- Email: [email protected]
Establishment and characterization of two human ovarian endometrioid carcinoma cell lines (with or without squamous cell component)
GYNECOLOGIC
ONCOLOGY
25, 95-107 (1986)
Establishment and Characterization of Two Human Ovarian Endometrioid Carcinoma Cell Lines (with or without Squamous Cell Component) ISAMU ISHIWATA,
M.D., *,t CHIEKO ISHIWATA, M.D.,* MASAYUKI SOMA, C.T.,* AND HIROSHI ISHIKAWA, M.D.7
*Ishiwata Obstetrics and Gynecologic Hospital, Mite-shi, Ibaraki-ken 310, and i-Department of Anatomy, Jikei University, School of Medicine, Minato-ku, Tokyo 105, Japan Received May 20, 1985 The cell lines designated HMOA and HNOA were established from human ovarian adenoacanthoma and from a mouse graft of human ovarian endometrioid adenocarcinoma, respectively. These cell lines grew well without interruption for over 20 months. The cultured cells of both HMOA and HNOA lines were spindle, polygonal, and columnar, and showed a jigsaw puzzle-like arrangement and a piling-up tendency devoid of contact inhibition. When the HMOA cells were maintained at the confluent stage, the cells formed cysts and/or squamous metaplasia. The chromosome number of both cell lines varied widely and showed aneuploidy, while the modal chromosome number was stable at the diploid range. Both of these cell lines, HMOA and HNOA, were trdnsplanted into the subcutis of BALB/c nude mice and produced well-differentiated adenoacanthoma and poorly differentiated endometrioid adenocarcinoma, respectively. HMOA cells were characterized as producing large amounts of CA125 (ovarian carcinoma marker), in vitro, in the cyst-forming phase. The HNOA cells, however, did not produce CA125. ‘a 1986 Academic Press, Inc.
INTRODUCTION Human ovarian endometrioid adenocarcinoma may develop histogenetically from ovarian endometriosis, or may originate from the ovarian surface layer of mesothelium [3,12]. On the other hand, squamous epithelial foci often coexist with a well-differentiated ovarian endometrioid adenocarcinoma. According to its morphologically benign or malignant characteristics, it is called an endometrioid adenoacanthoma or an endometrioid adenosquamous carcinoma, respectively
[5,121. Recently, we have cultured an ovarian endometrioid adenocarcinoma that did not contain squamous epithelial foci, and an ovarian endometrioid adenoacanthoma that did contain a squamous epithelial foci. We then established a cell line of each culture (HNOA and HMOA cell line, respectively). We report the methods we used to establish the above cell lines, as well as the biological characteristics ’ To whom requests for reprints should be addressed at I-4-21, Kami-Mito, Mito-shi, Ibaraki-ken 310, Japan 95 0090-8258186$1.50 Copyright Q 1986 by Academic Press, Inc. All rights of reproductmn in any Form reserved.
96
ISHIWATA
ET AL.
of these lines. We found that (1) the formation of squamous epithelial foci of an adenoacanthoma is caused by squamous metaplasia of the adenocarcinoma, (2) the HMOA is a cell line that produced the antigenic determinant CA125 [1,8] that can be defined by using the monoclonal antibody [OC125, obtained from mouse immunized with an human ovarian cystadenocarcinoma cell (OVCA-433)], and (3) the above CA125 was produced in a large quantity when a cyst and/or gland had formed, and in a small quantity when only squamous metaplasia occurred. MATERIALS AND METHODS I, Materials HMOA. The material (ovarian tumor) was obtained at radical hysterectomy and bilateral salpingo-oophorectomy in a 57-year-old Japanese woman on July 23, 1982. A portion of the tumor was placed in culture. HNOA. The material (ovarian tumor) was also obtained by the same operation in a 71-year-old Japanese woman on August 26, 1981. Eight 2-mm fragments were transplanted into the subcutis of three BALB/c nude mice Q-week-old female; Kurea Co., Tokyo) and the tumor measuring about 10 mm in diameter which developed in only one mouse was resected on December 9, 1981, and the mouse graft placed in culture. 2. Culture
Technique
and Culture Media
The materials for both HMOA and HNOA lines were rinsed twice with the growth medium, minced with scissors, and dissociated in 600 pronase Units/ml of dispase (Gohdo Shusei Co., Tokyo) for 30 min at room temperature and finally centrifuged at 1000 rpm for 10 min. The sediments were resuspended in the growth medium, placed on the 6-cm plastic dishes (Terumo Co., Tokyo) and cultured at 37°C in the 5% CO, incubator. The growth medium used was Ham’s F-12 (Grand Island Biological Co., Grand Island, N. Y.) supplemented with 15% fetal calf serum (Flow Laboratories, Beltsville, Md.), 40 pg kanamycin/ml, pH 7.2 to 7.4. In the initiation of the primary culture, epithelial colonies and fibroblast sheets appeared in both the HMOA and HNOA cultures. The pure epithelial cells were obtained by colony isolation technique (filter papers soaked with a mixed solution of 0.1% trypsin and 0.02% EDTA were placed directly on the epithelial cells; the papers together with the cells were shaken off into new plastic dishes [ 171). 3. Histology
The tumor was fixed with 10% buffered Formalin, embedded in paraffin, cut into thicknesses of 4 pm, and stained with hematoxylin and eosin, mucicarmine, alcian blue, periodic acid-Schiff, with or without 1% diastase, for 10 min. The cultured cells were also stained with Papanicolaou, Giemsa, periodic acid-Schiff, mucicarmine, and alcian blue. The grafts in nude mice and the cultured cells were observed with the JEM 1OOBelectron microscope. Details of the procedures can be found elsewhere [61.
