Antisense oligodeoxynucleotides directed against aspartyl (asparaginyl) β-hydroxylase suppress migration of cholangiocarcinoma cells

Journal of Hepatology 38 (2003) 615–622 www.elsevier.com/locate/jhep

Antisense oligodeoxynucleotides directed against aspartyl (asparaginyl) b-hydroxylase suppress migration of cholangiocarcinoma cells Takashi Maeda 1, Paul Sepe 1, Stephanie Lahousse 1, Seishu Tamaki 1, Munetomo Enjoji 2, Jack R. Wands 1, Suzanne M. de la Monte 1,* 1

Departments of Medicine and Pathology, Liver Research Center, Rhode Island Hospital, Brown Medical School, 55 Claverick Street, 4th Floor, Providence, RI 02903, USA 2 Department of Medicine and Bioregulatory Science, Graduate School of Medical Sciences, Kyushu University, 3-1-1 Maidashi, Higashiku, Fukuoka 812-8582, Japan

See Editorial, pages 671–673

Background: Aspartyl (asparaginyl) b-hydroxylase (AAH) is an a-ketoglutarate-dependent dioxygenase that hydroxylates aspartate and asparagine residues in EGF-like domains of proteins. The consensus sequence for AAH bhydroxylation occurs in signaling molecules such as Notch and Notch homologs, which have roles in cell migration. Aim: This study evaluated the potential role of AAH in cell migration using cholangiocarcinoma cell lines as models due to their tendency to widely infiltrate the liver. Methods: Five human cholangiocarcinoma cell lines established from human tumors were examined for AAH expression and motility. The effect of antisense oligodeoxynucleotide inhibition of AAH on cholangiocarcinoma cell migration was investigated. Results: Western blot analysis detected the ,86 kDa AAH protein in all five cholangiocarcinoma cell lines, and higher levels of AAH in cell lines derived from moderately or poorly differentiated compared with well-differentiated tumors. Immunocytochemical staining and fluorescence activated cell sorting analysis revealed both surface and intracellular AAH immunoreactivity. Using the phagokinetic non-directional migration assay and a novel ATPLite luminescencebased directional migration assay, we correlated AAH expression with motility. Correspondingly, antisense and not sense or mutated antisense AAH oligodeoxynucleotides significantly inhibited AAH expression and motility in cholangiocarcinoma cells. Conclusions: AAH over-expression may contribute to the infiltrative growth pattern of cholangiocarcinoma cells by promoting motility. q 2003 European Association for the Study of the Liver. Published by Elsevier Science B.V. All rights reserved. Keywords: Cell migration; Antisense oligodeoxynucleotide DNA

1. Introduction Cholangiocarcinomas are highly malignant epithelial neoplasms of intrahepatic biliary tract origin. The pathogenReceived 4 February 2002; received in revised form 19 December 2002; accepted 16 January 2003 * Corresponding author. Tel.: 11-401-444-7364; fax: 11-401-444-2939. E-mail address: [email protected] (S.M. de la Monte).

esis of these tumors is unknown, prognosis is exceedingly poor, and there are no effective treatments [1–3]. Despite attempts at surgical resections, liver transplantation, chemotherapy, and radiation therapy, survival among patients with advanced cholangiocarcinomas has not significantly changed over the past several decades [4,5]. Thus, new therapeutic modalities are needed. One novel approach would be to inhibit expression of genes that contribute to aggressive tumor behavior. For example, ras, myc, c-erb B-

0168-8278/03/$30.00 q 2003 European Association for the Study of the Liver. Published by Elsevier Science B.V. All rights reserved. doi:10.1016/S0 168-8278(03)00 052-7

