- Email: [email protected]
5-Fluorouracil-induced apoptosis in cultured oral cancer cells
Oral Oncology 36 (2000) 236±241
www.elsevier.com/locate/oraloncology
5-Fluorouracil-induced apoptosis in cultured oral cancer cells D. Tonga, M. Pootb, D. Huc, D. Odaa,* a
Department of Oral and Maxillofacial Surgery, School of Dentistry, University of Washington, Box 357134, Seattle, WA 98195-7134, USA b Department of Pathology, School of Medicine, University of Washington, Seattle, WA 98195-7134, USA c Department of Ophthalmology, School of Medicine, University of Washington, Seattle, WA 98195-7134, USA Received 5 May 1999; accepted 18 May 1999
Abstract Chemotherapy is commonly used to treat advanced oral squamous cell carcinoma (SCC) and is known to kill cancer cells through apoptosis. Our hypothesis states that 5-¯uorouracil (5FU) also kills cultured oral epithelial cells through programmed cell death or apoptosis. Cultured oral cancer cells were exposed to an optimum dose of 20 mg/ml of 5FU. Cells were analyzed for changes in cell cycle distribution and induction of cell death including apoptosis. Normal control, human papilloma virusimmortalized (PP), ATCC SCC cell line (CA1) and two primary oral SCC cell lines (CA3 and -4) were studied. Inhibition of apoptosis by a pan-caspase inhibitor was used. SYTO 11 ¯ow cytometry showed increased apoptosis in all 5FU-treated cell cultures compared to untreated controls. The results show biological variation in apoptotic response. CA1 had the lowest apoptotic rate of the cancer cell lines at 1.5%. Next lowest was CA3, followed by CA4 and PP. In addition, alteration in the G1 and S phase fractions were found. Untreated CA1 showed 28% G1, 53% S compared to 43% G1, and 40% S of treated. We investigated the pathway of apoptosis using the pan-caspase inhibitor IDN-1529 by methylthiazolyl diphenyl tetrazolium bromide (MTT) colorimetric analysis. Results showed mild inhibition of cell death when cells were incubated with 50 mM IDN-1529 for 24 h. This suggests a probable caspase-dependent apoptotic pathway. In conclusion, our data suggest that 5FU induces oral cancer cell death through apoptosis and that biological variation exists between normal and cancer cells and between dierent types of cancer cells themselves. Our data indicate that cultures of a useful in vitro model for chemosensitivity assays are possible. Our results also suggest a caspase-dependent pathway for chemocytotoxicity in oral SCC. # 2000 Elsevier Science Ltd. All rights reserved. Keywords: Apoptosis; Oral cancer; Epithelial; Flow cytometry; Caspase inhibitor
1. Introduction Oral squamous cell carcinoma (SCC) presents a signi®cant health problem that aects over half a million people in the world each year [1,2]. Treatment has traditionally been surgical, with radiation and chemotherapy used as supportive measures [3±5]. Recent advances in the understanding of chemotherapy have produced new strategies for the use of chemotherapeutic agents in a primary role, some studies suggesting that in 40% of patients with head and neck cancer, including oral cancer, chemotherapy followed by radiation therapy was comparable to surgery and radiation therapy [6]. A commonly used chemotherapeutic agent is 5¯uorouracil (5FU), an analog of naturally occurring * Corresponding author. Tel.: +1-206-616-4748; fax: +1-206-6857222. E-mail address: [email protected] (D. Oda).
pyrimidine uracil with a ¯uorine atom substituted at the carbon-5 position [7]. 5FU is commonly used in conjunction with other agents such as cisplatin and methotrexate [8]. 5FU alone or in combination with other drug(s) is used to treat cancer in a variety of sites: breast, cervix, ovaries, gastrointestinal tract and the head and neck [7,9]. Apoptosis is thought to mediate therapeutic ecacy in cancer chemotherapy [10]. It is described as an active and physiologic process of cell death with distinct morphological characteristics, which include overall cell shrinkage, nuclear chromatin condensation and internucleosomal DNA cleavage [11,12]. However, many morphologically based tests such as the terminal deoxytransferase-biotinylated-dUTP nick end labeling (TUNEL) are sensitive but not speci®c for apoptosis and may identify only late changes in the apoptotic process [13]. By comparison, ¯ow cytometry and the SYTO 11 assays appear to be very sensitive in the detection of apoptosis [14,15].