HISTOGENESIS,
ADENOACANTHOMA,
CELL
LINE
97
4. Growth Characteristics Growth curves, population doubling times, saturation density, and plating efficiency were studied at the 10th and 40th passage. Approximately 1 x IO5 cells were placed into 3.5cm plastic dishes and cultured for approximately 10 days. Cells in three dishes were counted using an Erma counting chamber (Erma Co., Tokyo) and the average number was plotted each day. Population doubling time and saturation density were determined using a growth curve [6]. The medium was changed every 2 days. To study the plating efficiency, 250 single suspended cells were placed into five plastic dishes (6 cm in diameter), cultured for about 14 days, and finally stained with Giemsa. Plating efficiency was determined as the ratio of the number of visible colonies (more than 10 cells) to the number of the inoculated cells. 4. Chromosomal Analysis Chromosomal analysis was done at the 5th and 20th passages. The cultured cells were treated with 0.1% trypsin solution for 15 set at room temperature, stained with 3% Giemsa, and analyzed with G-band karyotyping [7]. Histograms of chromosome number distribution were determined on more than 100 metaphase plates. 5. CA125 Assay Approximately 5 x 10’ cells of HMOA and HNOA at passage 20 were cultured respectively for 2 days and the CA125 level in the conditioned media was measured using a CA125 RIA KIT (monoclonal antibody; Centocor) with the sandwich methods of radioimmunoassay [II]. The cellular CA125-producing activity was determined as the amount of CA125 produced by one cell for 2 days. Triplicate assays were performed to determine the CA125 levels. 6. Heterotransplantation Approximately 1 x 106, I x IO’, and 1 x IO8HMOA and HNOA cells (both at passage 20) were transplanted respectively into the subcutis of 2 BALB/c nude mice (8 week-old female; Kurea). The tumors were examined histologically by light microscopy 2 months after transplantation. The serum CA125 was measured by radioimmunoassay at the same time. RESULTS
I. Pathology of Original Tumors The original ovarian tumor for the HMOA line was interpreted pathologically to be an endometrioid adenoacanthoma composed of well-differentiated endometrioid adenocarcinoma with benign squamous foci (Fig. 1). The cytoplasms of squamous foci were strongly stained with periodic acid-SchifI-digested diastase and negatively with mucicarmine and alcian blue. On the contrary, the cytoplasm of the glandular tissue was slightly stained with mucicarmine and periodic acidSchiff. The material (mouse graft) of the HNOA cell line was interpreted to be a poorly differentiated endometrioid adenocarcinoma. The original tumor (patient’s
98
ISHIWATA
ET AL.
FIG. 1. Histology of the original human ovarian tumor of HMOA line revealed an adenoacanthoma composed of well-differentiated endometrioid adenocarcinoma and squamous metaplasia. Hematoxylineosin stain, x 100.
ovarian tumor) was considered to be a moderately differentiated endometrioid adenocarcinoma (Fig. 2). 2. Morphology
of Cultured
Cells.
Both HMOA and HNOA (Fig. 3) cells were spindle, polygonal, and columnar in shape and had an epithelial cell arrangement (jigsaw puzzle-like). They revealed a piling-up tendency without contact inhibition and showed anaplastic and pleomorphic features. When HMOA cells were maintained in the confluent condition they formed various sized cysts or showed squamous metaplasia (Fig. 4). Squamous metaplastic cells reacted strongly when stained with periodic acid-Schiff. On the contrary, the cells in cyst formation stained only slightly with mucicarmine and periodic acid-Schiff. Electron microscopy revealed that the HMOA cells were dark and had numerous short microvilli on the free surface. They made contact with each other by numerous desmosomes. Metaplastic cells contained a number of bundles of tonofibrils and glycogen granules scattered throughout the cytoplasm. The rough endoplasmic reticulum was poorly developed. The nuclear shape was irregular and a large nucleolus was present (Fig. 5). Another type of HMOA cell and HNOA cell also had a large nucleolus in an irregularly shaped nucleus. There were, however, many microtubules, small vesicles, moderately developed lamellary rough-endoplasmic reticulum, oval-shaped mitochondria, and a few secretory-like granules in the cytoplasm (Fig. 6).
HISTOGENESIS,
ADENOACANTHOMA,
CELL LINE
FIG. 2. Histology of the original human ovarian tumor of HNOA line revealed a moderaltely differentiated endometrioid adenocarcinoma. Hematoxylin-eosin stain, x 100.
FIG. 3. The monolayer-cultured cells of passage 5 (HNOA line) were spindle, polygonal, and columnar, and revealed neoplastic and pleomorphic features. Papanicolaou stain, x 100.
100
ISHIWATA ET AL.
FIG. 4. (a) The cultured cells of passage 5 (HMOA line) formed cysts or (b) showed squamous metaplasia. Phase contrast; (a) x40, (b) x200.
FIG. 5. Some HMOA cells (passage 5) were dark cells and had a number of the bundles of tonotibrils and made contact by desmosomes (arrow). This cell was considered a metaplastic cell. Transmission electron microscopy x 7900.
HISTOGENESIS,
ADENOACANTHOMA,
CELL LINE
101
FIG. 6. The HNOA cells (passage 5) had many microtubules, small vesicles, moderately developed lamellary rough-endoplasmic reticulum and a few secretory-like granules (arrow) in the cytoplasm. Transmission electron microscopy x 7900.
3. Growth Characteristics The growth curves are shown in Fig. 7. The growth characteristics such as population doubling time, saturation density, and plating efficiency are shown in Table 1. The population doubling time was 72 hr and plating efficiency was about 20% in the case of HMOA and 28 hr and 65%, respectively, in the case of HNOA.
o.l!
, 12
3 culture
4
5
6 7 days
8910
FIG. 7. The growth curve of the cultured cells at the 10th (0) and 40th (0) passage of HMOA line and that of the cultured cells at the 10th (X) and 40th (0) passage of HNOA line.
102
ISHIWATA
GROWTH
Cell line
PN
HMOA
10 40 10 40
HNOA
CHARACTERISTICS
ET AL.
TABLE 1 OF HMOA
AND
HNOA
CELL LINES
SD (1 x 104/cmZ) 72 58 28 24
5.2 5.6 5.9 6.0
20 25 65 72
Note. PN, passage number; DT, population doubling time; SD, saturation density; PE, plating efficiency.