616

T. Maeda et al. / Journal of Hepatology 38 (2003) 615–622

2, and c-met could be targeted since previous studies demonstrated increased expression of these proto-oncogenes in high percentages of cholangiocarcinomas [6–11]. Another potential anti-tumor target is interleukin-6, which may contribute to autocrine or paracrine growth of transformed cholangiocytes through activation of p38 and p44/ p42 MAPK signaling cascades [12]. Since down-regulation of the p27kip1 cell cycle inhibitor protein has also been linked to poor prognosis of intrahepatic cholangiocarcinoma [3], strategies to increase its expression may prove beneficial. Finally, recent evidence suggests that the administration of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) may inhibit growth of cholangiocarcinomas [13], although the practical utility of this approach awaits execution and completion of rigorous clinical trials. The major feature of intrahepatic cholangiocarcinomas that renders them aggressive is their highly infiltrative growth pattern. In this regard, lymphatic permeation, intrahepatic metastasis, and lymph node metastasis independently correlate with poor prognosis [1,3]. Therefore, the identification of genes that promote infiltrative and metastatic growth of cholangiocarcinomas could prove helpful for selecting therapeutic targets to inhibit intrahepatic and local tissue spread of these neoplasms. Recently, we characterized the human aspartyl (asparaginyl) b-hydroxylase (AAH) gene [14,15] and in preliminary studies demonstrated its potential role in regulating motility of malignant neoplastic cells [16,17]. AAH is an a-ketoglutarate-dependent dioxygenase that stereospecifically catalyzes post-translational hydroxylation of b carbons of aspartyl and asparaginyl residues present in EGF-like domains of certain proteins [18–21]. The consensus sequence for AAH hydroxylation is present in Notch proteins, which have known roles in cell differentiation during development, and in mammalian Notch homologs, which are known to be oncogenic [22,23]. Previously, we reported over-expression of AAH in 20 of 20 primary human cholangiocarcinomas and undetectable AAH expression in normal, regenerating, and benign proliferating bile duct epithelial cells [15]. Moreover, we obtained experimental evidence that AAH over-expression is linked to cellular transformation [14] and in vivo evidence of an association between high levels of AAH and proneness toward infiltrative or metastatic growth of malignant neoplastic cells [15]. These observations led to the hypothesis that AAH has a role in regulating invasive or metastatic tumor cell growth. The present study demonstrates that high levels of AAH correlate with increased cell motility, and that antisense oligodeoxynucleotide (ODN) inhibition of AAH expression significantly reduces motility of cholangiocarcinoma cells. 2. Materials and methods 2.1. Cell lines Five human cholangiocarcinoma cell lines, SSP-25, NEC, ETK-1, RBE, and H-1 [24–26], originating from intrahepatic cholangiocarcinomas, were

kindly provided by Dr. Enjoji of Kyushu University in Japan. The cells were maintained in RPMI 1640 supplemented with 10% fetal calf serum, 10 mM non-essential amino acids, and 2 mM l-glutamine.

2.2. Protein studies Western blot analysis of cell lysates prepared in radioimmunoprecipitation assay buffer was performed using 60 mg protein samples as described [15,27]. Cells were fixed in Histofix (Amresco, Solon, CA), permeabilized with 0.05% Saponin for 10 min, and immunostained with FB-50 monoclonal antibodies to AAH using the avidin–biotin horseradish peroxidase method and diaminobenzidine as the substrate [15]. As negative controls, cells were incubated with non-relevant primary antibody or with the primary antibody omitted.

2.3. FACS analysis Cells were fixed in 2% formaldehyde/phosphate-buffered saline (PBS) and used either intact or after Saponin permeabilization. AAH immunoreactivity was detected with the FB50 antibody, biotinylated secondary antibody, and fluorescein isothiocyanate (FITC)-conjugated streptavidin. Fluorescence activated cell sorting (FACS) analysis was performed using a Becton Dickinson FACScan.

2.4. Non-directional motility assay The phagokinetic track motility assay which measures non-directional cell migration is based upon phagocytosis of gold particles in the paths of migrating cells. Glass coverslips, pre-coated with bovine serum albumin were coated in a colloidal gold solution [28] and then placed into 24-well plates. Cells were seeded at a density of 1000/ml/well. As a negative control, gold-coated coverslips were incubated with medium alone. After 4 h of migration, the cells were fixed with 4% paraformaldehyde and stained with hematoxylin. The coverslips were then mounted onto microscope slides. Phagocytosis resulted in gold particle-free plaques that were measured using the Kodak Digital Science Image Station.

2.5. Directional motility assay Directional motility was measured using Boyden chamber-type culture inserts equipped with porous membranes (8 mM pore diameters) [29]. The chambers were seated in 24-well plates. Viable cells (10 5) were placed into the upper chamber in serum-free medium (see cell culture conditions). The same medium supplemented with 1% FCS was placed in the well below to provide a trophic stimulus for migration. Migration was allowed to proceed for 4 h. ATPLite (Packard Instrument Co., Meriden, CT) was used to quantify viable cell density on the upper surface of the membrane (nonmotile), the undersurface of the membrane (migrated adherent), and at the bottom of the well (migrated, non-adherent) [29]. ATPLite luminescence was measured in a TopCount Microplate reader (Packard Instrument Co.). The percentages of motile adherent, motile non-adherent, total population of motile cells were calculated per chamber. Results from six replicate assays per group were analyzed statistically.

