A new human cholangiocellular carcinoma cell line (KMC-1)

Jourwal of Heparologfi 1992; 15: 28X-298 i:# I992 Elsevier Science Publishers B.V. All rights reserved. Ol68-8278/92iSO5.00

288 HEPAT 00934

A new human cholangiocellular carcinoma cell line ( dkihiro Iemura, Masafumi Maruiwa, Hirohisq Yano and Masamichi Kojiro The First

Depurtrnerlt

c$‘Pntholog_v,

K urwnc

UmYrrsit_v School oj Mnlicirte.

Kurutw, Jupun

(Received 26 September 1990)

We have recently established a cholangiocellular carcinoma (CCC) cell line, designated KMC-1, from a nude mouse subcutaneous tumor which developed after inoculation of a surgically resected peripheral type CCC from a 62-yearold Japanese male patient. KMC-I cells grew over a 26-month period and passaged 57 times. These cells retained the morphologic characteristics of both the original tumor and the subcutaneous tumor in the nude mouse, which mainly consisted of irregular tubules and invaded surrounding interstitial tissue in part with an indurate pattern. KMC-I cells grew in a monolayer pavement-like cell arrangement with tubular formation in part. Some cells and/oi* glands had a mucin-like substance inside. The doubling time of KMC-1 cells growing in serum-containing medium was 54 h at passage 31. Cell growth in serum-free medium was slow but steady. The number of chromosomes was distributed in range from 73 to 83 with modes of 76 and 78. KMC-1 cells secreted some tumor markers such as DUPAN-2, CA125 TPA, hCG, CA19-9 and ferritin, however, the secretion of DUPAN-2, and CA 19-9 and ferritin were only detectable in serum-containing and serum-free medium, respectively. These findings suggest that KMC-1 cells will provide a variety of experimental models for research on CCI.~ and “he mechanisms of tumor marker secretion. Key bVords:Cholangiocellular

carcinoma; Cell line; Bile duct carcinoma; Nude mouse; DUPAN-2

In spite of recent advances in diagnostic imaging and therapeutic techniques, prognosis of cholangiocellular carcinoma (CCC) is still unfavorable, because diagnosis of CCC, especially of the peripheral type, is difficult and because CCC tends to involve the vessels and to spread in early stages. Since the incidence of CCC is fairly low compared to hepatocehular carcinoma (HCC) ( I ,2), many scientists have studied HCC. Several established human HCC cell lines have been reported and have achieved favorable consequences (3-10). Established CCC cell lines would be applicable to many further investigations (I 1,12). However, since only five CCC cell lines have so far been established (13- 17), different and new CCC cell lines are needed. The present report describes the establishment and characterization of a new human CCC cell line from peripheral CCC transplanted into nude mice.

Corresponc/~ncc to: MasamichiKojiro. M.D., Kurume-shi, 830. Japan.

The First Department

Materials and Methods Patient

The tissue used for the culture was originally derived from the cholangiocellular carcinoma (CCC) of a 62year-old Japanese male patient who visited a hospital because of epigastric pain. A liver tumor was found in the left lateral segment by ultrasonography. Using computed tomography and angiography, the patient was clinically diagnosed as having CCC. On admission, HBsAg or HBsAb was not detected in his serum, the levels of serum r-fetoprotein (AFP), carcinoembryonic antigen (CEA), and ferritin were 4.0 ng/ml, 1.0 ng/ml, 121 ng/ml, respectively. Segmentectomy of the liver was performed on November 30, 1987. At the operation, the tumor was located in the lateral portion of left hepatic lobe without complication of cirrhosis. The cut surface of the resected liver showed a well-demarcated, but nonof Pathology, Kurume

University Sc;~ool of Medicine, 67 Asahi-machi,

A NEW CHOLANGIOMA

CELL LINE

encapsulated tumor (Fig. 1). Some pieces of the tumor were fixed in 10% formalin for light microscope examination. The paraffin-embedded sections were stained with hematoxylin and eosin (H & E), Mayer’s mucicarmine and perodic acid-Schiff (PAS) reaction. Transplantation into nude mouse