4. Chromosome
Analysis
Karyotype of both HMOA and HNOA showed them clearly to be human cells. The chromsome number of both HMOA and HNOA was widely distributed and showed aneuploidy. However, the modal chromosome number was stable at the diploid range (Fig. 8). The HMOA line has a marker chromosome and the HNOA line is without a marker chromosome. The abnormal chromosome (2p-) was detected in about 75% of the cells (Fig. 9). 5. CA125 Assay
The CA125 level in the conditioned media in which 5 x lo5 cells of HMOA (at passage 20) were cultured for 2 days was 1140 U/ml. The CA125 level in the culture media (cell free) and also in the case of HNOA (at passage 20) was less than 8 U/ml. Therefore, the HMOA is considered to be a CA125-producing cell line, while the HNOA line is a CA125-nonproducing cell line. The amount of CA125 produced by one cell of the HMOA line for 2 days was 2.24 mU during HMOA
passage
HMOA
passage20
HNOA
passage
HNOA
passage20
chromosomal
5
5
number
8. Distribution of the chromosomal number of passages 5 and 20 of both HMOA and HNOA lines shows that the modal chromosomal number was 46-47. FIG.
HISTOGENESIS,
ADENOACANTHOMA,
CELL
103
LINE
f
M 19,:
1
2P-
6
7
13
14
4
3
8
15
9
10
16
5
11
12
18
21 22 19 20 xx FIG. 9. G-band karyotyping of passage 20 of HMOA and HNOA cells. That of HMOA shows 46, XXq - , - 13,Sq+ (1lp - ,19q + , marker chromosome. The telocentric marker chromosome was detected in lOO%, and that of HNOA shows 46,XX,2p-. 2P- was shown in 75% of HNOA cells.
monolayer culture, that in cyst formation was 3.8 mu, while that in metaplasia was 1.63 mu. 6. Heterotransplantation
Heterotransplantability of both HMOA and HNOA is shown in Table 2. The tumors produced by HMOA were interpreted to be an adenocarcinoma (Fig. 10)
104
ISHIWATA
ET AL.
TABLE 2 ONCOGENECITYOF HMOA LINE (PASSAGE20) AND HNOA LINE (PASSAGE20) IN NUDE MICE
No. of mice Cell line HMOA
HNOA
Inoculum size
Injected
With tumors
1 x lo6 1 x IO’ 1 x 108
2 2 2
0 1 2
1 x lo6 1 x 10’
2 2
2 2
lx
2
2
108
Tumor size (mm) 8 x 8 x 12 x 11 x 12 x 12 x
Pathology Well-differentiated endometrioid adenocarcinoma
10x 9x 9I 10 .-:-. 1 l4 x’ 10 l3 x’ lo9 Poor,y diff-... erenuatcu 15 x 13 x 11 endometricrid adenocarcinoma 15 x 12 x 10 16 x 14 x 11
CA125 levels in sera (U/ml) 276 290 304 <8 <8 <8 <8 <8 <8
and resembled the original tumor. On the contrary, that produced by HNOA was a poorly differentiated endometrioid adenocarcinoma (Fig. 11). The CA125 level in mouse sera was also measured by radioimmunoassay. The average CA125 level in the sera of three mice with HMOA tumor was 290 U/ml, while the level in case of HNOA was less than 8 U/ml, as illustrated in Table 2.
FIG. 10. The histology of the tumor produced by heterotransplantation of passage 20 cells of HMOA into nude mouse reveals an adenoacanthoma. Hematoxylin-eosin stain; (a) x250, (b) x 100.
HISTOGENESIS,
ADENOACANTHOMA,
CELL
LINE
FIG. 11. The histology of the tumor produced by heterotransplantation of passage 20 cells of HNOA into nude mouse reveals a poorly differentiated endometrioid adenocarcinoma. Hematoxylineosin stain, x 10. Inset: the area enclosed by the rectangle of the low magnification micrograph x 100.
DISCUSSION The HMOA and HNOA cell lines were proven to be an human ovarian endometrioid adenoacanthoma cell line, and an ovarian endometrioid adenocarcinoma cell line, respectively, based on the following observations. 1. The materials cultured were from an ovarian endometrioid adenoacanthoma and an ovarian endometrioid adenocarcinoma, respectively. 2. Both the HMOA and the HNOA have remained viable in culture for over 2 years and over 60 serial passages have been made. 3. The karyotypes are of human aneuploid (stem line was 46, 47), and have abnormal chromosomes. 4. These cell lines have an epithelial cell arrangement with a monolayer system, no contact inhibition, and multilayered growth. Metaplasia can be observed with HMOA. 5. A heterotransplantation can be performed with these cell lines with the HMOA forming an adenoacanthoma (well-differentiated type), and the HNOA forming a poorly differentiated endometrioid adenocarcinoma. As for the human ovarian endometrioid adenocarcinoma cell line, three such cell lines have hitherto been established [4,9,13]. All of these cell lines are moderately differentiated ovarian endometrioid adenocarcinoma (containing no squamous foci) which have been established using ascites as the culture material. The tumors subsequently developing following a heterotransplantation of these cell lines are all poorly differentiated endometrioid adenocarcinomas. The cells
106
ISHIWATA
ET AL.
