JOURNAL OF PATHOLOGY, VOL.
164: 17-22 (199 1)
SIMULTANEOUS DETECTION O F FLUORESCENT IN SITU HYBRIDIZATION AND IN VIVO INCORPORATED BrdU IN A HUMAN BLADDER TUMOUR HERMAN VAN DEKKEN, EDWARD W. SCHERVISH, JOHN G. PIZZOLO, WILLIAM R. FAIR AND MYRON R. MELAMED
Department of Pathology and Urology Service, Department of Surgery. Memorial Sloan-Kettering Cancer Center, New York, N Y 10021, U.S.A. Received 4 September 1990 Accepted 14 November 1990
SUMMARY We have used fluorescent in situ hybridization and simultaneous in vivo bromodeoxyuridine labelling of a solid bladder cancer to examine tumour cell subsets for possible proliferative growth differences. In this dual-labelled preparation, most tumour cell nuclei exhibited monosomy 9, consistent with reported karyotypes of bladder cancer.' Incorporated bromodeoxyuridine was visualized with a fluoresceinated antibody in 5-6 per cent of the tumour cells, concordant with S-phase estimates by cell cycle analysis of the flow cytometric DNA histogram. A majority of the bromodeoxyuridine-positive cells also carried the monosomy 9 chromosome abnormality. This is the first report to demonstrate the feasibility of combined in situ hybridization and detection of bromodeoxyuridine incorporated in vivo in human tumour cells in order to provide information on the growth rate of specific subsets of tumour cells identified by chromosomal constitution. KEY WORDS-h
situ hybridization, in vivo BrdU labelling, fluorescence, double-labelling, monosomy 9, bladder cancer, subset analysis.
INTRODUCTION In recent years, fluorescent in situ hybridization (FISH) has emerged as a fast and powerful tool to study numerical chromosome changes in interphase nuclei of tumour cells and cell For this purpose labelled chromosome specific alpha-satellite DNA probes were applied.',' These probes bind to the (peri)centromeric repetitive sequences found on human chromosomes.8~9Owing to the repetitive nature of this DNA, the probes bind to a large target, which is easily visualized in the interphase nucleus. Although much is known about the specific chromosome aberrations in haematologic cancers, there is still little information concerning solid turnours.'.''''' The reason for this is the difficulty Addressee for correspondence: Herman van Dekken, Department of Pathology. University Hospital Nijmegen, Geert Grooteplein Zuid 24,6525 GA Nijmegen, The Netherlands.
0022-3417/91/01001746 $05.00 0 1991 byJohn Wiley&Sons, Ltd.
in preparing high-quality metaphase spreads for cytogenetic analysis. Flow cytometry, which has yielded important data on DNA ploidy measurements, cannot disclose the primary cytogenetic events. I 2,1 We have been interested in developing a method to investigate the relationship between tumour karyotype and tumour proliferation and progression, in particular with respect to tumour cell clones that carry tumour specific chromosome abnormalities. As a target for study, we have chosen bladder cancer, since a numerical chromosome aberration (monosomy 9) has been found repeatedly in this tumour'"O (Schervish et al., in preparation). By treating the tumour cells with bromodeoxyuridine (BrdU), information can be derived about the fraction of growing (i.e., DNA-synthesizing) cells within the tumour. In thispaperwedescribea method forstudying the proliferative differences in human bladder tumour I 4 3 l 5