2.6. Antisense oligodeoxynucleotide (ODN) studies Experiments were performed using three antisense AAH ODNs, one sense AAH ODN, and one non-relevant ODN. The AAH ODNs overlapped the initiation codon of the human AAH mRNA. To demonstrate inhibition of AAH expression, in vitro transcription/translation assays were performed using TnT Quick Coupled Transcription/Translation Systems (Promega, Madison, WI). The products were labeled with [ 35S]methionine and analyzed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and autoradiography. To examine the effects of antisense ODN inhibition of AAH expression in vivo, cells were transfected with ODNs using streptolysin O (SLO) permeabilization [30– 32]. For each assay, 10 6 cells were transfected with antisense AAH, sense AAH, or non-relevant

T. Maeda et al. / Journal of Hepatology 38 (2003) 615–622 ODNs and evaluated by Western blot analysis or ATPLite-based directional motility. In addition, to demonstrate specificity of the antisense effects, cells were transfected with mutated antisense AAH ODNs (Mut 1: TTAGGCTGGGGCATTGGAGG; Mut 2: TTACGGTGCGCCAATGCACC) where the underscored G/C bases were substituted for C/G, and the A/T was substituted for T/A in the wild-type AAH ODN.

617

tiated; 43%), consistent with the differences observed by Western blot analysis. 3.2. Analysis of non-directional and directional motility

3.1. AAH expression in cholangiocarcinoma cell lines

In the phagokinetic gold particle assay, motile cells phagocytose gold particles in their paths, and the areas of peri-cellular clearance correspond to the degrees of cell motility. H-1 cells which express low levels of AAH exhibited minimal peri-cellular gold particle clearance (Fig. 3A), whereas NEC (Fig. 3B) and ETK-1 (Fig. 3C) cells, which had higher levels of AAH and were derived from moderately or poorly differentiated cholangiocarcinomas, exhibited large areas of gold particle clearance (Fig. 3D). Similar results were obtained for the other cell lines generated from poorly differentiated cholangiocarcinomas and with high levels of AAH expression. Statistical analysis of results obtained by measuring 25 areas of gold particle clearance from each culture demonstrated significantly higher motility indices in ETK-1 (P , 0:001) and NEC (P , 0:01) compared with H-1 cells (Fig. 3D). An ATPLite-based assay was used to measure directional motility. This assay has the advantage of accounting for all cells in the culture system and therefore permits calculation of percentages of motile cells. In addition, the assay permits distinction between motile adherent and motile non-adher-

Western blot analysis using the FB50 mAb detected the expected ,86 kDa AAH protein in all five cholangiocarcinoma cell lines studied (Fig. 1A). Northern blot analysis confirmed the presence of major ,2.8 kb and minor ,5 kb mRNA transcripts (data not shown), consistent with previous observations in human hepatocellular carcinoma cell lines [15]. By Western blot analysis, similarly high levels of AAH expression were detected in two cell lines (ETK-1, SSP-25) that were generated from poorly differentiated (111) tumors, high levels were observed in RBE and NEC cells which were derived from moderately differentiated (1 1 ) tumors, and relatively low levels were observed in H-1 cells derived from a well-differentiated (1) cholangiocarcinoma (Fig. 1A,B). In contrast, a-tubulin expression was similarly abundant among the different cell lines (Fig. 1A). Immunocytochemical staining localized AAH immunoreactivity to both the cell surface and perinuclear regions (Fig. 2A,B). Perinuclear staining is consistent with AAH localization in the endoplasmic reticulum. The presence of both surface and intracellular AAH immunoreactivity was confirmed by FACS analysis of non-permeabilized and permeabilized cells. Representative results using NEC and H-1 cells are depicted in Fig. 2C–E. FACS analysis revealed more abundant labeling of permeabilized compared with non-permeabilized cells, reflecting higher intracellular levels of AAH. In addition, the percentage of NEC cells (moderately differentiated) with surface FB50 immunoreactivity was higher (81%) than for H-1 cells (well-differen-

Fig. 1. Human AAH and a-tubulin expression in five different cholangiocarcinoma cell lines demonstrated by Western blot analysis. AAH protein was detected with the FB50 monoclonal antibody, and immunoreactivity was revealed with horseradish peroxidase conjugated secondary antibody, enhanced chemiluminescence reagents, and film autography. (A) Western immunoblots demonstrating expression of the ,86 kDa AAH protein and a-tubulin (, 55 kDa) in the same homogenates. (B) Densitometric analysis of the above AAH Western autoradiograph demonstrating higher levels of AAH protein in cholangiocarcinoma cell lines generated from moderately (1 1 ) or poorly (111) differentiated tumors compared with the H-1 cell line which originated from a well-differentiated (1) tumor.

2.7. AAH reagents The ,2.3 kb coding region of the human AAH cDNA was ligated into the pcDNA3.1 vector (Invitrogen, Carlsbad, CA) in which gene expression is regulated by a CMV promoter. AAH immunoreactivity was detected with the FB50 [15] monoclonal antibody which binds to a core epitope comprising residues 286–291 (NPVEDS), which is contained within the proteolytically processed ,52 kDa and ,56 kDa forms of AAH [27].