An aseptically obtained specimen from the tumor nodule was immediately placed in a BO-mm dish (Falcon, Becton Dickinson Labware, Oxnard, CA) filled with DMEM + 20% FBS (Dulbecco’s Modified Eagle Medium (Nissui Seiyaku, Japan) supplemented with 100 U/ml penicillin and 100 pg/ml streptomycin (GIBCO, Chagrin Falls, OH), 12 mM sodium bicarbonate, and 20% fetal boGne serum (M.A. Bioproduct, Walkersville, MD), and was cut into 1 to 2 mm3 pieces with two surgical blades. Four to five of these pieces were inoculated S.C.into the back of 4--6-week-old female athymic nude mice (BAL ic, nu,+m, Clea Japan). When the subcutaneous tumor had grown to about 1 cm in diameter, mice were killed under general diethyl ether anesthesia. The developed tumors were resected aseptically, and were used for serial transplantation, and light

289 and electron microscope examinations. The paraffinembedded sections of S.C. tumors were stained in the same way as the original tumor. Some pieces of tumor were fixed in Karnovsky’s fixing fluid for 5 h at room temperature. After rinsing three times with 0.1 M sodium cacodylate buffer, the sample was postfixed with 2% 0~0, with milloning buffer ( I : 1). dehydrated in a graded series of ethanol, and embedded in Epon 8 12. They were cut on an ultramicrotome, stained with uranyl acetate and lead citrate, and examined under a Hitachi H-500 electron microscope (Hitachi, Japan). Tumor markers in nude mouse serum

Serum obtained from mice after killing was assayed for measurement of the following tumor markers, AFP, CEA and carbohydrate antigen 19-9 (CA19-9) using Radioimmunoassay kits (AFP: Celltech Diagnostics; CEA: DAINABOT, Japan; CA 19-9: CENTOCOR). Culture medium

DMEM + 20% FBS was used as a primary culture medium (PCM), and after stable growth of the cell line was obtained, serum concentration was decreased to

Fig. 1. The gross finding of the resected original tumor. A yellowish white, well-demarcated, non-encapsulated tumor in the resected liver without cirrhosis.

A. IEMURA et al.

290

10% or 5% for cell maintenance. For serum-free culture, RPM1 1640 (Nissui) supplemented with selenium (Na$eO,) and antibiotics as chemically defined medium (CDM) was used (18). Tissue culture A specimen from the S.C. tumor of nude mice was obtained under sterile conditions washed in PCM, and then minced into fine pieces with two surgical blades. These small tissue fragments were filtered through a 20pm pore mesh to remove blood, washed in fresh Ca and Mg-free phosphate-buffered saline (PBS) (Nissui), centrifuged (800 rpm 5 min at 4 “Cj, and the pellet, suspended in 20 ml of PBS with collagenase (type IV, 144 U/ml, 0.5 mg/nl, Sigma Chemicals, St. Louis, MO) was incubated in a 37°C water bath for 60 to 80 min. After incubation, the supernatant was centrifuged, the pellet was then resuspended in PCM, and filtered through a 75-pm pore mesh to separate digested cell suspension from undigested tissue. These were separately placed in 35-mm culture dishes (Falcon), and cultured at 37 “C with 5% CO* in air. After primary culture, cells were maintained by changing the medium three times a week. Obsercation Cells were observed daily by using a phase contrast microscope (Nikon, Japan). For light microscopic obsertvation, cells grown on Lab-Tek tissue culture chamber slides (Miles Laboratories, Naperville, IL) were washed with PBS three times, fixed in absolute methanol for 30 min, then stained with H & E, Mayer’s mucicarmine and PAS. For electron microscopic observation, cells grown on Lab-Tek tissue chamber slides were treated in the same manner as previously described, and examined under a Hitachi H-500 electron microscope. Growtlt curce The growth curve was initially obtained by seeding KMC-1 cells at the 31st passage growing in DMEM + 5% FBS at 4.5. 105/T-25 flask in duplicate. Twenty-four hours after seeding, the cells were detached with trypsinEDTA, and the average number of viable cells from two flasks was determined by counting cells that excluded Trypan blue in a Biirker-Turk hemocytometer. Afterwards, cells were counted every 2 days in the same manner. The doubling time of the cell population was estimated in the logarithmic growth phase. The growth curve of KMC-1 cells growing in serum-free medium was obtained in the same manner. The serum-containing spent media were centrifuged (3000 rpm, 10 min at 4 ‘C), and assayed for DUPAN-2 and CA 125 by radio-

kits (CA125 immunoassay DUPAN-2: KYOWAMEDIX).