1368-8375/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved. PII: S1368-8375(99)00079-2
D. Tong et al. / Oral Oncology 36 (2000) 236±241
In this study, we investigate the chemosensitivity of oral epithelial cells, including SCC cells, to 5FU by combining results from a ¯ow cytometric assay for apoptosis and a MTT assay. 2. Materials and methods 2.1. Reagents SYTO 11, Hoechst 33342, propidium iodide and 40 , 6diamino-2-phenylindole (DAPI) stains were obtained from Molecular Probes Inc. (Eugene, OR, USA). 5FU was obtained from Sigma Chemical Company (St. Louis, MO, USA) and prepared from an anhydrous crystalline powder dissolved in keratinocyte basal medium (KBM) (Clonetics Inc., Walkersville, MD, USA) adjusted to a pH of 9.0. Due to the short half-life of 5FU, all solutions were made immediately prior to their use. 2.2. Cell culture A total of ®ve cell types were used: normal oral epithelium (FH4); human papilloma virus type 16 E6:E7 gene-immortalized cells (PP); ATCC SCC (CA1) and two primary oral SCCs (CA3 and CA4). Cell line details are summarized in Table 1. Primary normal oral epithelial cell cultures were obtained from patients undergoing routine surgical removal of third molars. Our SCC cell cultures were obtained from patients undergoing de®nitive resection for oral SCC. Normal epithelial cell cultures were maintained in KBM with bronchial epithelial cell growth medium single-quots (Clonetics Inc., Walkersville, MD, USA). Cancer cells were maintained in a 50:50 mix of keratinocyte serumfree medium and Ham's F12 nutrient medium (Gibco, Grand Island, NY, USA). All cell culture media were supplemented with 10% fetal bovine serum. Tissue samples were cut into small pieces and incubated overnight in 4 mg/ml Dispase II (Boehringer Mannheim Biochemicals, Indianapolis, IN, USA) at 4 C. Surface epithelium was trypsinized to dissociate the cells into suspension and they were then centrifuged and resuspended in the respective cell media. The cells were seeded in 100-mm plastic culture plates (Falcon Glassware, Lincoln Park, NJ, USA) and fed every 48 h. The
237
cells were grown to 70±80% con¯uence (3±5 days) prior to being used for experiments. 2.3. Dose response A time and dose response for 5FU was performed for both normal and malignant cell cultures. The following dilutions were made: 10, 15, 20, 25, 30 and 35 mg/ml solutions. Normal and malignant cells were continuously treated at various time intervals: 1, 6, 24, 48, 72 h. Viable cell numbers were counted using Trypan Blue dye exclusion as described by Drake et al. [16]. Brie¯y, the cells were seeded onto Malloy slides at a concentration of 10,000 cells per well and treated with the various doses of 5FU at dierent time intervals. The cells were then stained with Trypan Blue for 60 s and washed. The cells were then counted under light microscopy at 20 magni®cation. The optimal dose and treatment time was found to be 20 mg/ml (0.15 M) over 24 h as this produced the best observations without reaching 50% cell death. A summary of these results is shown in Table 2. 2.4. Cell cycle distributions Flow cytometry was performed to observe cell cycle distributions of treated and control groups using DAPI as described by Rabinovitch [17]. The cells were treated continuously for 24 h with 5FU at 37 C. After treatment, the cells were trypsinized, centrifuged and resuspended directly in DAPI solution. A concentration of at least 5105 cells per 0.5 ml DAPI solution was prepared. Stained cell suspensions were kept on ice until ready to be analyzed. 2.5. Apoptosis assay SYTO 11 assay is a sensitive assay to detect apoptosis by Poot et al. [14]. The cells were treated with 5FU and prepared as described above. After centrifugation, the cells were resuspended in 1 ml of cell culture medium and stained sequentially with 20 mM HOECHST 33342, 100 nM SYTO 11 and 5 cdmg/ml propidium iodide. The cells were incubated for 30 min at 37 C prior to analysis. Fluorescence intensity after SYTO 11 staining allows dierentiation between normal and apoptotic cells.
Table 1 Cell type, passage number and site of the primary cell cultures used in this study Cell name
Cell type
Passages used
Location
FH4 PP CA1 CA3 CA4
Normal HPV immortalized SCC cell line Primary SCC Primary SCC
6,9,10 16,19,20 5,11,12 4,7,10 3,6,10
Gingiva/retromolar Gingiva Tongue Floor of mouth Tongue
238
D. Tong et al. / Oral Oncology 36 (2000) 236±241
Table 2 5-Fluorouracil (5FU) dose response over 24 h by Trypan Blue exclusion 5FU dose concentration
Trypan Blue positive
Trypan Blue negative
Percentage positive
Cell medium control 5 mg 10 mg 15 mg 20 mg 25 mg 30 mg 70% alcohol
4 12 16 37 51 122 176 202
231 235 285 222 226 235 267 243
1.73 5.10 5.61 16.67 22.57 51.91 65.92 83.13
2.6. Cell cycle assay Stained cell suspensions were analyzed using a Coulter Epics Elite1 ¯ow cytometer equipped with an argon laser set at a light-stabilized output of 15 mW of the 488nm beam. To read the HOECHST 33342 dye, a second argon laser was also set to a light-stabilized output of 10 mW at 360 nM. Phosphate-buered saline was used as the sheath ¯uid and typical sample ¯ow rates ranged from 50 to 350 particles per second. Data were stored on the hard disk of an IBM-compatible PC and analyzed with MPlus software designed by Rabinovitch [17]. 2.7. Caspase inhibitor assay A pan-caspase inhibitor, IDN-1529 was obtained from IDUN Pharmaceuticals (San Diego, CA, USA) and a stock concentration of 20 mM was made. A dose response was performed using 50, 100 and 200 mM dilutions of IDN-1529 and optical densities compared using methylthiazolyl diphenyl tetrazolium bromide (MTT) (Sigma, St. Louis, MO, USA) assay technique as described by Mossman [18]. Dilutions of 50 and 100 mM were added to cells plated in 96-well plates (Falcon Labware, Lincoln Park, NJ, USA) one hour prior to and 1 h after treatment of 5FU. The cells were treated with 5FU for 24 h, stained with 20 ml MTT and incubated for a further 4 h at 37 C prior to addition of 150 ml dimethyl sulfoxide (DMSO) (Sigma, St. Louis, MO, USA). Cell death populations cased on optical densities were observed using an MTT assay technique as described by Mossman [18]. Controls using cell medium alone, 5FU alone and IDN-1529 alone were also compared. A cell concentration of 10,000 cells per well was seeded. The experiment was repeated at a dose of 200 mM IDN1529 and processed as above. 3. Results 3.1. Cell cultures A total of ®ve cell types were used: normal oral epithelium (FH4); human papilloma virus type 16 E6:E7
