNA (siRNA) significantly inhibited endothelial cell proliferation [21], [22]. In the current study we asked whether MetAP2 associates with CCA development and inactivation of MetAP2 activity can suppress proliferation of CCA cell lines.

We showed here for the first time that dysplastic biliary epithelia, intrahepatic CCA and metastatic CCA tissues frequently and strongly expressed MetAP2. In the normal intrahepatic biliary tree, MetAP2 is occasionally expressed; therefore, the frequent MetAP2 expression in CCA seems to be categorized as an aberrant expression.

It is apparent that CCA develops in the biliary tract as the result of a multistep process that seems to follow a sequence that includes early hyperplastic and metaplastic changes, followed by biliary dysplastic lesions, and leading eventually to the development of cancer [23]. The frequent and aberrant MetAP2 expressions in dysplastic biliary epithelia and CCA suggest that MetAP2 is up-regulated as part of the molecular events that take place during the malignant formation of CCA. The biliary epithelial cells might have acquired MetAP2 through their dysplasia-carcinoma sequence. We also observed a small population of non-dysplastic biliary epithelial cells with positive MetAP2 expression. These cells may represent predysplastic features and may be used as a novel biomarker for the early detection of CCA using the immunohistochemical approach.

The intrahepatic CCA can be classified according to the apparent lesion in the liver and clinical behaviors into mass-forming, periductal-infiltrating and intraductal growth types [24]. MetAP2 was frequently and strongly expressed in both types of CCA. This observation suggests that up-regulation of MetAP2 may be a common feature of CCA development. In metastatic CCA, MetAP2 is also constantly and strongly expressed. The frequent expression of MetAP2 both in primary and metastatic CCA identified here indicates the requirement of MetAP2 in carcinogenesis and growth of CCA. High levels of MetAP2 expression were also observed in germinal center B cells of malignant lymphomas of various subtypes [25].

MetAP2 is a bifunctional protein that plays a critical role in the regulation of post-translational processing and protein synthesis. In this study, the post-translational processing activity of MetAP2 in cell proliferation was further investigated. Inactivation of MetAP2 activity by fumagillin significantly suppressed proliferation of CCA cells. Cells with low expression of MetAP2 (M055 and M214) were potently inhibited following fumagillin treatment as compared to those with high expression of MetAP2 (M213A and KKU100). The differences in the observed growth inhibitory response between CCA cell lines may be based on cell-type specific or the amount of total cellular MetAP2. It appears that the degree of growth inhibition is based on the remaining amount of free MetAP2 in cell which has not been inactivated by MetAP2 enzyme inhibitor and is below the necessary cellular threshold to support cell proliferation [22]. Currently, the critical molecular regulator of MetAP2 in cell proliferation has been confirmed [22].

The mechanism by which inhibition of MetAP2 might lead to growth inhibition may be through both tumor cell intrinsic and extrinsic mechanisms. A defect in removal of N-terminal methionine caused by MetAP2 inhibition might lead to aberrant levels of proteins important for cell proliferation and apoptosis [26], [27]. Non-proper processing of the N-terminal methionine residue by MetAP results in difference of the first N-terminal residue which may significantly alter the function or binding affinities of the molecules. Decreasing activity of glutamine-fructose-6-phosphate aminotransferase [28]; and reduction in binding of interleukin-1beta to its receptor [29] by dysfunction of MetAP2 are evident.

TNP-470, a derivative of fumagillin, has been shown to be safe and effective in the treatment of solid tumors and arthritis in several animal studies and preclinical trials [30], [31]. It has been used in several clinical trials for the treatment of AIDs-related Kaposi's sarcoma, androgen-independent prostate cancer, metastatic breast cancer, advanced squamous cell cancer of the cervix, metastatic renal carcinoma, pediatric solid tumors, lymphomas and acute leukemia [32–35]. Apart from TNP-470 and its precursor, fumagillin, four highly potent MetAP2 inhibitors, IDR-803, IDR-804, IDR-805, and CKD-732 were recently designed by structure-based molecular modeling and their anticancer effects were examined in hepatoma cell line [36]. These compounds potently exert an anti-angiogenic effect which inhibited the growth of cancers in vivo and could potentially be useful for the treatment of a variety of cancers.

Molecular alterations associated with CCA have been made progressively over the past decade. Dysregulation of cell growth and survival pathways, aberrant gene expression, invasion and recently metastasis, and tumor microenvironment have been described to relate to human cholangiocarcinogenesis [23], [37–42]. Aberrant expression of MetAP2 in hyperplastic-dysplastic biliary cells as reported in the present study may be one of the mechanisms that dysregulates cell growth and survival of bile duct epithelia.

Recently, genome wide analysis of gene expression of human intrahepatic CCA bile duct epithelium [37] and expressed sequence tags of CCA cell lines [38] have been reported. Using different approaches, a few genes were consistently found and matched in these two published data sets. MetAP2 was not listed in the data of upregulated genes in both studies. However, not only MetAP2 but also ERBB2 and cyclooxygenase-2 which have been cited for involvement in cholangiocarcinogenesis [39], did not appear in these lists. This may due to high stringent of the selection criteria (a fivefold increase in more than 50% of the tumor examined) used in the study [37] and the level of MetAP2 expression may not high enough to pass this cutoff filter.

Based on the fact that CCA is a high infiltrating growth cancer which exhibits lymphatic and intrahepatic metastases, together with the fact that this lethal cancer is being encountered more frequently worldwide, there is now an urgent need to focus on developing specific chemopreventive and therapeutic strategies aimed at exploiting select molecular targets aberrantly expressed during carcinogenesis and metastasis of CCA. The frequent and aberrant expression of MetAP2 in biliary epithelial dysplasia as well as in primary and metastatic CCA, suggests it may be required in the genesis and metastasis of biliary tract cancer. The higher concentration of MetAP2 in CCA than the corresponding normal cells suggests a greater dependence on this enzyme by malignant cells for their function and proliferation. Hence, reduction in the enzyme activity may be more harmful to cancer than normal cells. Aberrantly high expression of MetAP2 in primary and metastatic tissues of CCA demonstrated in this study suggests the potential use of inhibitors to MetAP2 as an adjunct between cycles of conventional cytotoxic therapy. Evaluation of specific inhibitors to MetAP2 in CCA cell lines and/or animal models should provide justification for future selection and evaluation of MetAP2 inhibitors in clinical trials for high-risk CCA patients.

This project was partially supported by Research Grant of the Liver Fluke and Cholangiocarcinoma Research Center, Faculty of Medicine, Khon Kaen University (FY 2006) , and JSPS Asian and African Science Platform Program (Khon Kaen University and Kumamoto University) and the Royal Golden Jubilee Ph.D. Program (Grant No. PHD/0211/2543 to Sawanyawisuth K. and Wongkham S.). The authors would like to thank Dr. Banchob Sripa, Department of Pathology and the Liver Fluke and Cholangiocarcinoma Research Center, Faculty of Medicine, Khon Kaen University, for providing the CCA cell lines in this study.