CENTOCOR

INC;

marker The tumor markers secreted by KMC-1 cells for 48 h in 5 ml of serum-containing medium and serum-free medium were examined. Each spent medium was obtained when the KMC-1 cells grew conlluently in the T-25 flask (Falcon). The supernatant, obtained after centrifuge (3000 rpm, 10 min at 4 “C) of spent media, was assayed for AFP, CEA, CA19-9, CA125, ferritin, TPA and DUPAN-2 by radioimmunoassay kits (ferritin: Gamma Dab Ferritin; TPA: ABSANGTEC MEDIA), and human chorionic gonadotropin (hCG) by enzyme immunoassay kit (hCG-CTP test, Takeda Chemicals, Japan). Tumor

Protein studies Spent serum-free medium which was collected at the time of medium change and stored at -20°C was used for protein studies. The spent medium was centrifuged (3000 rpm, 10 min at 4”C), and concentrated about 100 times by Stirred ultrafiltration cells model 8200 (Amicon, MA, U.S.A.) and Diaflo membrane YMlO (MWlO cut off) (Amicon) and assayed for the following 15 kinds of proteins by the method of Laurel1 using antibody immunoelectrophoresis (19); prealbumin, albumin, q-acid glycoprotein, x,-antitrypsin, q-antichymotripsin, z2macroglobulin, ceruloplasmin, transfer& complement C3, complement C4, fibrinogen, fibronectin, plasminogen, prothrombin, and retional-binding protein. Commercially obtained antibody (Medical & Biological, Japan, and DAKO, Denmark) was used for each test. Chromosome study KMC-1 cells at the 13th passage were used for chromosome analysis. The cells in the T-25 flask were treated with colcemid (0.04 rug/ml) and incubated at 37 “C for 2 to 4 h. Cells were detached with trypsin, and treated with hypotonic 0.075 M KC1 for 10 min, centrifuged, and fixed in methanol/acetic acid (3:l). The cells were dropped on glass slides and dried, and stained with Giemsa. G-banding was carried out by using the method described by Seabright which stains the cells with Giemsa after 5 min 0.0125% trypsin treatment (20). Flow cytometric analysis of nuclear DNA contents of KMC-1 cell One million KMC-1 cells suspended in 8 mM TrisHCI buffer (pH 7.6) were centrifuged (800 rpm, 5 min at 4 “C), the supernatant was then removed. The cells were

A NEW CHOLANGIOMA

CELL LlNE

291

Fig. 2. Histologic finding of the originai tumor showing proliferation

Fig 3. Histologic finding of the subcutaneous

of various sizes of tubules with an abundant x 100).

fibrous stroma (H & E s

tumor in nude mouse showing the tumor consisting of various sizes of tubules similar to the original tumor with less fibrous stroma (H & E stain. x 100).

A. IEMURA

292 resuspended with 4 ml of solution A (1 mM glycine, 0.3 M NaC1,O. 1% Triton-X 100, pH 1Ok.O.2mg of RNase Type II-A (Sigma) and 20 /lg of ethidium bromide (Wako Pure Chemicals, Japan) were incubated for 10 min at 4 C, stored in a dark place at 37 “C, and analysed within 1 h. Flow cytometric analysis was performed using an ORTHO CYTRON (Ortho Diagnostic Systems, Irvine, CA). Normal human peripheral blood lymphocytes were used as external staining controls for the diploid (2C) DNA content. Xenotmnsplantationof KMC-I cells Four to six-week-old female athymic nude mice (Clea) were used to examine the tumorigenicity of KMC1 cells. Ten and twenty million cells suspended in 1 ml DMEM + 5% FBS were inoculated subcutaneously into the backs of eight and three mice, respectively. The inoculated mice were killed within 4 months and their tissues were observed by light microscopy. Mycoplasmadetection Mycoplasma contamination was examined by Mycoplasma T.C. rapid detection system (Gen-probe, San Diego, California, U.S.A.). Spent medium was mixed by

et al.