gene-immortalized cells (PP); ATCC SCC (CA1) and two primary oral squamous cell carcinomas (CA3 and CA4). Cell line details are summarized in Table 2. 3.2. Assay for apoptosis All ®ve cell lines were divided into two groups: 5FU treated and untreated. One normal primary cell line (FH4) acted as the control and showed a decrease in live cells from 63% untreated versus 5% treated. Dead cells increased from 32% untreated to 82% in the treated groups. The focus of this paper is on apoptosis; therefore, the following results will emphasize on apoptotic versus live proportions. Treatment with 5FU increased apoptotic cell number from 5 to 9%. PP cells showed a smaller decrease in live cells from 53 to 34% but a larger dierence in apoptosis from 6 to 20%. CA1 cells showed an increase in apoptosis when treated with 5FU, 0.3% in untreated to 1.5% in treated. Apoptosis in CA3 increased from 2 to 10%. CA4 showed increase in apoptosis from 0.5 to 19.5%. A summary of these results is shown in Fig. 1. 3.3. 5FU eects on cell cycle distribution The cells were divided into three groups: normal control; control with pH-adjusted medium; and 20 mg/ml 5FU treated. pH adjustment represents a medium of pH 9.0. 5FU crystals only dissolve in alkaline solution. We used an additional control for pH 9.0. Controls using pH-adjusted medium results were similar to the de®ned medium alone control. G1 and G2 proportions were higher in the FH4-untreated group than the treated group, 22.1% versus 11.6% for G1 and 25.2% versus 1.1% for G2, whereas the S-phase was similar both at 14.3% versus 13.1%. The remaining percentage represents dead cells. The trend was similar with PP cells where the cells in G1 phase in the untreated group were higher than the treated groups at 32.7% versus 13.9%. The S-phase in PP cells went up from 6.6 to 9.8%. Similar decrease in the G1 was noted with the ATCC cancer cells, where G1 decreased from 38.7 to 12.9%. Similar results were noted with the CA3 and CA4
D. Tong et al. / Oral Oncology 36 (2000) 236±241
239
Fig. 1. SYTO 11 apoptosis assay showing live, apoptotic and dead population percentages between 5-¯uorouracil (5FU)-treated (20 mg/ ml over 24 h) and untreated cells (n=10,000).
cancer cell lines. A summary of results is shown in Fig. 2 and Table 3. 3.4. Pan-caspase inhibitor There were four groups per cell line: medium only; IDN-1529 added 1 h prior to treatment with 5FU; 5FU
Fig. 2. Cell cycle distributions shown as percentage populations between 5FU-treated (20 mg/ml over 24 h) and untreated cells (n 10; 000).
240
D. Tong et al. / Oral Oncology 36 (2000) 236±241
Table 3 Summary table of changes found in cell cycle distribution after 5¯uorouracil (5FU) treatment (20 mg/ml over 24 h) Cell type
Cell cycle phase G1
FH4 PP CA1 CA3 CA4
# # # # #
S $ # # $ #
G2 # # # # "
treatment alone; and IDN-1529 alone. Dose concentrations of IDN-1529 used were 50, 100 and 200 mM. Results from 100 and 200 mM indicate cytotoxicity of IDN-1529. At 50 mM, results showed decreased optical density in 5FU-treated cells as compared to normal controls. Cells treated with IDN-1529 alone were similar to controls with mild reduction in CA1, -3 and -4 cell lines. Cells treated with IDN-1529 and 5FU showed increased optical density when compared to 5FU-treated alone in all cell lines. A summary of these results is shown in Fig. 3. 4. Discussion Apoptosis or programmed cell death is commonly described as a physiologic form of cell death, with characteristic morphologic changes, the most important being internucleosomal DNA fragmentation [11]. Apoptosis has been studied in terms of cancer development and treatment, with attempts made to identify its role in chemotherapeutic agent-induced cytotoxicity. Chemotherapeutic agents such as 5FU, paclitaxel and cyclophosphamide, have been shown to induce apoptosis in various types of tumor cells including gastric,
Fig. 3. MTT assay demonstrating optical densities between untreated, 5-¯uorouracil (5FU)-treated (20 mg/ml over 24 h) and IDN-1529treated (50 mM IDN-1529) cells.
colon and breast cancers [19,20]. 5FU is often used in the treatment of oral SCC in combination with cisplatin [7]; however, studies involving chemotherapy and apoptosis in oral cancer are somewhat scarce. 5FU is a pyrimidine analog which by interfering with the activity of thymidylate synthase, preventing the formation of thymidylate, the precursor for thymidine triphosphate. 5FU also interferes with RNA synthesis but the exact mechanisms are unclear as yet. We performed a time±dose response study to determine an optimum dose concentration of 0.15 M (20 mg/ ml) 5FU given continuously over 24 h. 5FU dose concentrations used as high as 0.385 M were used in other studies of head and neck SCC. The need for higher doses suggests that head and neck SCC may require higher concentrations of 5FU to undergo apoptosis. Fluorescence intensity after SYTO 11 staining allows deciphering `normal' from apoptotic cells. An advantage of this technique is that changes early in the process of apoptosis can be identi®ed which precede morphologic changes. Our results show that, in all cell lines, apoptosis increased after treatment, with 5FU using the SYTO 11 method. Normal FH4 cells appeared to be the most sensitive to 5FU treatment with decrease in live cell populations from 63 to 5%. CA1 cells appeared to be least sensitive to 5FU treatment with a smaller decrease from 92 to 65%. These results indicate biological variation between normal and malignant and between malignant cell lines, i.e. CA1 cells being less sensitive than CA3. Cell cycle analysis results showed marked reduction in cell cycle distribution in treated cells compared to controls (Table 3). This reduction in cell cycling suggests that our dose concentration of 0.15 M 5FU, despite being relatively high, was needed to induce cell apoptosis. This indicates that oral epithelium and oral cancer are more resistant to chemotherapy and require relatively higher doses of 5FU. It is well known clinically that oral SCC is much more aggressive in behavior than skin SCC, although both have the same cellular originÐ the strati®ed epithelium. In other cell cycle analyses, the action of 5FU during S-phase leading to G2 phase arrest has been demonstrated in breast cancer cells [21]. This eect was not noted in our studies. Identi®cation of human gene products involved with apoptosis in the nematode, C. elegans has focused the attention of many researchers on the biochemical pathways of cell death including apoptosis. At least one pathway has been identi®ed, involving cysteine proteases or caspases. The pathway of caspase activation leading to cell death follows a de®nite sequence, a cascade of precursors and co-enzymes similar to the serine proteases in the coagulation cascade [13]. The presence of a caspase-dependent pathway for cell death in oral cancer cells was apparent in our studies. We used a pan-caspase inhibitor, IDN-1529, with which
D. Tong et al. / Oral Oncology 36 (2000) 236±241
the cell lines were pre-treated prior to exposure to 5FU. This inhibitor was most eective at 50 mM, when added 1 h prior 24-h treatment with 5FU (Fig. 3). At 100 and 200 mM concentrations resulted in toxicity. Other researchers have seen similar caspase-dependent cell death when using 5FU, but not with cisplatin [22]. Our ®ndings may have a potential clinical application for in vitro chemosensitivity assay. This application has already been described in determining toxicity levels of 5FU and mitomycin-c in corneal endothelium of the eye [23] and in ovarian cancer cell lines [24]. The variable apoptosis levels identi®ed in our cell lines emphasizes the potential of our working model for this application. Acknowledgments VA PSHCS RDIS 0001 and Dean CDE Project 29, University of Washington, Seattle, USA, supported this study.
