Journal of Surgical Oncology 44:260-267 (1990)
Characterization of Four Newly Established Human Colorectal Adenocarcinoma Cell Lines From Chinese Patients JHY-YOUNG CHENG, MD, DMSC, CHING-LIANG MENG, MD, DMSc, JIH-CHANG LIN, MD, CHING-CHERNG TZENG, MD, LI-TE CHIN, MS, AND KUO-LIANG SHEN, MD, PhD From the Department of Surgery, Section of Colorectal Surgery (I.-Y.C., ].-C.l., K.A.S.) and the Department of Medical Research (C.-L.M., C.-C.T., L.-T.C.), Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, Republic of China
Four colon adenocarcinoma cell lines, CC-M2, CC-M3, CC-M4, and CC-M2NM, have been established from surgical specimens of 18 unselected patients without the use of “feeder” cells and additional growth factors (e.g., insulin, hydrocortisone, etc.) in the culture medium. The methods of primary cultivation of tissue explants are described. Studies of determination of morphology, growth curve, plating efficiency, chromosomal analysis, CEA and beta-HCG synthesis, and tumorigenicity , were done to characterize the cell lines. Significant variations have been found in one of the four cell lines, both in vitro and in vivo studies. There are distinct phenotypes in the established cell lines which may be useful in studying the cell differentiation and progression of colorectal cancer. KEY WORDS: immortal cells, establishment, colon cancer
INTRODUCTION The incidence of colorectal cancer in Taiwan has increased rapidly in recent years. In order to facilitate the treatment of patients with colon cancer, studies of the biology of human colorectal cancer are necessary and urgent. Although many colon cell lines derived from colon carcinoma have been established in western countries [ 1-61, they are rare in Asia. Moreover, the more the cell lines can be established, the more the biological characteristics can be determined. Therefore, four colorectal adenocarcinoma cell lines derived from Chinese patients have been recently established in our laboratory in order to investigate geographic and ethnic variations of colorectal cancer abnormalities (71. There is evidence indicating the loss of requirements for specific growth factors which may lead to autologous growth of cancer cells [&lo]. Therefore, inappropriate supplement of growth factors may also induce cell transformation, and may not properly reflect molecular properties of original cancer cells [ 101. The four cell lines we hereby describe in this report were established and characterized in the culture medium containing no additional 0 1990 Wiley-Liss, Inc.
growth factors, i.e., insulin, hydrocortisone etc., which had usually been used in previous studies. They were directly derived from primary human colorectal adenocarcinomas, with one retaining well-differentiated morphology. The methods of successful cultivation and detailed characterization of these cell lines are thus presented.
MATERIALS AND METHODS Clinical Specimens Biopsies of 18 primary colorectal carcinomas were obtained at the time of exploratory surgery and colectomy. A representative portion of the surgical specimen was taken aseptically and immediately transported to the laboratory in cold (4°C) tissue culture medium for culti-
Accepted for publication April 19, 1990. Address reprint requests to Jhy Y. Cheng, M.D., DMSc, Immunology Laboratory, Department of Medical Research, Tri-Service General Hospital, Taipei, Taiwan, R.O.C., 10713.
Colorectal Cancer Cell Lines in Chinese
261
vation and processing. Another representative portion of the tissue was fixed, embedded, and stained.
sential medium) with 10% fetal bovine serum were also determined.
Cell Culture Materials The growth medium consisted of RPMI-1640 (Roswell Park Memorial Institute) with 10% FBS (fetal bovine serum) (Gibco-Europe Ltd., Paisley, Scotland), 50 pglml penicillin, 50 pg/ml streptomycin (Bristol Laboratories, Langley, Slough), amphotericin B 2.5 pg/ml (E.R. Squibb, Morton, Cheshire), 4 mM L-glutamine, and 2.5 mM sodium pyruvate (all from Grand Island Biological C o . , Grand Island, NY).
Plating and Seeding Efficiency Five milliliters of RPMI 10% fetal bovine serum culture medium with 1,250 cells of each cell line (passage 50) was seeded into a 25 cm2 flask. Colonies (16-50 cells/cluster) were counted to obtain plating efficiency . Duplicate culture flasks were seeded with 5 X 10' cells in 5 ml culture medium. Medium was then removed after a 24 hr culture, and the attached cells were counted by trypan blue exclusion test to determine seeding cfficiency [141.