[3H]mycoplasma DNA probe and incubated (72 “C, 2 h). After hybridization, hybrid was detected by scintillation counter.

Results Morphology of resected liver Histologically, the original tumor consisted of tubules of various sizes with abundant fibrous stroma. Invadrng interstitial tissue with indurate pattern was observed in part. Tumor cells had large nuclei with prominent nucleoli, and had Mayer’s mucicarmine- and PAS-positive mucin in the cytoplasm (Fig. 2). According tc’ the classification of cholangiocellular carcinoma by Sugibara and Kojiro (21), the original tumor was diagnosed as tubular adenocarcinoma, moderately differentiated type. Morphology of nude mouse tumor Two months after inoculation, S.C. tumors grew to about 1 cm in diameter. Tumors were elastic hard withotit adhesion to surrounding tissues. Microscopically, the tumor consisted of tubules of various size with a papillary growth pattern similar to the original tumor, but with less fibrous stroma (Fig. 3). Some tumor cells

Fig 4. Electron microscopic finding of the subcutaneous tumor in nude mouse showing some intracytoplasmic organelles and an equipment of microvilli on the cell surface (x

4133).

A NEW CHOLANGIOMA CELL LINE

had large nuclei at the basement membrane side and Mayer’s mucicarmine- and PAS-positive mucin in the cytoplasm. Ultrastructurally, column-shaped tumor cells formed irregular glands of various sizes with abun&ant microvilli toward the lumen and the basement membrane at the base. Nuclei were variable in size with irregular margins, and small number of mitochondria, lysosome, smooth endoplasmic reticulum. Golgi apparatus, and ribosomes were found in the cytoplasm (Fig. 4). Cells with an intracytoplasmic lumen were rarely observed.

293

AFP, CEA, CA19-9 in the serum of a nude mouse were at undetectable levels.

of 26 months, they were considered an established cell line, and named KMC-I as a new human cholangiocellular carcinoma cell line. KMC-1 proliferate in a monolayer pavement-like cell arrangement with no contact inhibition. These cells grow with a partly tubular formation (Fig. 5). KMC-1 cells have clear cytoplasm and oval-shaped nuclei. Multinucleated giant cells are also observed among them. Some KMC-I cells have Mayer’s mucicarmine- and PAS-positive mucin in the cytoplasm. By electron microscopy, certain types of junctional complexes, i.e., tight junctions and desmosomes, and microvilli are observed on the cell surface between the cells (Fig. 6a and b). Cytoplasm and nuclei features are similar to those of original tumor.

Establishment and morphology ofthe cell line

Growth kinetics

A few days after the primary culture, small scattered colonies composed of epithelial-like cells were observed in some dishes. Fibroblast-like cells were also found around these colonies. Subsequently, the colonies of epithelial-like cells increased in size, and drove away any surrounding least fibroblast-like cells. These colonies were transferred to T-25 flasks by trypsin-EDTA treatment. Since these cells had been growing continuously and had been serially passaged 56 times over a period

The growth curve was initially obtained by seeding cells growing in DMEMA -I-5% FBS at the 3 1st passage. Population doubling time was 54 h. The correlation between cell growth and DUPAN-2 or CA125 secretion is shown in Fig. 7a and b, respectively. The concentration of DUPAN-2 per flask cc per IO6 cells was almost parallel to the growth curve from the late logarithmic growth phase to the middle stationary phase. DUPAN2 was undetectable in the early logarithmic phase. The

Tumor makers of nude mouse

Fig.5. Light microscopic finding of KMC-I cells showing a pavement-like

x loo).

cell arrangement

with glandular structure in part (H 8~ E stain.