of each cell line contain a specific marker chromosome. The HNOA is a cell line that has been established using the primary tumor (mouse graft) of a poorly differentiated endometrioid adenocarcinoma, and the cellular characteristics bear similar resemblance to those of the three cell lines. On the contrary, the HMOA is a cell line established using the primary tumor of a well-differentiated ovarian endometrioid adenoacanthoma, forms cysts in monolayer culture, and shows metaplasia. Since a large quantity of fibrilar bundles and/or tonofilaments and glycogen granules are contained in the cytoplasm of the polygonal squamous cells of the HMOA line, these polygonal cells are probably responsible for the observed metaplasia. The tumor that developed through a heterotransplantion of the HMOA is an adenoacanthoma closely resembling the original tumor. All these findings appear to suggest that the endometrioid adenocarcinoma cell retains the capability to differentiate into a squamous cell, even in vitro. There are at least two separate mechanisms responsible for the existence of squamous foci within an adenocarcinoma, namely (1) simultaneous tumor developments from separate cells of different origin (glandular cell and the squamous cell), and (2) squamous metaplasia of an adenocarcinoma. If the cultures were to be continued indefinitely, the adenocarcinoma and squamous cell would probably not remain viable with mixed growth. In cases of developing a tumor through heterotransplantation, the chances are remote that the normal squamous cell could successfully be heterotransplanted. Moreover, the tumor which developed following heterotransplantation of the cloned cell line we established through the culture of a single cell was an adenoacanthoma. All of these facts strongly support the thesis that the adenoacanthoma could have developed through squamous metaplasia of the endometrioid adenocarcinoma cells. The CA125 is a new type of an antigenic determinant specific for ovarian cancer that can be found using a monoclonal antibody (OC125) [ 11. The OC125 was obtained from mice immunized with an ovarian cystadenocarcinoma cell (OVCA433) by Bast et al. [1,2]. An interesting finding in a culture of the HMOA line is that the CA125 was produced in a large quantity when the endometrioid adenocarcinoma cells differentiated into the glandular cells and in small quantities when the above cells differentiated more into the squamous cells through metaplasia. These findings suggested that the CA125 could have been produced primarily from the adenocarcinoma in the same organ. It has been reported that CA125 is produced in a far larger quantity by cervical adenocarcinoma than by cervical epidermoid carcinoma [ 101. The HNOA is a cell line established from an ovarian endometrioid adenocarcinema, and the biological characteristics bear resemblance to those of the ovarian endometrioid carcinoma cell line established elsewhere [4,9,13]. On the other hand, the HMOA is a cell line established from an ovarian endometrioid adenoacanthoma, and shows characteristics of adenoacanthoma, both in vivo and in vitro. The mechanisms related to squamous metaplasia are poorly understood. This HMOA line should prove to be an important tool for investigating the mechanism of metaplasia in cases of adenoacanthoma, as well as the mechanism for production/secretion of the CA125, at the cellular level.
HISTOGENESIS,
ADENOACANTHOMA,
CELL
LINE
107
ACKNOWLEDGMENTS We thank Dr. I. Kondo, Division of Human Genetics, Institute of Basic Medical Science, Tsukuba University, for help in the karyotyping and M. Ohara of Kyushu University for critical comments on the manuscript. This study was supported in part by Grant-in-Aid for Scientific Research 59570016 from the Ministry of Education, Japan.
REFERENCES 1. Bast, R. C., Feeney, M., Iazarus, H., Nadler, L. M., Colvin, P. B., and Knapp, R. C. Reactivity of a monoclonal antibody with human ovarian carcinoma, J. C/in. Invest. 68, 1331-1337 (1981). 2. Bast, R. C., Klug, T. L., John, E. T., Jenison, E., Niloff, J. M. Lazarus, H., Berkowitz, R. S., Leavitt, T., Griffiths, C. T., Parker, L., Zurawski, V. R., and Knapp, R. C. A radioimmunoassay using a monoclonal antibody to monitor the course of epithelial ovarian cancer, N. Engl. J. Med. 309, 883-887 (1983).
3. Czernobilsky, B. Primary epithelial tumors of the ovary, in Purhology of the female genital tract (A. Blaustein, Ed.), Springer-Verlag, New York/Heidelberg/Berlin, pp. 532-538 (1982). 4. Freedman, R. S., Pihl, E., Kusyk, C., Gallager, H. S., and Rutledge, F. Characterization of an ovarian carcinoma cell line, Cancer 42, 2352-2359 (1978). 5. Fu, Y. S., Stock, P. J., Reagan, J. W., Storaasli, J. P., and Wentz, W. B. Significance of squamous components in endometrioid carcinoma of the ovary, Cancer 44, 616-621 (1979). 6. Ishiwata, I., Ishiwata, C., Soma, M., Kobayaski, N., and Ishikawa, H. Establishment and characterization of an estrogen-producing human ovarian granulosa tumor cell line, J. Null. Cancer Inst. 72, 789-800 (1984). 7. Ishiwata, I., Nozawa, S., Inoue, T., and Okumura, H. Development and characterization of established cell lines from primary and metastatic regions of human endometrial adenocarcinoma, Cancer Res. 37, 1777-1785 (1977). 8. Klug, T. L., Bast, R. C., Niloff, J. M., Knapp, R. C., Zurawski, V. R. Monoclonal antibody immunoradiometric assay for an antigenic determinant (Ca125) associated with human epithelial ovarian carcinoma, Cancer Res. 44, 1048-1053 (1984). 9. Nagai, S., Nozawa, S., Kurihara, S., and Mukai, M. Cytologic and biologic studies of endometrioid carcinoma of the ovary, Actu Cytol. 27, 676-682 (1983). 10. Niloff, J. M., Klug, T. L., Schaetzl, E., Zurawski, V. R., Knapp, R. C., and Bast, R. C. Elevation of serum Cal25 in carcinomas of the Fallopian tube, endometrium, and endocervix, Amer. /. Obstet. Gynecol.
148, 1057-1058 (1984).
11. Nishi, S., and Hirai, H. A new radioimmunoassay
of alpha-fetoprotein and curcinoembryonic of the biologicaljuid, Pergamon, Vol. 23, pp. 303-307 (1976). 12. Scully, R. E. Endometrioid tumors, in Tumors of the ovary and maldeveloped gonads (W. H. antigen:
The 23rd proteides
Hartmann, Ed.), the Armed Forces Institute of Pathology, Washington, D.C., pp. 92-l 10 (1982).
13. Uehara, S., Soh, K., Hoshiai, H., Yajima, A., Suzuki, M., and Abe, H. Establishment and characterization of a human ovarian endometrioid carcinoma cell line, Actu Obstet. Gynuecol. Japan. 35, 19-26 (1983).