2.8. Statistical analysis Data depicted in the graphs represent means ^ SDs generated with results obtained from individual experiments with 3–6 replicate assays, and all in vitro experimental results were reproduced at least three times. Between-group comparisons were made with Student t-tests or analysis of variance (ANOVA) using the Fisher least significant difference (LSD) post hoc test.

3. Results

618

T. Maeda et al. / Journal of Hepatology 38 (2003) 615–622

membrane (Fig. 4C). Future studies will determine the significance of these results with respect to the metastatic potential of these neoplasms. 3.3. Antisense inhibition of AAH expression and its effect on cell motility To further investigate the role of AAH as a mediator of cell motility, AAH expression was inhibited with antisense ODNs. Three antisense and one sense ODNs that targeted the initiation codon of human AAH mRNA were used for in vitro transcription/translation studies (Fig. 5A). In addition, parallel studies were conducted using non-relevant ODNs. The in vitro transcription/translation studies demonstrated expression of the expected ,86 kDa AAH protein using the human AAH cDNA as a template (Fig. 5B). Inclusion of antisense AAH ODN in the reactions produced dose-dependent reductions in AAH expression, while the same concentrations of sense or non-relevant ODN had no effect on the levels of AAH protein synthesized (Fig. 5B). Although inhibition of AAH was achieved with each of the antisense ODNs used, the best results were obtained using ODNs that bound to AAH mRNA beginning at positions 26 and 211 relative to the AUG initiation codon. Therefore, we used the Location 26 ODN to inhibit AAH expression in cells. Most mammalian cell types take up antisense ODNs by highly inefficient endocytic mechanisms, and most intracellular oligomers are rapidly sequestered into vesicles, separating them from their target mRNA molecules. In contrast,

Fig. 2. Distribution of AAH immunoreactivity in cholangiocarcinoma cells demonstrated by immunocytochemical staining (A,B) and FACS analysis (C–E). (A,B) Low (A) and high (B) magnification images of permeabilized NEC cells immunostained with the FB50 mAb using the avidin-biotin horseradish peroxidase method with diaminobenzidine as the chromogen. Immunoreactivity was detected on the surfaces and perinuclear regions of the cells. (C–E) FACS analysis of non-permeabilized (C2,D2) and permeabilized (C3,D3) NEC (C) and H-1 (D) cells labeled with the FB50 mAb. Profiles of unlabeled control cells are illustrated in panels C1 (NEC) and D1 (H-1). The percentages of non-permeabilized and permeabilized cells labeled with the FB50 mAb are graphed in panels E1 and E2, respectively.

ent populations, which may be relevant to invasive versus metastatic growth of neoplastic cells. Using the ATPLite directional motility assay, significantly higher motility indices were measured in NEC and RBE cells compared with H-1 cells (P , 0:01; Fig. 4A). In addition, the assay revealed differences between the NEC and RBE cells in that substantially higher percentages of the motile RBE cells remained adherent to the undersurfaces of the porous membranes relative to NEC cells (Fig. 4B), whereas the majority of motile NEC cells were non-adherent to the

Fig. 3. Cholangiocarcinoma cell motility correlates with AAH expression. (A–C) Non-directional motility was examined in H-1 (A), NEC (B), and RBE (C) cells using the phagokinetic track motility assay in which motile cells phagocytosed gold particles leaving cleared (dark) areas corresponding to the paths of migration. Larger areas of clearance correspond to greater degrees of cell motility. (D) Quantification of non-directional motility by measuring 25 pericellular areas of clearance per culture. Graphed data reflect mean ^ SD of cleared areas measured as square-pixels/cell. (*P , 0.01; **P , 0.001 relative to H1 cells).

T. Maeda et al. / Journal of Hepatology 38 (2003) 615–622

619

AAH, non-relevant, or mutated antisense ODN as demonstrated by Western blot analysis (Fig. 6A). In addition, the AAH protein levels were similar in cells transfected with mutated antisense AAH, sense AAH, or non-relevant ODN (Fig. 6A). The Western blot signals for AAH shown in Fig. 6 are less intense than those depicted in Fig. 1 because only 10 6 cells were used for individual ODN transfections and

Fig. 4. Increased directional motility with high levels of AAH expression. Directional motility was measured in well-differentiated H-1 cells which had low levels of AAH and in moderately differentiated tumor cells with intermediate (NEC) or high (RBE) levels of AAH using Boyden chamber-type cell culture inserts and an ATPLite luminescence-based assay to determine percentages non-motile, motile adherent, and motile non-adherent cells (see Section 2). Note differences in the percentages of motile adherent and motile non-adherent NEC and RBE cell populations, despite similar overall motility indices. Graphs depict mean ^ SD of results obtained from six replicate assays (*P , 0.005, **P , 0.001 relative to the other groups by ANOVA).