A. IEMURA et al.

294

Fig. 6. Electron mxroscopic finding of KMC-I cells: (a) similar features as those of the nude mouse tumor; inset: desmosomes between the cells (x 4133; inset: x 12857); (b) poorly developed organelles, containing some microfilaments in lhe cytoplasms and junctional complexes and an Gquipment of microvilli between the cells (x 8266).

logarithmic secretion of CAI 25 by KMC-1 cells was identified from the early logarithmic to middle stationary growth phase. The concentration of CA 125 per IO6cells was constant from the beginning to the late logarithmic growth phase, then it increased, and nearly reached a pIateau. The growth curve of cells in serum-free medium could not be obtained because cells grew very slowly

and their attachment to dishes was poor. In spite of the slow growth, however, a confluent monolayer was formed after slow but steady growth. Tumor markers

KMC-1 cells growing in a serum-containing medium secreted CA125 (750 U/ml), TPA (15 000 U/ml), hCG

A NEW CHOLANGIOMA

CELL LINE

295 DUPAN-2 (U/ml) ,.O@O

105

I:i

10 Dsyr

Fig.7. (a)Correlation 2 (U/ml per 10 ceils) (

efkv

. 0

reecung

Days

alter

t

,

I

10

20

1

seedlng

ween KMC-I :eli growth and DUPAN-2 secretion. Number of cells ( _,).DUPAN-2 (U:ml per flask (b) Correlation between KMC-I cell growth and CA125 se ion. Number of cells (Z?). CA13 ( and CA125 (U mt per 10 cells)(

(13 mu/ml), DUPAN 2 (180 U/ml), CEA (P.7 nglml), whereas the cells growing in a serum-free medium secreted not only CA125 (270 U/ml), TPA (94 U/l), hCG (24 mU/mi j, but also CA 19-9 (I 40 U/ml), ferritin (6.9 ng/ ml). Neither CAl9-9 nor DUPAN 2 was detected in the serum-containing medium. AFP was not detected in either the serum-containing or free medium. Proreins

The I5 kinds of proteins examined were not detected in the spent serum-free medium, using the method of Laurel1 containing antibody immunoelectrophoresis. Chromosome stltdq The number of chromosomes was ranged from 73 to 83 with the modes of 76 and 78 (Fig. 8). G-banded karyotypes of the cell containing 77 chromosomes are shown (Fig. 9). The karyotype was human type and abnormal structures (7qf, 12pf, 13p+, l5p+, l7p’,

narisomy of 15, I7 and 22, and trisomy of I and 10) were observed in all of the cells with 77 chromosomes, Many cells had a tetrasom} of 2.

Aneuploid cell populations were observed with a DNA index of 1.76 by flow cytometry using ethidium bromide technique (Fig. IO).

Only one of the mice inocuiated with 2. IO7 KMC-I cells developed tumor. The others did not have an apparent tumor. Histologically, although most of the tumor changed into necrosis and hemorrhage, various sizes of tubules with abundant fibrous stroma were observed in the tumor. My~oplasrnn derection

Mycoplasma

contamination

was negative.

Discussion

73

7.4

75

76

77

Number

70 of

79

a0

8i

a2

03

chromosomes

Fig. 8. The distribution of chromosome

number al the 13th passage.

CCC is a malignant neoplasm arising from the intrahepatic bile duct epithelium. The current incidence of CCC is only 5.57r among primary hepatic malignancies in Japan, but this percentage is increasing (l,2), and more available CCC cell lines need to be established, to provide a variety of in vitro models which will be helpful in understanding the etiology, diagnosis and therapy of CCC. Only five CCC cell lines (I 3-l 7) and one nude mouse line (22) have been reported in the literature until now. The reasons for this small number of CCC cell

A. IEMURA et al.

296

B (4-5)

A (l-3)

C

(6-12)

E (16-18)

D (13-15)

F (19-20)

Unidentified

G (21-22)

chromosome

Fig. 9. A G-banded karyotype of KMC-1

2c

Fig. 10.