ONCOLOGY
25, 95-107 (1986)
Establishment and Characterization of Two Human Ovarian Endometrioid Carcinoma Cell Lines (with or without Squamous Cell Component) ISAMU ISHIWATA,
M.D., *,t CHIEKO ISHIWATA, M.D.,* MASAYUKI SOMA, C.T.,* AND HIROSHI ISHIKAWA, M.D.7
*Ishiwata Obstetrics and Gynecologic Hospital, Mite-shi, Ibaraki-ken 310, and i-Department of Anatomy, Jikei University, School of Medicine, Minato-ku, Tokyo 105, Japan Received May 20, 1985 The cell lines designated HMOA and HNOA were established from human ovarian adenoacanthoma and from a mouse graft of human ovarian endometrioid adenocarcinoma, respectively. These cell lines grew well without interruption for over 20 months. The cultured cells of both HMOA and HNOA lines were spindle, polygonal, and columnar, and showed a jigsaw puzzle-like arrangement and a piling-up tendency devoid of contact inhibition. When the HMOA cells were maintained at the confluent stage, the cells formed cysts and/or squamous metaplasia. The chromosome number of both cell lines varied widely and showed aneuploidy, while the modal chromosome number was stable at the diploid range. Both of these cell lines, HMOA and HNOA, were trdnsplanted into the subcutis of BALB/c nude mice and produced well-differentiated adenoacanthoma and poorly differentiated endometrioid adenocarcinoma, respectively. HMOA cells were characterized as producing large amounts of CA125 (ovarian carcinoma marker), in vitro, in the cyst-forming phase. The HNOA cells, however, did not produce CA125. ‘a 1986 Academic Press, Inc.
INTRODUCTION Human ovarian endometrioid adenocarcinoma may develop histogenetically from ovarian endometriosis, or may originate from the ovarian surface layer of mesothelium [3,12]. On the other hand, squamous epithelial foci often coexist with a well-differentiated ovarian endometrioid adenocarcinoma. According to its morphologically benign or malignant characteristics, it is called an endometrioid adenoacanthoma or an endometrioid adenosquamous carcinoma, respectively
[5,121. Recently, we have cultured an ovarian endometrioid adenocarcinoma that did not contain squamous epithelial foci, and an ovarian endometrioid adenoacanthoma that did contain a squamous epithelial foci. We then established a cell line of each culture (HNOA and HMOA cell line, respectively). We report the methods we used to establish the above cell lines, as well as the biological characteristics ’ To whom requests for reprints should be addressed at I-4-21, Kami-Mito, Mito-shi, Ibaraki-ken 310, Japan 95 0090-8258186$1.50 Copyright Q 1986 by Academic Press, Inc. All rights of reproductmn in any Form reserved.
96
ISHIWATA
ET AL.
of these lines. We found that (1) the formation of squamous epithelial foci of an adenoacanthoma is caused by squamous metaplasia of the adenocarcinoma, (2) the HMOA is a cell line that produced the antigenic determinant CA125 [1,8] that can be defined by using the monoclonal antibody [OC125, obtained from mouse immunized with an human ovarian cystadenocarcinoma cell (OVCA-433)], and (3) the above CA125 was produced in a large quantity when a cyst and/or gland had formed, and in a small quantity when only squamous metaplasia occurred. MATERIALS AND METHODS I, Materials HMOA. The material (ovarian tumor) was obtained at radical hysterectomy and bilateral salpingo-oophorectomy in a 57-year-old Japanese woman on July 23, 1982. A portion of the tumor was placed in culture. HNOA. The material (ovarian tumor) was also obtained by the same operation in a 71-year-old Japanese woman on August 26, 1981. Eight 2-mm fragments were transplanted into the subcutis of three BALB/c nude mice Q-week-old female; Kurea Co., Tokyo) and the tumor measuring about 10 mm in diameter which developed in only one mouse was resected on December 9, 1981, and the mouse graft placed in culture. 2. Culture
Technique
and Culture Media
The materials for both HMOA and HNOA lines were rinsed twice with the growth medium, minced with scissors, and dissociated in 600 pronase Units/ml of dispase (Gohdo Shusei Co., Tokyo) for 30 min at room temperature and finally centrifuged at 1000 rpm for 10 min. The sediments were resuspended in the growth medium, placed on the 6-cm plastic dishes (Terumo Co., Tokyo) and cultured at 37°C in the 5% CO, incubator. The growth medium used was Ham’s F-12 (Grand Island Biological Co., Grand Island, N. Y.) supplemented with 15% fetal calf serum (Flow Laboratories, Beltsville, Md.), 40 pg kanamycin/ml, pH 7.2 to 7.4. In the initiation of the primary culture, epithelial colonies and fibroblast sheets appeared in both the HMOA and HNOA cultures. The pure epithelial cells were obtained by colony isolation technique (filter papers soaked with a mixed solution of 0.1% trypsin and 0.02% EDTA were placed directly on the epithelial cells; the papers together with the cells were shaken off into new plastic dishes [ 171). 3. Histology
The tumor was fixed with 10% buffered Formalin, embedded in paraffin, cut into thicknesses of 4 pm, and stained with hematoxylin and eosin, mucicarmine, alcian blue, periodic acid-Schiff, with or without 1% diastase, for 10 min. The cultured cells were also stained with Papanicolaou, Giemsa, periodic acid-Schiff, mucicarmine, and alcian blue. The grafts in nude mice and the cultured cells were observed with the JEM 1OOBelectron microscope. Details of the procedures can be found elsewhere [61.