ODNs introduced directly into the cytoplasm by microinjection, rapidly accumulate in the nucleus. Poor target cell delivery is another major problem limiting the routine use of antisense ODNs. Consequently, a great deal of effort has been placed on improving the methods of gene delivery for in vitro experimentation. One method developed to enhance ODN delivery involves streptolysin O treatment, which reversibly permeabilizes the plasma membrane, thereby enabling a biochemical-type ‘microinjection’ of ODNs directly into the cytoplasm [30–32]. In a previous study, SLO treatment of KY01 myelogenous leukemia cells resulted in the accumulation of more than 100-fold higher intracellular levels of ODNs with substantial internalization into nuclei and specific suppression of the target mRNA compared with other method of ODN delivery. We used the SLO method to transfect ODNs into cholangiocarcinoma cells. To verify DNA transfection into 100% of the cells, FITC-labeled ODNs were used and the cells were subjected to FACS analysis. Time course studies demonstrated intracellular persistence of FITC-labeled ODN for at least 6 h. Initial studies of the effects of antisense and sense AAH ODNs on AAH expression and motility were performed with each of the CCC lines. However, after demonstrating similar responses among the different CCC lines, RBE and SSP-25 cells were arbitrarily selected for more detailed analysis. Representative data obtained with RBE cells are shown in Fig. 6, but similar results were obtained using SSP-25 cells. Transfection with Location 26 antisense AAH ODN substantially reduced AAH protein expression relative to cells transfected with sense

Fig. 5. Antisense inhibition of AAH expression. (A) Sequences of oligodeoxynucleotides (ODNs) tested using in vitro transcription/translation assays. Locations 21, 26, and -11) refer to the positions of the first nucleotide relative to the AUG start codon. The sense primer overlaps with Location 26 of AAH mRNA. (B) Results of in vitro transcription/ translation assays of AAH expression in the presence of AAH antisense or sense negative control ODNs. In vitro transcription/translation assays were performed with the Tn nT Quick Master Mix. [ 35S]Methioninelabeled products were analyzed by SDS–PAGE and autoradiography. Lane 1 (no ODN) represents the positive control assay of AAH cDNA plasmid only. Lane 2 (Luc) shows a ,65 kDa protein generated in reactions containing a luciferase cDNA construct. Additional lanes depict results from parallel studies in which reaction mixtures contained AAH cDNA plasmid plus 200 £ , 2000 £ , or 20 000 £ molar excess of antisense (Loc 21, Loc 211, Loc 26) or sense AAH ODN.

620

T. Maeda et al. / Journal of Hepatology 38 (2003) 615–622

Fig. 6. Antisense oligodeoxynucleotides (ODNs) inhibition of AAH expression and motility in RBE cells. 10 6 RBE cholangiocarcinoma cells were transfected with Location 26 antisense AAH, sense AAH, mutated (Mut-1, Mut-2) antisense AAH, or non-relevant ODNs using SLO permeabilization (see Section 2). (A) AAH and a-tubulin expression were evaluated by Western blot analysis. (B–D) Directional motility was measured using the ATPLite luminescence-based assay (see Section 2). Graphs depict the mean (^SD) percentages of total migrated, migrated adherent, and migrated non-adherent populations calculated for six replicate assays per group (*P , 0.001 relative to all other groups by ANOVA with post hoc Fisher LSD testing for significance).

immunoblot analysis, whereas 60 mg protein was used for the conventional Western blot analysis. Cells transfected with antisense AAH ODN had significantly reduced mean directional motility indices relative to cells transfected with sense AAH, mutated antisense AAH, or non-relevant ODN, as demonstrated with the ATPLitebased directional motility assay. In RBE cells transfected with antisense AAH ODN, the mean percentage of motile cells was reduced to 24% from ,74% observed in control cultures transfected with sense AAH, mutated antisense AAH, or non-relevant ODN (P , 0:001; Fig. 6B). The subpopulations of migratory cells were calculated to determine if the adherent or non-adherent pools were preferentially affected. Those analyses demonstrated that the antisense AAH significantly reduced the populations of both migrated adherent and migrated non-adherent cells (P , 0:001; Fig. 6C,D); however, the effect was somewhat greater with respect to migrated non-adherent population (78

versus 56% reduction; P ¼ 0:05). Small differences in the proportions of migrated adherent and migrated non-adherent cells were noted between transfected and non-transfected control cells. This effect was likely due to technical aspects of the procedure since we have detected alterations in motility following transient transfection of other cell types.