Primary Culture and Subculture Procedures The tissue was cultured by using the tissue explant culture method 151. In brief, fresh colorectal carcinoma tissues were minced with apposite scalpels into 0.5 mm pieces and washed with washing medium which contained only RPMI-1640 and 2 X antibiotics. Then, 2 ml of culture medium was used to wet the tissue explants. Tissue fragments were simultaneously inoculated into the subcutaneous tissue of nude mice. Explants were incubated in a humidified atmosphere with 5 % CO, in air at 37°C. Multiple explants with cell outgrowth were evident during the early stage of cultivation and initial cell passage was not attempted until the 2nd to 4th week. Selective removal of fibroblasts and explants were attempted when the rate of epithelial outgrowth decreased [ 1 1 ) . Medium was changed twice weekly. Fibroblast removal was achieved by mechanical scraping under an Olyrnpus IMT phase contrast microscope, differential trypsinization with 0.025% trypsin (w/v) in Versene (Gibco-Europe Ltd., Paisley, Scotland) [ 121, or the toxicity of collagenase (25 U/ml, Grand Island Biological Co., Grand Island, NY) [ 131. When 70% of the bottom surface of the flask was covered by a heavy growth of cells, subculture could be attempted. Conditioned medium was used for initial subculture. The conditioned medium was made from the decanted culture medium while changing the medium in each flask. It was centrifuged with 250g, 10 min, at 4"C, distributed into small vials, and then stored in a refrigerator for usuage. The initial passage ratio was maintained no greater than 1.2. Passage frequency, ranging from 2 to 5 days, was controlled by the growth rate of cells.
Chromosomal Studies The cytogenetic studies of all cancer cells were undertaken at passage 50. G-banding karyotype was analysed with Giemsa stain according to the Yunis method [ 151. Screening for CEA and Beta-HCG Synthesis The supernatant of 10' cells which were incubated in 10 ml of RPMI 10% fetal bovine serum culture medium at 37°C for 21 days without refeeding was collected for CEA (carcinoembryonic antigen) and beta-HCG (betahuman chorionic gonadotropin) assays. Radioimmunoassay was performed with a Roche kit (Hoffmann-LaRoche Inc., Nutley, NJ). Tumorigenicity Cells were subcutaneously injected into bilateral flanks of six adult female BALB/C nude mice. Approximately 5 x 10' cells were injected per site. In 1 to 2 months, the newly grown tumors were removed and representative sections were stained with hematoxylin-eosin or Alcian blue (pH:2,5)-periodic acid-Schiff stain (ABPAS). A portion of the tumor tissue was also processed for electron microscopy.
Electron Microscopy Tumor fragments were fixed in 3% glutaraldehyde in 0.1 M phosphate buffer at 4°C for 3 days. After fixation in 2% osmium tetroxide, dehydrating with graded ethanol, infiltrating with propylene oxide, and embedding in Epon 812, tissue fragments were cut into 80-nm sections, then stained with lead citrate and uranyl acetate Growth Curve and examined with a Zeiss Model EM109 electron miA concentration of 3 X lo4 cells per ml of RPMI 10% croscope at 80 kV [ 161. fetal bovine serum was cultivated in each well of the Cytokeratin Detection 24-well culture plates (Costar, Cambridge, MA), and the The modified immunofluorescent method of Debus et number was counted daily by hemocytometer (Reichert, Buffalo, NY) for 6 days without refeeding. Population al. [ 171 was administered for detection of cytokeratin in doubling time and growth ability in the RPMI 0.5%, 2%, cancer cells by mouse monoclonal anti-cytokeratin 8.13 5 % fetal bovine serum and MEM (Eagle's minimum es- antibody (Sigma Chemical Co., St. Louis, MO).
262
Cheng et al.