HISTOGENESIS,
ADENOACANTHOMA,
CELL
LINE
97
4. Growth Characteristics Growth curves, population doubling times, saturation density, and plating efficiency were studied at the 10th and 40th passage. Approximately 1 x IO5 cells were placed into 3.5cm plastic dishes and cultured for approximately 10 days. Cells in three dishes were counted using an Erma counting chamber (Erma Co., Tokyo) and the average number was plotted each day. Population doubling time and saturation density were determined using a growth curve [6]. The medium was changed every 2 days. To study the plating efficiency, 250 single suspended cells were placed into five plastic dishes (6 cm in diameter), cultured for about 14 days, and finally stained with Giemsa. Plating efficiency was determined as the ratio of the number of visible colonies (more than 10 cells) to the number of the inoculated cells. 4. Chromosomal Analysis Chromosomal analysis was done at the 5th and 20th passages. The cultured cells were treated with 0.1% trypsin solution for 15 set at room temperature, stained with 3% Giemsa, and analyzed with G-band karyotyping [7]. Histograms of chromosome number distribution were determined on more than 100 metaphase plates. 5. CA125 Assay Approximately 5 x 10’ cells of HMOA and HNOA at passage 20 were cultured respectively for 2 days and the CA125 level in the conditioned media was measured using a CA125 RIA KIT (monoclonal antibody; Centocor) with the sandwich methods of radioimmunoassay [II]. The cellular CA125-producing activity was determined as the amount of CA125 produced by one cell for 2 days. Triplicate assays were performed to determine the CA125 levels. 6. Heterotransplantation Approximately 1 x 106, I x IO’, and 1 x IO8HMOA and HNOA cells (both at passage 20) were transplanted respectively into the subcutis of 2 BALB/c nude mice (8 week-old female; Kurea). The tumors were examined histologically by light microscopy 2 months after transplantation. The serum CA125 was measured by radioimmunoassay at the same time. RESULTS
I. Pathology of Original Tumors The original ovarian tumor for the HMOA line was interpreted pathologically to be an endometrioid adenoacanthoma composed of well-differentiated endometrioid adenocarcinoma with benign squamous foci (Fig. 1). The cytoplasms of squamous foci were strongly stained with periodic acid-SchifI-digested diastase and negatively with mucicarmine and alcian blue. On the contrary, the cytoplasm of the glandular tissue was slightly stained with mucicarmine and periodic acidSchiff. The material (mouse graft) of the HNOA cell line was interpreted to be a poorly differentiated endometrioid adenocarcinoma. The original tumor (patient’s
98
ISHIWATA
ET AL.
FIG. 1. Histology of the original human ovarian tumor of HMOA line revealed an adenoacanthoma composed of well-differentiated endometrioid adenocarcinoma and squamous metaplasia. Hematoxylineosin stain, x 100.
ovarian tumor) was considered to be a moderately differentiated endometrioid adenocarcinoma (Fig. 2). 2. Morphology
of Cultured
Cells.
Both HMOA and HNOA (Fig. 3) cells were spindle, polygonal, and columnar in shape and had an epithelial cell arrangement (jigsaw puzzle-like). They revealed a piling-up tendency without contact inhibition and showed anaplastic and pleomorphic features. When HMOA cells were maintained in the confluent condition they formed various sized cysts or showed squamous metaplasia (Fig. 4). Squamous metaplastic cells reacted strongly when stained with periodic acid-Schiff. On the contrary, the cells in cyst formation stained only slightly with mucicarmine and periodic acid-Schiff. Electron microscopy revealed that the HMOA cells were dark and had numerous short microvilli on the free surface. They made contact with each other by numerous desmosomes. Metaplastic cells contained a number of bundles of tonofibrils and glycogen granules scattered throughout the cytoplasm. The rough endoplasmic reticulum was poorly developed. The nuclear shape was irregular and a large nucleolus was present (Fig. 5). Another type of HMOA cell and HNOA cell also had a large nucleolus in an irregularly shaped nucleus. There were, however, many microtubules, small vesicles, moderately developed lamellary rough-endoplasmic reticulum, oval-shaped mitochondria, and a few secretory-like granules in the cytoplasm (Fig. 6).
HISTOGENESIS,
ADENOACANTHOMA,
CELL LINE
FIG. 2. Histology of the original human ovarian tumor of HNOA line revealed a moderaltely differentiated endometrioid adenocarcinoma. Hematoxylin-eosin stain, x 100.
FIG. 3. The monolayer-cultured cells of passage 5 (HNOA line) were spindle, polygonal, and columnar, and revealed neoplastic and pleomorphic features. Papanicolaou stain, x 100.
100
ISHIWATA ET AL.
FIG. 4. (a) The cultured cells of passage 5 (HMOA line) formed cysts or (b) showed squamous metaplasia. Phase contrast; (a) x40, (b) x200.
FIG. 5. Some HMOA cells (passage 5) were dark cells and had a number of the bundles of tonotibrils and made contact by desmosomes (arrow). This cell was considered a metaplastic cell. Transmission electron microscopy x 7900.
HISTOGENESIS,
ADENOACANTHOMA,
CELL LINE
101
FIG. 6. The HNOA cells (passage 5) had many microtubules, small vesicles, moderately developed lamellary rough-endoplasmic reticulum and a few secretory-like granules (arrow) in the cytoplasm. Transmission electron microscopy x 7900.
3. Growth Characteristics The growth curves are shown in Fig. 7. The growth characteristics such as population doubling time, saturation density, and plating efficiency are shown in Table 1. The population doubling time was 72 hr and plating efficiency was about 20% in the case of HMOA and 28 hr and 65%, respectively, in the case of HNOA.
o.l!
, 12
3 culture
4
5
6 7 days
8910
FIG. 7. The growth curve of the cultured cells at the 10th (0) and 40th (0) passage of HMOA line and that of the cultured cells at the 10th (X) and 40th (0) passage of HNOA line.
102
ISHIWATA
GROWTH
Cell line
PN
HMOA
10 40 10 40
HNOA
CHARACTERISTICS
ET AL.
TABLE 1 OF HMOA
AND
HNOA
CELL LINES
SD (1 x 104/cmZ) 72 58 28 24
5.2 5.6 5.9 6.0
20 25 65 72
Note. PN, passage number; DT, population doubling time; SD, saturation density; PE, plating efficiency.