[9] [10] [11] [12] [13]
[14] [15] [16] [17] [18]
References [1] Parkin DM, Pisani P, Ferlay J. Estimates of the worldwide incidence of 18 major cancers in 1985. International Journal of Cancer 1993;54:594±606. [2] Morse D, Pendrys DG, Neely AL, Psoter WJ. Trends in the incidence of lip, oral and pharyngeal cancer: Connecticut, 1935± 94. Oral Oncology 1999;35:1±8. [3] McAndrew PG. Oral cancer and precancer: treatment. British Dental Journal 1990;168:191±8. [4] Strong E. Surgical management of oral cancer. Dental Clinics of North America 1990;34:185±203. [5] Gluckman JL, Savoury LW. Carcinoma of the oral cavity. In: Paparella MM, editor. Otolaryngology. Philadelphia: WB Saunders, 1991. p. 2041±67. [6] Posner MR, Colevas AD. Induction chemotherapy in the management of squamous cell cancer of the head and neck. The Cancer Journal from Scienti®c American 1997;3:73±5. [7] MacDonald JS, editor. Manual of oncologic therapeutics, 3rd ed. Philadelphia: JB Lippincott, 1991. p. 111±13. [8] Merlano M, Benasso M, Cavallari M, Blengio F, Rosso M.
[19]
[20] [21] [22] [23] [24]
241
Chemotherapy in head and neck cancer. European Journal of Cancer B, Oral Oncology 1994;30B:283±9. Perry MC, editor. The chemotherapy source book. 2nd ed. Baltimore: Williams and Wilkins, 1996. p. 318, 323±325. Darzynkiewicz Z, Juan G, Li X, Gorczyca W, Murakami T, Tragonos F. Cytometry in cell necrobiology: analysis of apoptosis and accidental cell death (necrosis). Cytometry 1997;27:1±20. Bowen ID, Bowen SM. Programmed cell death in tumours and tissues. London: Chapman and Hall, 1990. p. 1±68. Hockenbery D. De®ning apoptosis. American Journal of Pathology 1995;146:16±19. Dong Z, Saikumar P, Weinberg JM, Venkatachalam MA. Internucleosomal DNA cleavage triggered by plasma membrane damage during necrotic cell death. American Journal of Pathology 1997;151:1205±13. Poot M, Gibson LL, Singer VL. Detection of apoptosis in live cells by Mito Tracker red CMXRos and SYTO dye ¯ow cytometry. Cytometry 1997;27:358±64. Darzynkiewicz Z, Bruno S, Del Bino G, Gorczyca W, Hotz MA, Lassota P, Tragonos F. Features of apoptotic cells measured by ¯ow cytometry. Cytometry 1992;13:795±808. Drake W, Ungaro PC, Mardiney MR. Formalin ®xed cell preparations as standards for use in the automated trypan blue cytotoxic assay. Transplantation 1972;14:127±30. Rabinovitch P. DNA content histogram and cell cycle analysis. Methods in Cell Biology 1994;41:263±96. Mossman T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. Journal of Immunological Methods 1983;65:55±63. Inada T, Ichikawa A, Kubota T, Ogata Y, Moosa AR, Homan RM. 5FU induced apoptosis correlates with ecacy against human gastric and colon cancer xenografts in nude mice. Anticancer Research 1997;17:1965±71. Inada T, Ichikawa A, Igarashi S, Kubota T, Ogata Y. Eect of preoperative 5-¯uorouracil on apoptosis of advanced gastric cancer. Journal of Surgical Oncology 1997;65:106±10. Okamoto S, Sakai M, Uchida J, Saito H. 5-¯uorouracil induces apoptotic cell death with G2 phase arrest in human breast cancer grafted in nude mice. Anticancer Research 1996;16:2699±704. Ohtani T, Hatori M, Ohashi M, Kamijo R Involvement of caspases in 5FU induced apoptosis (Abstract). 1999 IADR Meeting, Vancouver, BC, Canada. Nuyts RMMA, Pels E, Greve EL. The eects of 5-¯uorouracil and mitomycin C on the corneal endothelium. Current Eye Research 1992;6:565±70. Havrilesky LJ, Elbendary A, Hurteau JA, Whitaker RS, Rodriguez GC, Berchuck A. Chemotherapy induced apoptosis in epithelial ovarian cancers. Obstetrics and Gynecology 1995;85: 1007±1010.