TABLE I. Characteristics of CC-M Cell Lines* Cell lines Characteristics
CC-M2
CC-M3
CC-M4
CC-M2NM -
Age (years) Sex Site Histological grade Modified Dukes’ stage Population doubling time Passage 25 Passage 50 Efficiency Plating (5%) Seeding Synthesis CEA (ngiml) beta-HCG (mIUlml) Tuniorigenicity Passage 5 Passage 50
58 M SC” MDb B1 (hr) 20.0 18.5
80 F TC MD D
92 M R MD B2
16.6 16.1
11.2 16.8
64.5 51.4
34 81
25 16
22 68
Characterization of Four Newly Established Human Colorectal Adenocarcinoma Cell Lines From Chinese Patients JHY-YOUNG CHENG, MD, DMSC, CHING-LIANG MENG, MD, DMSc, JIH-CHANG LIN, MD, CHING-CHERNG TZENG, MD, LI-TE CHIN, MS, AND KUO-LIANG SHEN, MD, PhD From the Department of Surgery, Section of Colorectal Surgery (I.-Y.C., ].-C.l., K.A.S.) and the Department of Medical Research (C.-L.M., C.-C.T., L.-T.C.), Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, Republic of China
Four colon adenocarcinoma cell lines, CC-M2, CC-M3, CC-M4, and CC-M2NM, have been established from surgical specimens of 18 unselected patients without the use of “feeder” cells and additional growth factors (e.g., insulin, hydrocortisone, etc.) in the culture medium. The methods of primary cultivation of tissue explants are described. Studies of determination of morphology, growth curve, plating efficiency, chromosomal analysis, CEA and beta-HCG synthesis, and tumorigenicity , were done to characterize the cell lines. Significant variations have been found in one of the four cell lines, both in vitro and in vivo studies. There are distinct phenotypes in the established cell lines which may be useful in studying the cell differentiation and progression of colorectal cancer. KEY WORDS: immortal cells, establishment, colon cancer
INTRODUCTION The incidence of colorectal cancer in Taiwan has increased rapidly in recent years. In order to facilitate the treatment of patients with colon cancer, studies of the biology of human colorectal cancer are necessary and urgent. Although many colon cell lines derived from colon carcinoma have been established in western countries [ 1-61, they are rare in Asia. Moreover, the more the cell lines can be established, the more the biological characteristics can be determined. Therefore, four colorectal adenocarcinoma cell lines derived from Chinese patients have been recently established in our laboratory in order to investigate geographic and ethnic variations of colorectal cancer abnormalities (71. There is evidence indicating the loss of requirements for specific growth factors which may lead to autologous growth of cancer cells [&lo]. Therefore, inappropriate supplement of growth factors may also induce cell transformation, and may not properly reflect molecular properties of original cancer cells [ 101. The four cell lines we hereby describe in this report were established and characterized in the culture medium containing no additional 0 1990 Wiley-Liss, Inc.
growth factors, i.e., insulin, hydrocortisone etc., which had usually been used in previous studies. They were directly derived from primary human colorectal adenocarcinomas, with one retaining well-differentiated morphology. The methods of successful cultivation and detailed characterization of these cell lines are thus presented.
MATERIALS AND METHODS Clinical Specimens Biopsies of 18 primary colorectal carcinomas were obtained at the time of exploratory surgery and colectomy. A representative portion of the surgical specimen was taken aseptically and immediately transported to the laboratory in cold (4°C) tissue culture medium for culti-
Accepted for publication April 19, 1990. Address reprint requests to Jhy Y. Cheng, M.D., DMSc, Immunology Laboratory, Department of Medical Research, Tri-Service General Hospital, Taipei, Taiwan, R.O.C., 10713.
Colorectal Cancer Cell Lines in Chinese
261
vation and processing. Another representative portion of the tissue was fixed, embedded, and stained.
sential medium) with 10% fetal bovine serum were also determined.
Cell Culture Materials The growth medium consisted of RPMI-1640 (Roswell Park Memorial Institute) with 10% FBS (fetal bovine serum) (Gibco-Europe Ltd., Paisley, Scotland), 50 pglml penicillin, 50 pg/ml streptomycin (Bristol Laboratories, Langley, Slough), amphotericin B 2.5 pg/ml (E.R. Squibb, Morton, Cheshire), 4 mM L-glutamine, and 2.5 mM sodium pyruvate (all from Grand Island Biological C o . , Grand Island, NY).
Plating and Seeding Efficiency Five milliliters of RPMI 10% fetal bovine serum culture medium with 1,250 cells of each cell line (passage 50) was seeded into a 25 cm2 flask. Colonies (16-50 cells/cluster) were counted to obtain plating efficiency . Duplicate culture flasks were seeded with 5 X 10' cells in 5 ml culture medium. Medium was then removed after a 24 hr culture, and the attached cells were counted by trypan blue exclusion test to determine seeding cfficiency [141.