4c

64

6t

Flow cytometric analysis of nuclear DNA contents of KMCI cells shows aneuploidy cell population.

might be as follows; (i) the incidence of CCC is fairly low. (ii) Obtaining specimens suitable for primary culture is difficult due to complications as the tumor grows, of cholestasis, cholangitis, and abscess formation for example. (iii) The mincing of CCC tissue and the management of specimens by enzymes tend to be incomplete due to abundant interstitial tissue. (iv) Colonies of cancer cells do not grow well in the over growth of surrounding fibroblast-like cells. In the case of KMC-1 cell line, the second and fourth conditions noted above were avoided in our study, since our transplanted tumors lines

x Y

cell at the 13th passage with 77 chromosomes.

in nude mice showed no evidence of bacterial infection, and had less fibrous stroma than the patient’s original tumor. For these reasons, in the ftiture establishing a cell line through serially transplantable tumors in nude mouse might be easier and desirable, if the characteristics of the original tumor can be retained. The glandular structure and mucin production of the original tumor was diagnosed as tubular adenocarcinoma. KMC-1 cells also grow with tubular formations and with mucin productions. Tridimensional culture methods with type-l collagen gel matrix also showed apparent tridimensional glandular structures by KMC1 cells (data not shown). Equipment of microvilli is generally one of the ultrastructural characteristics of cholangiocellular carcinoma and was observed on the surface of KMC-1 cells. From these findings, it can be considered that KMC-I cells successfully retain the morphologic characteristics of the original tumor. In addition to the morphologic characteristics, KMC- I cells possess the following characteristics of a malignant transformed cell line: (i) KMC-1 cells secrete some tumor markers, (ii)abnormal structures of chromosomes including many unidentified chromosomes have been identified in KMC-1 cells, (iii) the DNA index was 1.76 and DNA contents of KMC-1 cells showed aneuploid cell popula-

A NEW CHOLANGiOMA

CELL LINE

tions, and (iv) KMC-I cells produced a tumor in a nude mouse. For the diagnosis of hepatocellular carcinoma (I-ICC), most patients have chronic liver diseases, such as chronic hepatitis and/or cirrhosis (21,23,24), and have been followed up periodically, resulting in the early diagnosis of HCC. In cases of CCC, most patients do not have chronic liver diseases, and are usually found with an enlarged liver tumor in the advanced stages without any clinical symptoms (23,25). Thorotorast, hepatolithiasis, parasites and a few other factors have however been reported to be related to the course and etiology of some CCC cases (21,24,26). For these reasons, it is necessary to find sensitive and specific tumor markers to detect CCC in the early stages. KMC-1 cells secrete various tumor markers, including DUPAN-2, CA1 25, TPA, hC6, CA19-9 and ferritin, in both serumcontaining and serum-free mediums. DUPAN-2 is a murine monoclonal antibody against the human pancreatic adenocarcinoma cell line elicited by Metzgar et al. (27,28) who have suggested that it can recognize heavily glycosylated antigens, and for its antigenicity, sialic acid is essential (27). We consider that DUPAN-2 was not detected in the spent serum-free media of KMC-1 cells for the following reasons: (i) non-DUPAN-2 producing cells were selectively cultured; (ii) the synthesis of DUPAN-2 by KMC-1 cells was too slight to detect; (iii) serum-free media did not contain sialic acid. Availability of DUPAN-2 as a tumor marker is an arguable point (29-31). KMC-1 cells may provide useful formations to clarify the synthesis and secretion mechanisms of DUPAN-2. CAl9-9, which was discovered by Koprowsky et al. (32) in 1979, is a monoclonal antibody against human colon cancer, but is also known to be positive for patients with bile duct carcinoma and pancreatic carcinoma (33-35). It would be possible to obtain more sensitive and specific monoclonal antibodies against intrahepatic cholangiocellular carcinoma (bile duct carcinoma). In this report, we have described the establishment and characterization of a new human CCC ceil line, KMC-I. KMC-1 cells have retained the morphological characteristics of the original CCC tumor, and have been showing a steady growth from their primary culture. KMC-1 cells can provide a variety of in vitro models for research on CCC and its associated tumor marker synthesis and secretion mechanisms. For further study, rloning of KMC-I is required to clarify the morphologic heterogeneity of the cells and the relationships between secretion of tumor makers and one specific kind of cell types.