4. Discussion The studies reported herein demonstrate over-expression of AAH in cholangiocarcinoma cell lines and a positive association between high levels of AAH and cell motility. In addition, antisense experiments showed that inhibition of AAH expression substantially reduces cell motility, a feature required for tumor invasiveness and metastatic spread. These results support the hypothesis that AAH has a functional role in cell migration.

T. Maeda et al. / Journal of Hepatology 38 (2003) 615–622

Previously, we demonstrated high levels of AAH expression in 20 of 20 primary human cholangiocarcinomas, and low or undetectable AAH expression in normal or proliferating intrahepatic bile duct epithelium [14]. In addition, previous experiments linked stable over-expression of AAH and constitutive AAH hydroxylase activity to transformation of NIH-3T3 cells [14]. Therefore, increased AAH gene expression and enzymatic activity contribute to cellular transformation and are not strictly related to proliferation. The findings in the present study suggest that AAH contributes to cellular transformation by rendering neoplastic cells more motile, a quality required for infiltrative and metastatic spread of tumors. Cholangiocarcinomas pose a challenging problem due to their resistance to currently available therapy and generally poor prognosis associated with infiltrative growth that is not amenable to surgical resection [1–5]. Although cancer cell invasion and metastasis are mediated by complex processes that involve regulation and expression of proteolytic enzymes, adhesion molecules, and growth factors [33], the findings herein suggest that AAH represents an important target for inhibiting infiltrative growth of cholangiocarcinomas. In this regard, cholangiocarcinoma cells transfected with antisense ODNs that inhibit AAH expression exhibited significantly reduced motility in directional migration assays. Since motility is required for invasive and metastatic tumor cell growth, and the AAH gene is overexpressed in a very high percentage of human cholangiocarcinomas (100% examined), antisense AAH therapy may prove beneficial for treating and preventing spread of cholangiocarcinomas in vivo. Similarly, this approach may aid in the treatment of hepatocellular carcinomas since AAH is overexpressed in 40% of those neoplasms as well [15]. Antisense effects of ODNs in mammalian cells have been reported in numerous tissue culture experiments, and in several in vivo studies [27,30,34–38]. Antisense ODN therapy offers the potential to block expression of specific genes; however, the in vivo use of conventional phosphodiester-linked ODNs is limited by low physiological stability. In preliminary experiments, we tested several antisense and sense phosphorothioate-linked AAH ODNs, but encountered significant problems related to non-specific binding and inhibition of non-relevant gene expression as noted in recent review article [39]. Therefore, the studies reported herein were conducted using phosphodiesterlinked ODNs, limiting the timeframe for analysis of gene expression and motility to a relatively brief interval to ensure stability of the transfected DNA. Recently, the employment of compounds that have greater nuclease resistance combined with enhanced target specificity, such as occurs with phosphoramidate- [35] or 2 0 -O-(2-methoxy)ethyl-linked [40] ODNs has helped to overcome problems that may hamper in vivo use of antisense gene therapy. Continued rapid development of the chemistry required to enhance the stability and binding specificity of ODNs will augment our capacity to selectively disrupt the expression of