www.elsevier.com/locate/oraloncology
5-Fluorouracil-induced apoptosis in cultured oral cancer cells D. Tonga, M. Pootb, D. Huc, D. Odaa,* a
Department of Oral and Maxillofacial Surgery, School of Dentistry, University of Washington, Box 357134, Seattle, WA 98195-7134, USA b Department of Pathology, School of Medicine, University of Washington, Seattle, WA 98195-7134, USA c Department of Ophthalmology, School of Medicine, University of Washington, Seattle, WA 98195-7134, USA Received 5 May 1999; accepted 18 May 1999
Abstract Chemotherapy is commonly used to treat advanced oral squamous cell carcinoma (SCC) and is known to kill cancer cells through apoptosis. Our hypothesis states that 5-¯uorouracil (5FU) also kills cultured oral epithelial cells through programmed cell death or apoptosis. Cultured oral cancer cells were exposed to an optimum dose of 20 mg/ml of 5FU. Cells were analyzed for changes in cell cycle distribution and induction of cell death including apoptosis. Normal control, human papilloma virusimmortalized (PP), ATCC SCC cell line (CA1) and two primary oral SCC cell lines (CA3 and -4) were studied. Inhibition of apoptosis by a pan-caspase inhibitor was used. SYTO 11 ¯ow cytometry showed increased apoptosis in all 5FU-treated cell cultures compared to untreated controls. The results show biological variation in apoptotic response. CA1 had the lowest apoptotic rate of the cancer cell lines at 1.5%. Next lowest was CA3, followed by CA4 and PP. In addition, alteration in the G1 and S phase fractions were found. Untreated CA1 showed 28% G1, 53% S compared to 43% G1, and 40% S of treated. We investigated the pathway of apoptosis using the pan-caspase inhibitor IDN-1529 by methylthiazolyl diphenyl tetrazolium bromide (MTT) colorimetric analysis. Results showed mild inhibition of cell death when cells were incubated with 50 mM IDN-1529 for 24 h. This suggests a probable caspase-dependent apoptotic pathway. In conclusion, our data suggest that 5FU induces oral cancer cell death through apoptosis and that biological variation exists between normal and cancer cells and between dierent types of cancer cells themselves. Our data indicate that cultures of a useful in vitro model for chemosensitivity assays are possible. Our results also suggest a caspase-dependent pathway for chemocytotoxicity in oral SCC. # 2000 Elsevier Science Ltd. All rights reserved. Keywords: Apoptosis; Oral cancer; Epithelial; Flow cytometry; Caspase inhibitor
1. Introduction Oral squamous cell carcinoma (SCC) presents a signi®cant health problem that aects over half a million people in the world each year [1,2]. Treatment has traditionally been surgical, with radiation and chemotherapy used as supportive measures [3±5]. Recent advances in the understanding of chemotherapy have produced new strategies for the use of chemotherapeutic agents in a primary role, some studies suggesting that in 40% of patients with head and neck cancer, including oral cancer, chemotherapy followed by radiation therapy was comparable to surgery and radiation therapy [6]. A commonly used chemotherapeutic agent is 5¯uorouracil (5FU), an analog of naturally occurring * Corresponding author. Tel.: +1-206-616-4748; fax: +1-206-6857222. E-mail address: [email protected] (D. Oda).
pyrimidine uracil with a ¯uorine atom substituted at the carbon-5 position [7]. 5FU is commonly used in conjunction with other agents such as cisplatin and methotrexate [8]. 5FU alone or in combination with other drug(s) is used to treat cancer in a variety of sites: breast, cervix, ovaries, gastrointestinal tract and the head and neck [7,9]. Apoptosis is thought to mediate therapeutic ecacy in cancer chemotherapy [10]. It is described as an active and physiologic process of cell death with distinct morphological characteristics, which include overall cell shrinkage, nuclear chromatin condensation and internucleosomal DNA cleavage [11,12]. However, many morphologically based tests such as the terminal deoxytransferase-biotinylated-dUTP nick end labeling (TUNEL) are sensitive but not speci®c for apoptosis and may identify only late changes in the apoptotic process [13]. By comparison, ¯ow cytometry and the SYTO 11 assays appear to be very sensitive in the detection of apoptosis [14,15].