297

The authors gratefully acknowledge Mr. K. Takashim photomicrographs, and Ms. S. Ogasawara, iI4S. atsuo, Ms. K. Najima, and Ms. M. Mori for technical assistance. This study was supported in part by the Sarah Cousins Fund, Boston, MA.

eferenees I Liver Cancer Study Group of Japan. Survey and ~oIIow-up study of primary liver cancer in Japan. Acta Hepato] Jpn 1988; 29:

428-56. 2 The Liver Cancer Study Group of Japan. Primary liver cancer in

Japan. sixth report. Cancer 1987; 60: 1400-I 1. T, Yano H, Maruiwa M, Sugihara S. Kojiro bl. Establishment and characterization of a human combined hepatocholangiocarcinoma cell line and its heterologous transplantation in nude mice. Hepatology 1987: 7(3): 551-6. 4 Yano H, Kojiro M, Nakashima T. A new humsn hepatocellula: carcinoma ceil line (KYN-1) with a transformation to adenocarcinoma. I? Vitro Cell Dev Biol 1986; 22( I I): 637-46. 5 Yano H, Maruiwa M, Murakami T, et al. A new human pleomorphic hcpatocellular carcinoma cell line, KYN-2. Acta Path01 Jpn 1988: 38(P): 953%6b. 6 Alexander JJ, Bey EM, Geddes EW, Lecatsas G. Establishment of a continuously growing cell line from primary carcinoma of the liver. S Afr Med J 1976; 50: 2124-8. 7 Das PK. Nayak NC, Tsiquaye KN, Zuckerman AJ. Establishment of a human hepatocellular carcinoma cell line releasing hepatitis B virus surface antigen. Br J Exp Pathol 1980; 61: 648-54. 8 Doi I. Namba M. Sato J. Establishment and some biological characteristics of human hepatoma cell lines. Gann 1975; 66: 385-92. 9 He L. Isselbacher KJ, Wands JR, Goodman HM, Shih C. Quaroni A. Establishment and characterization of a new human hepatocellular carcinoma cell line. In Vitro 1984: 2016):493-503. 10 Laohathai K, Bhamaraprava~i N. Culturing of human hepatoce!lular carcinoma: a simple and reproducible method. Am J Pathol 1985; I IS: 203-8 II Homma S, Nagamori S, Fujise K. et al. Effects of hyperthermia on the proliferation and viability of cultured human hepatobiliary carcinoma cells. Jpn J Gastroenterol 1985; 8219):2043-8. I2 Hudd C. Euhus DM. LaRegina MC, Herbald DR, Palmer DC, Johnson FE. Effect of cholecystokinin on human cholangiocarcinoma xenografted into nude mice. Cancer Res. 1985; 45: 1372-7. 13 Yamaguchi N. Morioka H. Ohkusa H. Hirohashi S. Kawai K. Establishment and characterization of the human cholangiocarcinnma cell line HChol-Yl in a serum-free. chemically defined medium. J Natl Cancer Inst 1985; 75(l): 29.-35. 14 Miyagiwa M, Ichida T, Tokiwa T. Sato J. Sasaki H. A new human cholangiocellular carcinoma cell line (HuCC-Tl) producing carbohydrate antigen 19i9 in serum-free medium. In Vitro Cell Dev Biol 1989: 25(6): 503-10. 1.5 Homma S. Nagamori S, Fujise K. et al. Human bile duct carcinoma cell line producing abundant mucin in vitro. Gastroenterol Jpn 1987; 22(4): 474-9. 16 Kusaka Y, Tokiwa T, Sato J. Establishment and characterization of a cell line from a human cholangiocellular carcinoma. Res Exp Med 1988; 188: 367-75. Y. Cholangiocarcinoma-H-I cell line. In: Ohboshi S. 17 Hayada Sugano H ed. Tissue Culture of Human Cancer Cells. Japan, Asakura, 1975: 181-I. H, Taketa K. Yamane T, Oda M. Sat0 J. Hormonal 18 Nakabay3shi 3 Murakami