Primary Culture and Subculture Procedures The tissue was cultured by using the tissue explant culture method 151. In brief, fresh colorectal carcinoma tissues were minced with apposite scalpels into 0.5 mm pieces and washed with washing medium which contained only RPMI-1640 and 2 X antibiotics. Then, 2 ml of culture medium was used to wet the tissue explants. Tissue fragments were simultaneously inoculated into the subcutaneous tissue of nude mice. Explants were incubated in a humidified atmosphere with 5 % CO, in air at 37°C. Multiple explants with cell outgrowth were evident during the early stage of cultivation and initial cell passage was not attempted until the 2nd to 4th week. Selective removal of fibroblasts and explants were attempted when the rate of epithelial outgrowth decreased [ 1 1 ) . Medium was changed twice weekly. Fibroblast removal was achieved by mechanical scraping under an Olyrnpus IMT phase contrast microscope, differential trypsinization with 0.025% trypsin (w/v) in Versene (Gibco-Europe Ltd., Paisley, Scotland) [ 121, or the toxicity of collagenase (25 U/ml, Grand Island Biological Co., Grand Island, NY) [ 131. When 70% of the bottom surface of the flask was covered by a heavy growth of cells, subculture could be attempted. Conditioned medium was used for initial subculture. The conditioned medium was made from the decanted culture medium while changing the medium in each flask. It was centrifuged with 250g, 10 min, at 4"C, distributed into small vials, and then stored in a refrigerator for usuage. The initial passage ratio was maintained no greater than 1.2. Passage frequency, ranging from 2 to 5 days, was controlled by the growth rate of cells.
Chromosomal Studies The cytogenetic studies of all cancer cells were undertaken at passage 50. G-banding karyotype was analysed with Giemsa stain according to the Yunis method [ 151. Screening for CEA and Beta-HCG Synthesis The supernatant of 10' cells which were incubated in 10 ml of RPMI 10% fetal bovine serum culture medium at 37°C for 21 days without refeeding was collected for CEA (carcinoembryonic antigen) and beta-HCG (betahuman chorionic gonadotropin) assays. Radioimmunoassay was performed with a Roche kit (Hoffmann-LaRoche Inc., Nutley, NJ). Tumorigenicity Cells were subcutaneously injected into bilateral flanks of six adult female BALB/C nude mice. Approximately 5 x 10' cells were injected per site. In 1 to 2 months, the newly grown tumors were removed and representative sections were stained with hematoxylin-eosin or Alcian blue (pH:2,5)-periodic acid-Schiff stain (ABPAS). A portion of the tumor tissue was also processed for electron microscopy.
Electron Microscopy Tumor fragments were fixed in 3% glutaraldehyde in 0.1 M phosphate buffer at 4°C for 3 days. After fixation in 2% osmium tetroxide, dehydrating with graded ethanol, infiltrating with propylene oxide, and embedding in Epon 812, tissue fragments were cut into 80-nm sections, then stained with lead citrate and uranyl acetate Growth Curve and examined with a Zeiss Model EM109 electron miA concentration of 3 X lo4 cells per ml of RPMI 10% croscope at 80 kV [ 161. fetal bovine serum was cultivated in each well of the Cytokeratin Detection 24-well culture plates (Costar, Cambridge, MA), and the The modified immunofluorescent method of Debus et number was counted daily by hemocytometer (Reichert, Buffalo, NY) for 6 days without refeeding. Population al. [ 171 was administered for detection of cytokeratin in doubling time and growth ability in the RPMI 0.5%, 2%, cancer cells by mouse monoclonal anti-cytokeratin 8.13 5 % fetal bovine serum and MEM (Eagle's minimum es- antibody (Sigma Chemical Co., St. Louis, MO).
262
Cheng et al.
TABLE I. Characteristics of CC-M Cell Lines* Cell lines Characteristics
CC-M2
CC-M3
CC-M4
CC-M2NM -
Age (years) Sex Site Histological grade Modified Dukes’ stage Population doubling time Passage 25 Passage 50 Efficiency Plating (5%) Seeding Synthesis CEA (ngiml) beta-HCG (mIUlml) Tuniorigenicity Passage 5 Passage 50
58 M SC” MDb B1 (hr) 20.0 18.5
80 F TC MD D
92 M R MD B2
16.6 16.1
11.2 16.8
64.5 51.4
34 81
25 16
22 68