1368-8375/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved. PII: S1368-8375(99)00079-2
D. Tong et al. / Oral Oncology 36 (2000) 236±241
In this study, we investigate the chemosensitivity of oral epithelial cells, including SCC cells, to 5FU by combining results from a ¯ow cytometric assay for apoptosis and a MTT assay. 2. Materials and methods 2.1. Reagents SYTO 11, Hoechst 33342, propidium iodide and 40 , 6diamino-2-phenylindole (DAPI) stains were obtained from Molecular Probes Inc. (Eugene, OR, USA). 5FU was obtained from Sigma Chemical Company (St. Louis, MO, USA) and prepared from an anhydrous crystalline powder dissolved in keratinocyte basal medium (KBM) (Clonetics Inc., Walkersville, MD, USA) adjusted to a pH of 9.0. Due to the short half-life of 5FU, all solutions were made immediately prior to their use. 2.2. Cell culture A total of ®ve cell types were used: normal oral epithelium (FH4); human papilloma virus type 16 E6:E7 gene-immortalized cells (PP); ATCC SCC (CA1) and two primary oral SCCs (CA3 and CA4). Cell line details are summarized in Table 1. Primary normal oral epithelial cell cultures were obtained from patients undergoing routine surgical removal of third molars. Our SCC cell cultures were obtained from patients undergoing de®nitive resection for oral SCC. Normal epithelial cell cultures were maintained in KBM with bronchial epithelial cell growth medium single-quots (Clonetics Inc., Walkersville, MD, USA). Cancer cells were maintained in a 50:50 mix of keratinocyte serumfree medium and Ham's F12 nutrient medium (Gibco, Grand Island, NY, USA). All cell culture media were supplemented with 10% fetal bovine serum. Tissue samples were cut into small pieces and incubated overnight in 4 mg/ml Dispase II (Boehringer Mannheim Biochemicals, Indianapolis, IN, USA) at 4 C. Surface epithelium was trypsinized to dissociate the cells into suspension and they were then centrifuged and resuspended in the respective cell media. The cells were seeded in 100-mm plastic culture plates (Falcon Glassware, Lincoln Park, NJ, USA) and fed every 48 h. The
237
cells were grown to 70±80% con¯uence (3±5 days) prior to being used for experiments. 2.3. Dose response A time and dose response for 5FU was performed for both normal and malignant cell cultures. The following dilutions were made: 10, 15, 20, 25, 30 and 35 mg/ml solutions. Normal and malignant cells were continuously treated at various time intervals: 1, 6, 24, 48, 72 h. Viable cell numbers were counted using Trypan Blue dye exclusion as described by Drake et al. [16]. Brie¯y, the cells were seeded onto Malloy slides at a concentration of 10,000 cells per well and treated with the various doses of 5FU at dierent time intervals. The cells were then stained with Trypan Blue for 60 s and washed. The cells were then counted under light microscopy at 20 magni®cation. The optimal dose and treatment time was found to be 20 mg/ml (0.15 M) over 24 h as this produced the best observations without reaching 50% cell death. A summary of these results is shown in Table 2. 2.4. Cell cycle distributions Flow cytometry was performed to observe cell cycle distributions of treated and control groups using DAPI as described by Rabinovitch [17]. The cells were treated continuously for 24 h with 5FU at 37 C. After treatment, the cells were trypsinized, centrifuged and resuspended directly in DAPI solution. A concentration of at least 5105 cells per 0.5 ml DAPI solution was prepared. Stained cell suspensions were kept on ice until ready to be analyzed. 2.5. Apoptosis assay SYTO 11 assay is a sensitive assay to detect apoptosis by Poot et al. [14]. The cells were treated with 5FU and prepared as described above. After centrifugation, the cells were resuspended in 1 ml of cell culture medium and stained sequentially with 20 mM HOECHST 33342, 100 nM SYTO 11 and 5 cdmg/ml propidium iodide. The cells were incubated for 30 min at 37 C prior to analysis. Fluorescence intensity after SYTO 11 staining allows dierentiation between normal and apoptotic cells.
Table 1 Cell type, passage number and site of the primary cell cultures used in this study Cell name
Cell type
Passages used
Location
FH4 PP CA1 CA3 CA4
Normal HPV immortalized SCC cell line Primary SCC Primary SCC
6,9,10 16,19,20 5,11,12 4,7,10 3,6,10
Gingiva/retromolar Gingiva Tongue Floor of mouth Tongue
238
D. Tong et al. / Oral Oncology 36 (2000) 236±241
Table 2 5-Fluorouracil (5FU) dose response over 24 h by Trypan Blue exclusion 5FU dose concentration
Trypan Blue positive
Trypan Blue negative
Percentage positive
Cell medium control 5 mg 10 mg 15 mg 20 mg 25 mg 30 mg 70% alcohol
4 12 16 37 51 122 176 202
231 235 285 222 226 235 267 243
1.73 5.10 5.61 16.67 22.57 51.91 65.92 83.13
2.6. Cell cycle assay Stained cell suspensions were analyzed using a Coulter Epics Elite1 ¯ow cytometer equipped with an argon laser set at a light-stabilized output of 15 mW of the 488nm beam. To read the HOECHST 33342 dye, a second argon laser was also set to a light-stabilized output of 10 mW at 360 nM. Phosphate-buered saline was used as the sheath ¯uid and typical sample ¯ow rates ranged from 50 to 350 particles per second. Data were stored on the hard disk of an IBM-compatible PC and analyzed with MPlus software designed by Rabinovitch [17]. 2.7. Caspase inhibitor assay A pan-caspase inhibitor, IDN-1529 was obtained from IDUN Pharmaceuticals (San Diego, CA, USA) and a stock concentration of 20 mM was made. A dose response was performed using 50, 100 and 200 mM dilutions of IDN-1529 and optical densities compared using methylthiazolyl diphenyl tetrazolium bromide (MTT) (Sigma, St. Louis, MO, USA) assay technique as described by Mossman [18]. Dilutions of 50 and 100 mM were added to cells plated in 96-well plates (Falcon Labware, Lincoln Park, NJ, USA) one hour prior to and 1 h after treatment of 5FU. The cells were treated with 5FU for 24 h, stained with 20 ml MTT and incubated for a further 4 h at 37 C prior to addition of 150 ml dimethyl sulfoxide (DMSO) (Sigma, St. Louis, MO, USA). Cell death populations cased on optical densities were observed using an MTT assay technique as described by Mossman [18]. Controls using cell medium alone, 5FU alone and IDN-1529 alone were also compared. A cell concentration of 10,000 cells per well was seeded. The experiment was repeated at a dose of 200 mM IDN1529 and processed as above. 3. Results 3.1. Cell cultures A total of ®ve cell types were used: normal oral epithelium (FH4); human papilloma virus type 16 E6:E7