621

genes that promote the malignant phenotype. In this regard, clinical trials are currently in progress to evaluate the therapeutic potential of antisense ODNs in human disease. Since AAH expression is low-level or undetectable in many normal tissues including intrahepatic biliary epithelium, the findings herein suggest that the AAH gene could serve as a target in cholangiocarcinomas using antisense ODN technology. Moreover, the finding of abundant AAH immunoreactivity on the cell surfaces of moderately or poorly differentiated tumors as demonstrated by FACS analysis, suggests that immunotargeting of surface AAH epitopes using ‘humanized’ monoclonal antibodies or high affinity single chain antibody fragments [41,42] offers an additional therapeutic approach to cholangiocarcinomas. Acknowledgements Supported by Grants CA35711 and AA02666 from the NIH and a grant from the Uehara Memorial Foundation, Japan. References [1] Kajiyama K, Maeda T, Takenaka K, Sugimachi K, Tsuneyoshi M. The significance of stromal desmoplasia in intrahepatic cholangiocarcinoma: a special reference of ‘scirrhous-type’ and ‘nonscirrhoustype’ growth. Am J Surg Pathol 1999;23:892–902. [2] Maeda T, Adachi E, Kajiyama K, Sugimachi K, Tsuneyoshi M. Combined hepatocellular and cholangiocarcinoma: proposed criteria according to cytokeratin expression and analysis of clinicopathologic features. Hum Pathol 1995;26:956–964. [3] Taguchi K, Aishima S, Asayama Y, Kajiyama K, Kinukawa N, Shimada M, et al. The role of p27kip1 protein expression on the biological behavior of intrahepatic cholangiocarcinoma. Hepatology 2001;33:1118–1123. [4] de Groen PC, Gores GJ, LaRusso NF, Gunderson LL, Nagorney DM. Biliary tract cancers. N Engl J Med 1999;341:1368–1378. [5] Shimada M, Takenaka K, Kawahara N, Yamamoto K, Shirabe K, Maehara Y, Sugimachi K. Chemosensitivity in primary liver cancers: evaluation of the correlation between chemosensitivity and clinicopathological factors. Hepatogastroenterology 1996;43:1159–1164. [6] Kiba T, Tsuda H, Pairojkul C, Inoue S, Sugimura T, Hirohashi S. Mutations of the p53 tumor suppressor gene and the ras gene family in intrahepatic cholangiocellular carcinomas in Japan and Thailand. Mol Carcinog 1993;8:312–318. [7] Kubicka S, Kuhnel F, Flemming P, Hain B, Kezmic N, Rudolph KL, et al. K-ras mutations in the bile of patients with primary sclerosing cholangitis. Gut 2001;48:403–408. [8] Ohashi K, Nakajima Y, Kanehiro H, Tsutsumi M, Taki J, Aomatsu Y, et al. Ki-ras mutations and p53 protein expressions in intrahepatic cholangiocarcinomas: relation to gross tumor morphology. Gastroenterology 1995;109:1612–1617. [9] Tada M, Omata M, Ohto M. High incidence of ras gene mutation in intrahepatic cholangiocarcinoma. Cancer 1992;69:1115–1118. [10] Terada T, Nakanuma Y, Sirica AE. Immunohistochemical demonstration of MET overexpression in human intrahepatic cholangiocarcinoma and in hepatolithiasis. Hum Pathol 1998;29:175–180. [11] Voravud N, Foster CS, Gilbertson JA, Sikora K, Waxman J. Oncogene expression in cholangiocarcinoma and in normal hepatic development. Hum Pathol 1989;20:1163–1168. [12] Park J, Tadlock L, Gores GJ, Patel T. Inhibition of interleukin 6-

622

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

T. Maeda et al. / Journal of Hepatology 38 (2003) 615–622 mediated mitogen-activated protein kinase activation attenuates growth of a cholangiocarcinoma cell line. Hepatology 1999;30:1128–1133. Tanaka S, Sugimachi K, Shirabe K, Shimada M, Wands JR. Expression and antitumor effects of TRAIL in human cholangiocarcinoma. Hepatology 2000;32:523–527. Ince N, de la Monte SM, Wands JR. Overexpression of human aspartyl (asparaginyl) beta-hydroxylase is associated with malignant transformation. Cancer Res 2000;60:1261–1266. Lavaissiere L, Jia S, Nishiyama M, de la Monte S, Stern AM, Wands JR, Friedman PA. Overexpression of human aspartyl(asparaginyl)beta-hydroxylase in hepatocellular carcinoma and cholangiocarcinoma. J Clin Invest 1996;98:1313–1323. Lahousse SA, Sepe PS, Wands JR, de la Monte SM. Role of aspartyl(asparaginyl)-b-hydroxylase (AAH) in neuroblastoma migration [abstract]. Lab Invest 2001;81:208A. Tamaki S, de la Monte SM, Ince N, Torimura T, Ueno T, Wands JR. Over-expression of the aspartyl (asparaginyl) beta hydroxylase gene enhances invasive growth of hepatocellular carcinoma cells [abstract]. Hepatology 2000;32:322A. McGinnis K, Ku GM, Fu J, Stern AM, Friedman PA. The five cysteine residues located in the active site region of bovine aspartyl (asparaginyl) beta-hydroxylase are not essential for catalysis. Biochim Biophys Acta 1998;1387:454–456. McGinnis K, Ku GM, VanDusen WJ, Fu J, Garsky V, Stern AM, Friedman PA. Site-directed mutagenesis of residues in a conserved region of bovine aspartyl (asparaginyl) beta-hydroxylase: evidence that histidine 675 has a role in binding Fe 21. Biochemistry 1996;35:3957–3962. Monkovic DD, VanDusen WJ, Petroski CJ, Garsky VM, Sardana MK, Zavodszky P, et al. Invertebrate aspartyl/asparaginyl beta-hydroxylase: potential modification of endogenous epidermal growth factorlike modules. Biochem Biophys Res Commun 1992;189:233–241. Wang QP, VanDusen WJ, Petroski CJ, Garsky VM, Stern AM, Friedman PA. Bovine liver aspartyl beta-hydroxylase. Purification and characterization. J Biol Chem 1991;266:14004–14010. Jia S, McGinnis K, VanDusen WJ, Burke CJ, Kuo A, Griffin PR, et al. A fully active catalytic domain of bovine aspartyl (asparaginyl) betahydroxylase expressed in Escherichia coli: characterization and evidence for the identification of an active-site region in vertebrate alpha-ketoglutarate-dependent dioxygenases. Proc Natl Acad Sci USA 1994;91:7227–7231. Jia S, VanDusen WJ, Diehl RE, Kohl NE, Dixon RA, Elliston KO, et al. cDNA cloning and expression of bovine aspartyl (asparaginyl) beta- hydroxylase. J Biol Chem 1992;267:14322–14327. Enjoji M, Nakashima M, Honda M, Sakai H, Nawata H. Hepatocytic phenotypes induced in sarcomatous cholangiocarcinoma cells treated with 5-azacytidine. Hepatology 1997;26:288–294. Enjoji M, Sakai H, Nakashima M, Nawata H. Integrins: utility as cell type- and stage-specific markers for hepatocellular carcinoma and cholangiocarcinoma. In Vitro Cell Dev Biol Anim 1998;34:25–27. Enjoji M, Sakai H, Nawata H, Kajiyama K, Tsuneyoshi M. Sarcoma-