298

19

20 21

22

23

24

25

26

27

control of r-fetoprotein in human Ilepatoma cell lines proliferating in chemically defined medium. Cancer Res. 1985; 45: 6379-83. Laurel1 CB. Quantitative estimation of proteins by electrophoresis in agarose gel containing antibodies. Anal Biochem 1966: 15: 45-52. Seabright M. A rapid banding technique for human chromosomes. Lancet 1971; ii: 971-2. Sugihara S, Kojiro M. Pathology of cholangiocarcinoma In: Okuda K, lshak KG, eds. Neoplasms of the Liver. Tokyo: Springer-Verlag. 1987: 143-58. Kubota T, Hanatani Y, Yamada Y, et al. Experimental chemoand radio-therapy on human cholarlgiocarcinoma transplanted to nude mice. J Jpn Sot Cancer Ther 1981: 16(3):414-21. Okuda K. Kubo Y, Okazaki N, et al. Clinical aspects of intrahepatic bile duct carcinoma including hilar carcinoma. Cancer 1977; 39: 232-46. Mizumoto R, Kawarada Y. Diagnosis and treatment of cholangiocarcinoma and cystic adenocarcinoma of the liver. In: Okuda K. lshak KG, eds. Neoplasms of the Liver. Tokyo: Springer-Verlag, 1987: 381-96. Nakajima T, Kondo Y. Miyazaki M, Okui K. A histopathologic study of 102 cases of intrahepatic cholangiocarcinoma: histologic classification and modes of spreading. Hum Pdthol 1988; 19(10): 1228-34. Nakanuma Y. Terada T. Tanaka Y, Ohta G. Are hepatolithiasis and cholangiocarcinoma aetiologically related? Virchow’s Arch. A (Pathol. Anat.) 1985; 406: 45-58. Metzgar RS, Rodriquez N, Finn OJ, et al. Detection of a pancreatic cancer-associate antigen (DUPAN-2 antigen) in serum and ascites of patients with adenocarcinoma. Proc Nat1 Acad Sci USA 1984: 8 I: 5242-6.

A. IEMURA et al. 28 Metzgar RS, Gaillard MT, Levine SJ, Tuck FL, Bossen EH, Borowitz MJ. Antigens of human pancreatic adenocarcinoma cells defined by murine monoclonal antibodies. Cancer Res 1982: 42: 601-8. 29 Oguro M, Aoyagi Y, Saitou A, et al. Clinical evaluation of serum concentration and fucosylation index of AFP. PIVKA.l! sialyl SSEA-I (SLX). CA-50 and Dupan-2 for the diagnosis of hepatocellular carcinoma. Acta Hcpatol Jpn 1989; 30( I I): 1589-95. 30 Lan MS, Finn OJ. Ferns&n PD, Metzgar RS. Isolation and properties of a human pancreatic adenocarcinoma-associated antigen, DUPAN-2. Cancer Res 1985; 45: 305-10. 31 Ohkura H. Ohta N, Yoshino M. Enzyme-linked immunosorbent assay of DUPAN 2. Jpn J Clin Pathol 1987; 35(7): 769-73. 32 Koprowski H, Steplewski Z. Mitchell K, Herlyn MDH. Fuhrer P. Colorectal carcinoma antigens detected by hybridoma antibodies. Somat Cell Genet 1979; 5: 957-71. 33 Gupta MK, Arciaga R, Bocci L. Tubbs R, Bukowski R, Deodhar SD. Measurement of a monoclonal-antibody-defined antigen (CAl9-9) in the sera of patients with malignant and nonmalignant diseases: comparison with carcinoembryonic antigen. Cancer 1985; 56: 277-83. 34 Takahashi H, Miyazaki H, Deura M, Shimizu Y, Asukata I, Kameda H. Clinical evaluation of CAl9-9. TPA, IAP and 5’-NPDV as tumor markers of hcpatocellular, bile duct and pancreas carcinoma. Jpn J Cancer Clin 1985; 31(6): 623-30. 35 Yamanaka 0. Kawanami M. Ozaki H. et al. Usefulness ofcarbohydrate antigen 19-9 (CAl9-9) for the diagnosis of cholangioma As a role of a screening tumor marker in patients with the elevation of serum alkalinphosphatase. lcucine aminopeptidase and/or ;-glutamyl transpeptidase. Jpn J Gastrocnterol 1985: 82(5): 1419.