4. Chromosome
Analysis
Karyotype of both HMOA and HNOA showed them clearly to be human cells. The chromsome number of both HMOA and HNOA was widely distributed and showed aneuploidy. However, the modal chromosome number was stable at the diploid range (Fig. 8). The HMOA line has a marker chromosome and the HNOA line is without a marker chromosome. The abnormal chromosome (2p-) was detected in about 75% of the cells (Fig. 9). 5. CA125 Assay
The CA125 level in the conditioned media in which 5 x lo5 cells of HMOA (at passage 20) were cultured for 2 days was 1140 U/ml. The CA125 level in the culture media (cell free) and also in the case of HNOA (at passage 20) was less than 8 U/ml. Therefore, the HMOA is considered to be a CA125-producing cell line, while the HNOA line is a CA125-nonproducing cell line. The amount of CA125 produced by one cell of the HMOA line for 2 days was 2.24 mU during HMOA
passage
HMOA
passage20
HNOA
passage
HNOA
passage20
chromosomal
5
5
number
8. Distribution of the chromosomal number of passages 5 and 20 of both HMOA and HNOA lines shows that the modal chromosomal number was 46-47. FIG.
HISTOGENESIS,
ADENOACANTHOMA,
CELL
103
LINE
f
M 19,:
1
2P-
6
7
13
14
4
3
8
15
9
10
16
5
11
12
18
21 22 19 20 xx FIG. 9. G-band karyotyping of passage 20 of HMOA and HNOA cells. That of HMOA shows 46, XXq - , - 13,Sq+ (1lp - ,19q + , marker chromosome. The telocentric marker chromosome was detected in lOO%, and that of HNOA shows 46,XX,2p-. 2P- was shown in 75% of HNOA cells.
monolayer culture, that in cyst formation was 3.8 mu, while that in metaplasia was 1.63 mu. 6. Heterotransplantation
Heterotransplantability of both HMOA and HNOA is shown in Table 2. The tumors produced by HMOA were interpreted to be an adenocarcinoma (Fig. 10)
104
ISHIWATA
ET AL.
TABLE 2 ONCOGENECITYOF HMOA LINE (PASSAGE20) AND HNOA LINE (PASSAGE20) IN NUDE MICE
No. of mice Cell line HMOA
HNOA
Inoculum size
Injected
With tumors
1 x lo6 1 x IO’ 1 x 108
2 2 2
0 1 2
1 x lo6 1 x 10’
2 2
2 2
lx
2
2
108
Tumor size (mm) 8 x 8 x 12 x 11 x 12 x 12 x
Pathology Well-differentiated endometrioid adenocarcinoma
10x 9x 9I 10 .-:-. 1 l4 x’ 10 l3 x’ lo9 Poor,y diff-... erenuatcu 15 x 13 x 11 endometricrid adenocarcinoma 15 x 12 x 10 16 x 14 x 11
CA125 levels in sera (U/ml) 276 290 304 <8 <8 <8 <8 <8 <8
and resembled the original tumor. On the contrary, that produced by HNOA was a poorly differentiated endometrioid adenocarcinoma (Fig. 11). The CA125 level in mouse sera was also measured by radioimmunoassay. The average CA125 level in the sera of three mice with HMOA tumor was 290 U/ml, while the level in case of HNOA was less than 8 U/ml, as illustrated in Table 2.
FIG. 10. The histology of the tumor produced by heterotransplantation of passage 20 cells of HMOA into nude mouse reveals an adenoacanthoma. Hematoxylin-eosin stain; (a) x250, (b) x 100.
HISTOGENESIS,
ADENOACANTHOMA,
CELL
LINE
FIG. 11. The histology of the tumor produced by heterotransplantation of passage 20 cells of HNOA into nude mouse reveals a poorly differentiated endometrioid adenocarcinoma. Hematoxylineosin stain, x 10. Inset: the area enclosed by the rectangle of the low magnification micrograph x 100.
DISCUSSION The HMOA and HNOA cell lines were proven to be an human ovarian endometrioid adenoacanthoma cell line, and an ovarian endometrioid adenocarcinoma cell line, respectively, based on the following observations. 1. The materials cultured were from an ovarian endometrioid adenoacanthoma and an ovarian endometrioid adenocarcinoma, respectively. 2. Both the HMOA and the HNOA have remained viable in culture for over 2 years and over 60 serial passages have been made. 3. The karyotypes are of human aneuploid (stem line was 46, 47), and have abnormal chromosomes. 4. These cell lines have an epithelial cell arrangement with a monolayer system, no contact inhibition, and multilayered growth. Metaplasia can be observed with HMOA. 5. A heterotransplantation can be performed with these cell lines with the HMOA forming an adenoacanthoma (well-differentiated type), and the HNOA forming a poorly differentiated endometrioid adenocarcinoma. As for the human ovarian endometrioid adenocarcinoma cell line, three such cell lines have hitherto been established [4,9,13]. All of these cell lines are moderately differentiated ovarian endometrioid adenocarcinoma (containing no squamous foci) which have been established using ascites as the culture material. The tumors subsequently developing following a heterotransplantation of these cell lines are all poorly differentiated endometrioid adenocarcinomas. The cells
106
ISHIWATA
ET AL.