[27]

[28]

[29] [30]

[31]

[32]

[33] [34]

[35]

[36] [37]

[38]

[39] [40]

[41]

[42]

tous and adenocarcinoma cell lines from the same nodule of cholangiocarcinoma. In Vitro Cell Dev Biol Anim 1997;33:681–683. Dinchuk JE, Henderson NL, Burn TC, Huber R, Ho SP, Link J, et al. Aspartyl beta -hydroxylase (Asph) and an evolutionarily conserved isoform of Asph missing the catalytic domain share exons with junctin. J Biol Chem 2000;275:39543–39554. Kin M, Sata M, Ueno T, Torimura T, Inuzuka S, Tsuji R, et al. Basic fibroblast growth factor regulates proliferation and motility of human hepatoma cells by an autocrine mechanism. J Hepatol 1997;27:677–687. de la Monte SM, Lahousse SA, Carter J, Wands JR. ATP luminescence-based motility-invasion assay. Biotechniques 2002;33:98–100. Clark RE, Grzybowski J, Broughton CM, Pender NT, Spiller DG, Brammer CG, et al. Clinical use of streptolysin-O to facilitate antisense oligodeoxyribonucleotide delivery for purging autografts in chronic myeloid leukaemia. Bone Marrow Transplant 1999;23:1303–1308. Giles R, Grzybowski J, Spiller D, Tidd D. Enhanced antisense effects resulting from an improved streptolysin-O protocol for oligodeoxynucleotide delivery. Nucleosides Nucleotides 1997:1155–1163. Spiller DG, Tidd DM. Nuclear delivery of antisense oligodeoxynucleotides through reversible permeabilization of human leukemia cells with streptolysin O. Antisense Res Dev 1995;5:13–21. Fidler IJ. Critical factors in the biology of human cancer metastasis. Am Surg 1995;61:1065–1066. Broughton CM, Spiller DG, Pender N, Komorovskaya M, Grzybowski J, Giles RV, et al. Preclinical studies of streptolysin-O in enhancing antisense oligonucleotide uptake in harvests from chronic myeloid leukaemia patients. Leukemia 1997;11:1435–1441. Faria M, Spiller DG, Dubertret C, Nelson JS, White MR, Scherman D, et al. Phosphoramidate oligonucleotides as potent antisense molecules in cells and in vivo. Nat Biotechnol 2001;19:40–44. Wagner RW. Gene inhibition using antisense oligodeoxynucleotides. Nature 1994;372:333–335. Wakita T, Wands JR. Specific inhibition of hepatitis C virus expression by antisense oligodeoxynucleotides. In vitro model for selection of target sequence. J Biol Chem 1994;269:14205–14210. Zimrin AB, Pepper MS, McMahon GA, Nguyen F, Montesano R, Maciag T. An antisense oligonucleotide to the notch ligand jagged enhances fibroblast growth factor-induced angiogenesis in vitro. J Biol Chem 1996;271:32499–32502. Stein CA. The experimental use of antisense oligonucleotides: a guide for the perplexed. J Clin Invest 2001;108:641–644. Zhang H, Cook J, Nickel J, Yu R, Stecker K, Myers K, Dean NM. Reduction of liver Fas expression by an antisense oligonucleotide protects mice from fulminant hepatitis. Nat Biotechnol 2000;18:862–867. Boder ET, Midelfort KS, Wittrup KD. Directed evolution of antibody fragments with monovalent femtomolar antigen-binding affinity. Proc Natl Acad Sci USA 2000;97:10701–10705. Shusta EV, Kieke MC, Parke E, Kranz DM, Wittrup KD. Yeast polypeptide fusion surface display levels predict thermal stability and soluble secretion efficiency. J Mol Biol 1999;292:949–956.