gene-immortalized cells (PP); ATCC SCC (CA1) and two primary oral squamous cell carcinomas (CA3 and CA4). Cell line details are summarized in Table 2. 3.2. Assay for apoptosis All ®ve cell lines were divided into two groups: 5FU treated and untreated. One normal primary cell line (FH4) acted as the control and showed a decrease in live cells from 63% untreated versus 5% treated. Dead cells increased from 32% untreated to 82% in the treated groups. The focus of this paper is on apoptosis; therefore, the following results will emphasize on apoptotic versus live proportions. Treatment with 5FU increased apoptotic cell number from 5 to 9%. PP cells showed a smaller decrease in live cells from 53 to 34% but a larger dierence in apoptosis from 6 to 20%. CA1 cells showed an increase in apoptosis when treated with 5FU, 0.3% in untreated to 1.5% in treated. Apoptosis in CA3 increased from 2 to 10%. CA4 showed increase in apoptosis from 0.5 to 19.5%. A summary of these results is shown in Fig. 1. 3.3. 5FU eects on cell cycle distribution The cells were divided into three groups: normal control; control with pH-adjusted medium; and 20 mg/ml 5FU treated. pH adjustment represents a medium of pH 9.0. 5FU crystals only dissolve in alkaline solution. We used an additional control for pH 9.0. Controls using pH-adjusted medium results were similar to the de®ned medium alone control. G1 and G2 proportions were higher in the FH4-untreated group than the treated group, 22.1% versus 11.6% for G1 and 25.2% versus 1.1% for G2, whereas the S-phase was similar both at 14.3% versus 13.1%. The remaining percentage represents dead cells. The trend was similar with PP cells where the cells in G1 phase in the untreated group were higher than the treated groups at 32.7% versus 13.9%. The S-phase in PP cells went up from 6.6 to 9.8%. Similar decrease in the G1 was noted with the ATCC cancer cells, where G1 decreased from 38.7 to 12.9%. Similar results were noted with the CA3 and CA4
D. Tong et al. / Oral Oncology 36 (2000) 236±241
239
Fig. 1. SYTO 11 apoptosis assay showing live, apoptotic and dead population percentages between 5-¯uorouracil (5FU)-treated (20 mg/ ml over 24 h) and untreated cells (n=10,000).
cancer cell lines. A summary of results is shown in Fig. 2 and Table 3. 3.4. Pan-caspase inhibitor There were four groups per cell line: medium only; IDN-1529 added 1 h prior to treatment with 5FU; 5FU
Fig. 2. Cell cycle distributions shown as percentage populations between 5FU-treated (20 mg/ml over 24 h) and untreated cells (n 10; 000).
240
D. Tong et al. / Oral Oncology 36 (2000) 236±241
Table 3 Summary table of changes found in cell cycle distribution after 5¯uorouracil (5FU) treatment (20 mg/ml over 24 h) Cell type
Cell cycle phase G1
FH4 PP CA1 CA3 CA4
# # # # #
S $ # # $ #
G2 # # # # "
treatment alone; and IDN-1529 alone. Dose concentrations of IDN-1529 used were 50, 100 and 200 mM. Results from 100 and 200 mM indicate cytotoxicity of IDN-1529. At 50 mM, results showed decreased optical density in 5FU-treated cells as compared to normal controls. Cells treated with IDN-1529 alone were similar to controls with mild reduction in CA1, -3 and -4 cell lines. Cells treated with IDN-1529 and 5FU showed increased optical density when compared to 5FU-treated alone in all cell lines. A summary of these results is shown in Fig. 3. 4. Discussion Apoptosis or programmed cell death is commonly described as a physiologic form of cell death, with characteristic morphologic changes, the most important being internucleosomal DNA fragmentation [11]. Apoptosis has been studied in terms of cancer development and treatment, with attempts made to identify its role in chemotherapeutic agent-induced cytotoxicity. Chemotherapeutic agents such as 5FU, paclitaxel and cyclophosphamide, have been shown to induce apoptosis in various types of tumor cells including gastric,
Fig. 3. MTT assay demonstrating optical densities between untreated, 5-¯uorouracil (5FU)-treated (20 mg/ml over 24 h) and IDN-1529treated (50 mM IDN-1529) cells.
colon and breast cancers [19,20]. 5FU is often used in the treatment of oral SCC in combination with cisplatin [7]; however, studies involving chemotherapy and apoptosis in oral cancer are somewhat scarce. 5FU is a pyrimidine analog which by interfering with the activity of thymidylate synthase, preventing the formation of thymidylate, the precursor for thymidine triphosphate. 5FU also interferes with RNA synthesis but the exact mechanisms are unclear as yet. We performed a time±dose response study to determine an optimum dose concentration of 0.15 M (20 mg/ ml) 5FU given continuously over 24 h. 5FU dose concentrations used as high as 0.385 M were used in other studies of head and neck SCC. The need for higher doses suggests that head and neck SCC may require higher concentrations of 5FU to undergo apoptosis. Fluorescence intensity after SYTO 11 staining allows deciphering `normal' from apoptotic cells. An advantage of this technique is that changes early in the process of apoptosis can be identi®ed which precede morphologic changes. Our results show that, in all cell lines, apoptosis increased after treatment, with 5FU using the SYTO 11 method. Normal FH4 cells appeared to be the most sensitive to 5FU treatment with decrease in live cell populations from 63 to 5%. CA1 cells appeared to be least sensitive to 5FU treatment with a smaller decrease from 92 to 65%. These results indicate biological variation between normal and malignant and between malignant cell lines, i.e. CA1 cells being less sensitive than CA3. Cell cycle analysis results showed marked reduction in cell cycle distribution in treated cells compared to controls (Table 3). This reduction in cell cycling suggests that our dose concentration of 0.15 M 5FU, despite being relatively high, was needed to induce cell apoptosis. This indicates that oral epithelium and oral cancer are more resistant to chemotherapy and require relatively higher doses of 5FU. It is well known clinically that oral SCC is much more aggressive in behavior than skin SCC, although both have the same cellular originÐ the strati®ed epithelium. In other cell cycle analyses, the action of 5FU during S-phase leading to G2 phase arrest has been demonstrated in breast cancer cells [21]. This eect was not noted in our studies. Identi®cation of human gene products involved with apoptosis in the nematode, C. elegans has focused the attention of many researchers on the biochemical pathways of cell death including apoptosis. At least one pathway has been identi®ed, involving cysteine proteases or caspases. The pathway of caspase activation leading to cell death follows a de®nite sequence, a cascade of precursors and co-enzymes similar to the serine proteases in the coagulation cascade [13]. The presence of a caspase-dependent pathway for cell death in oral cancer cells was apparent in our studies. We used a pan-caspase inhibitor, IDN-1529, with which
D. Tong et al. / Oral Oncology 36 (2000) 236±241
the cell lines were pre-treated prior to exposure to 5FU. This inhibitor was most eective at 50 mM, when added 1 h prior 24-h treatment with 5FU (Fig. 3). At 100 and 200 mM concentrations resulted in toxicity. Other researchers have seen similar caspase-dependent cell death when using 5FU, but not with cisplatin [22]. Our ®ndings may have a potential clinical application for in vitro chemosensitivity assay. This application has already been described in determining toxicity levels of 5FU and mitomycin-c in corneal endothelium of the eye [23] and in ovarian cancer cell lines [24]. The variable apoptosis levels identi®ed in our cell lines emphasizes the potential of our working model for this application. Acknowledgments VA PSHCS RDIS 0001 and Dean CDE Project 29, University of Washington, Seattle, USA, supported this study.