of each cell line contain a specific marker chromosome. The HNOA is a cell line that has been established using the primary tumor (mouse graft) of a poorly differentiated endometrioid adenocarcinoma, and the cellular characteristics bear similar resemblance to those of the three cell lines. On the contrary, the HMOA is a cell line established using the primary tumor of a well-differentiated ovarian endometrioid adenoacanthoma, forms cysts in monolayer culture, and shows metaplasia. Since a large quantity of fibrilar bundles and/or tonofilaments and glycogen granules are contained in the cytoplasm of the polygonal squamous cells of the HMOA line, these polygonal cells are probably responsible for the observed metaplasia. The tumor that developed through a heterotransplantion of the HMOA is an adenoacanthoma closely resembling the original tumor. All these findings appear to suggest that the endometrioid adenocarcinoma cell retains the capability to differentiate into a squamous cell, even in vitro. There are at least two separate mechanisms responsible for the existence of squamous foci within an adenocarcinoma, namely (1) simultaneous tumor developments from separate cells of different origin (glandular cell and the squamous cell), and (2) squamous metaplasia of an adenocarcinoma. If the cultures were to be continued indefinitely, the adenocarcinoma and squamous cell would probably not remain viable with mixed growth. In cases of developing a tumor through heterotransplantation, the chances are remote that the normal squamous cell could successfully be heterotransplanted. Moreover, the tumor which developed following heterotransplantation of the cloned cell line we established through the culture of a single cell was an adenoacanthoma. All of these facts strongly support the thesis that the adenoacanthoma could have developed through squamous metaplasia of the endometrioid adenocarcinoma cells. The CA125 is a new type of an antigenic determinant specific for ovarian cancer that can be found using a monoclonal antibody (OC125) [ 11. The OC125 was obtained from mice immunized with an ovarian cystadenocarcinoma cell (OVCA433) by Bast et al. [1,2]. An interesting finding in a culture of the HMOA line is that the CA125 was produced in a large quantity when the endometrioid adenocarcinoma cells differentiated into the glandular cells and in small quantities when the above cells differentiated more into the squamous cells through metaplasia. These findings suggested that the CA125 could have been produced primarily from the adenocarcinoma in the same organ. It has been reported that CA125 is produced in a far larger quantity by cervical adenocarcinoma than by cervical epidermoid carcinoma [ 101. The HNOA is a cell line established from an ovarian endometrioid adenocarcinema, and the biological characteristics bear resemblance to those of the ovarian endometrioid carcinoma cell line established elsewhere [4,9,13]. On the other hand, the HMOA is a cell line established from an ovarian endometrioid adenoacanthoma, and shows characteristics of adenoacanthoma, both in vivo and in vitro. The mechanisms related to squamous metaplasia are poorly understood. This HMOA line should prove to be an important tool for investigating the mechanism of metaplasia in cases of adenoacanthoma, as well as the mechanism for production/secretion of the CA125, at the cellular level.
HISTOGENESIS,
ADENOACANTHOMA,
CELL
LINE
107
ACKNOWLEDGMENTS We thank Dr. I. Kondo, Division of Human Genetics, Institute of Basic Medical Science, Tsukuba University, for help in the karyotyping and M. Ohara of Kyushu University for critical comments on the manuscript. This study was supported in part by Grant-in-Aid for Scientific Research 59570016 from the Ministry of Education, Japan.
REFERENCES 1. Bast, R. C., Feeney, M., Iazarus, H., Nadler, L. M., Colvin, P. B., and Knapp, R. C. Reactivity of a monoclonal antibody with human ovarian carcinoma, J. C/in. Invest. 68, 1331-1337 (1981). 2. Bast, R. C., Klug, T. L., John, E. T., Jenison, E., Niloff, J. M. Lazarus, H., Berkowitz, R. S., Leavitt, T., Griffiths, C. T., Parker, L., Zurawski, V. R., and Knapp, R. C. A radioimmunoassay using a monoclonal antibody to monitor the course of epithelial ovarian cancer, N. Engl. J. Med. 309, 883-887 (1983).
3. Czernobilsky, B. Primary epithelial tumors of the ovary, in Purhology of the female genital tract (A. Blaustein, Ed.), Springer-Verlag, New York/Heidelberg/Berlin, pp. 532-538 (1982). 4. Freedman, R. S., Pihl, E., Kusyk, C., Gallager, H. S., and Rutledge, F. Characterization of an ovarian carcinoma cell line, Cancer 42, 2352-2359 (1978). 5. Fu, Y. S., Stock, P. J., Reagan, J. W., Storaasli, J. P., and Wentz, W. B. Significance of squamous components in endometrioid carcinoma of the ovary, Cancer 44, 616-621 (1979). 6. Ishiwata, I., Ishiwata, C., Soma, M., Kobayaski, N., and Ishikawa, H. Establishment and characterization of an estrogen-producing human ovarian granulosa tumor cell line, J. Null. Cancer Inst. 72, 789-800 (1984). 7. Ishiwata, I., Nozawa, S., Inoue, T., and Okumura, H. Development and characterization of established cell lines from primary and metastatic regions of human endometrial adenocarcinoma, Cancer Res. 37, 1777-1785 (1977). 8. Klug, T. L., Bast, R. C., Niloff, J. M., Knapp, R. C., Zurawski, V. R. Monoclonal antibody immunoradiometric assay for an antigenic determinant (Ca125) associated with human epithelial ovarian carcinoma, Cancer Res. 44, 1048-1053 (1984). 9. Nagai, S., Nozawa, S., Kurihara, S., and Mukai, M. Cytologic and biologic studies of endometrioid carcinoma of the ovary, Actu Cytol. 27, 676-682 (1983). 10. Niloff, J. M., Klug, T. L., Schaetzl, E., Zurawski, V. R., Knapp, R. C., and Bast, R. C. Elevation of serum Cal25 in carcinomas of the Fallopian tube, endometrium, and endocervix, Amer. /. Obstet. Gynecol.
148, 1057-1058 (1984).
11. Nishi, S., and Hirai, H. A new radioimmunoassay
of alpha-fetoprotein and curcinoembryonic of the biologicaljuid, Pergamon, Vol. 23, pp. 303-307 (1976). 12. Scully, R. E. Endometrioid tumors, in Tumors of the ovary and maldeveloped gonads (W. H. antigen:
The 23rd proteides
Hartmann, Ed.), the Armed Forces Institute of Pathology, Washington, D.C., pp. 92-l 10 (1982).
13. Uehara, S., Soh, K., Hoshiai, H., Yajima, A., Suzuki, M., and Abe, H. Establishment and characterization of a human ovarian endometrioid carcinoma cell line, Actu Obstet. Gynuecol. Japan. 35, 19-26 (1983).