[9] [10] [11] [12] [13]
[14] [15] [16] [17] [18]
References [1] Parkin DM, Pisani P, Ferlay J. Estimates of the worldwide incidence of 18 major cancers in 1985. International Journal of Cancer 1993;54:594±606. [2] Morse D, Pendrys DG, Neely AL, Psoter WJ. Trends in the incidence of lip, oral and pharyngeal cancer: Connecticut, 1935± 94. Oral Oncology 1999;35:1±8. [3] McAndrew PG. Oral cancer and precancer: treatment. British Dental Journal 1990;168:191±8. [4] Strong E. Surgical management of oral cancer. Dental Clinics of North America 1990;34:185±203. [5] Gluckman JL, Savoury LW. Carcinoma of the oral cavity. In: Paparella MM, editor. Otolaryngology. Philadelphia: WB Saunders, 1991. p. 2041±67. [6] Posner MR, Colevas AD. Induction chemotherapy in the management of squamous cell cancer of the head and neck. The Cancer Journal from Scienti®c American 1997;3:73±5. [7] MacDonald JS, editor. Manual of oncologic therapeutics, 3rd ed. Philadelphia: JB Lippincott, 1991. p. 111±13. [8] Merlano M, Benasso M, Cavallari M, Blengio F, Rosso M.
[19]
[20] [21] [22] [23] [24]
241
Chemotherapy in head and neck cancer. European Journal of Cancer B, Oral Oncology 1994;30B:283±9. Perry MC, editor. The chemotherapy source book. 2nd ed. Baltimore: Williams and Wilkins, 1996. p. 318, 323±325. Darzynkiewicz Z, Juan G, Li X, Gorczyca W, Murakami T, Tragonos F. Cytometry in cell necrobiology: analysis of apoptosis and accidental cell death (necrosis). Cytometry 1997;27:1±20. Bowen ID, Bowen SM. Programmed cell death in tumours and tissues. London: Chapman and Hall, 1990. p. 1±68. Hockenbery D. De®ning apoptosis. American Journal of Pathology 1995;146:16±19. Dong Z, Saikumar P, Weinberg JM, Venkatachalam MA. Internucleosomal DNA cleavage triggered by plasma membrane damage during necrotic cell death. American Journal of Pathology 1997;151:1205±13. Poot M, Gibson LL, Singer VL. Detection of apoptosis in live cells by Mito Tracker red CMXRos and SYTO dye ¯ow cytometry. Cytometry 1997;27:358±64. Darzynkiewicz Z, Bruno S, Del Bino G, Gorczyca W, Hotz MA, Lassota P, Tragonos F. Features of apoptotic cells measured by ¯ow cytometry. Cytometry 1992;13:795±808. Drake W, Ungaro PC, Mardiney MR. Formalin ®xed cell preparations as standards for use in the automated trypan blue cytotoxic assay. Transplantation 1972;14:127±30. Rabinovitch P. DNA content histogram and cell cycle analysis. Methods in Cell Biology 1994;41:263±96. Mossman T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. Journal of Immunological Methods 1983;65:55±63. Inada T, Ichikawa A, Kubota T, Ogata Y, Moosa AR, Homan RM. 5FU induced apoptosis correlates with ecacy against human gastric and colon cancer xenografts in nude mice. Anticancer Research 1997;17:1965±71. Inada T, Ichikawa A, Igarashi S, Kubota T, Ogata Y. Eect of preoperative 5-¯uorouracil on apoptosis of advanced gastric cancer. Journal of Surgical Oncology 1997;65:106±10. Okamoto S, Sakai M, Uchida J, Saito H. 5-¯uorouracil induces apoptotic cell death with G2 phase arrest in human breast cancer grafted in nude mice. Anticancer Research 1996;16:2699±704. Ohtani T, Hatori M, Ohashi M, Kamijo R Involvement of caspases in 5FU induced apoptosis (Abstract). 1999 IADR Meeting, Vancouver, BC, Canada. Nuyts RMMA, Pels E, Greve EL. The eects of 5-¯uorouracil and mitomycin C on the corneal endothelium. Current Eye Research 1992;6:565±70. Havrilesky LJ, Elbendary A, Hurteau JA, Whitaker RS, Rodriguez GC, Berchuck A. Chemotherapy induced apoptosis in epithelial ovarian cancers. Obstetrics and Gynecology 1995;85: 1007±1010.