18
H. VAN DEKKEN E T A L .
cells with and without monosomy 9. FISH with chromosome 1 and 9 specific DNA probes, labelled with biotin, were used to assess the presence of the chromosome aberration in interphase nuclei. The DNA-synthesizing cells of the tumour were labelled in vivo,just prior to surgery, by BrdU administration anddetected later bymeansofafluoresceinated antiBrdU antibody. By combining these two methods (BrdU-FISH) in a dual-parameter detection system, we were able to investigate the relationship between growth rate and the presence of monosomy 9 in interphase cells of this solid cancer. MATERIALS AND METHODS Sample preparation BrdU (200 mg/m2) was administered to the patient intravenously in 150 ml of D5W (5 per cent dextrose in H 2 0 ) over 30 min, about 1 h prior to cystoprostatectomy. This patient was one of the first in an approved clinical trial involving in vivo BrdU administration to patients with bladder cancer. Histopathology of the surgical specimen showed in situ and deeply invasive transitional cell carcinoma of the right lateral wall and base of the bladder, extending through the bladder wall and into ducts and stroma of the prostate. Multifocal carcinoma in situ was found elsewhere in the bladder on the anterior wall, left base, and dome, with superficial invasion in the dome. There was metastatic carcinoma in pelvic lymph nodes bilaterally. The specimen described in this study was taken from the invasive carcinoma of the right lateral wall of the bladder. A nuclear suspension from this fresh surgical specimen was obtained by mechanical disaggregation as described b e f ~ r e . ' ~One . ' ~ million nuclei were used for DNA flow cytometry; the rest were fixed twice in ethanol-acetic acid (3: I ; v/v) for 15 min on ice and stored in the fixative at - 20°C. Prior to in situ hybridization (ISH) and/or BrdU detection, the cells were dropped onto ethanol-cleaned microscope glass slides and air-dried at room temperature for 4-5 h. Probes andprobe labelling The chromosome 1 specific repetitive alphasatellite DNA probe was kindly supplied by Dr H. Cook (MRC, Edinburgh, U.K.).7 The alphasatellite DNA probe for chromosome 9 was a gift from Dr R. K. Moyzis (Los Alamos National Laboratory, Los Alamos, NM).8 These probes, in plasmids, were labelled with biotin- 16-dUTP
(Boehringer Mannheim, Indianapolis, IN) by nick translation (BRL, Gaithersburg, MD) according to the manufacturers' directions. All probes were stored at - 20°C. Pretreatment and in situ hybridization Before making the cellular DNA single-stranded, the slides were incubated with RNase A (Boehringer Mannheim; 200 pg/ml in 2 x SSC, pH 7; 1 x SSC is 0.15 M NaCI+O.OI 5 M sodium citrate) for 30 min at 37°C. Further pretreatment included protein removal and post-fixation with f ~ r m a l i n . 'Pro~ duction of single-stranded target DNA was accomplished either with heat d e n a t u r a t i ~ n or ~~~~'~ with an enzymatic procedure employing Exonuclease I11 (BRL).'* The slides were then dehydrated in an ethanol series. Subsequently, in situ hybridization was performed essentially as de~cribed.',~,'~ After TSH, the slides were washed under stringent conditions and then immersed in 4 x SSC+O.l per cent Tween 20, pH 7.5 (4SSCT). D N A probe detection Detection of the biotinylated probes was accomplished with fluorescein- (FISH only) or rhodamine-labelled avidin (BrdU-FISH) as described They were incubated with the fluorescent avidin conjugate (5 pg/ml in 4SSCT 5 per cent non-fat dried milk) and washed in 4SSCT. The probe-linked fluorescence was amplified, if necessary, by incubation with biotin-labelled goatanti-avidin (5 pg/ml in 4SSCT+ 5 per cent non-fat dried milk), washed in 4SSCT, and again incubated for 20 min with the avidin conjugate (fluorescein avidin DCS, rhodamine avidin D, and anti-avidin from Vector Laboratories, Burlingame, CA). For FISH only, the DNA of the cell nuclei was counterstained with propidium iodide (5 pglml) to allow simultaneous observation of total DNA and hybridized probe. For BrdU-FISH, the slides were immersed in PBS 0.05 per cent Tween 20 (PBST), labelled with the anti-BrdU antibody (see below), and counterstained with 4',6-diamidino-2-phenylindole (DAPI; 2pg/ml). The DNA stains (both from Sigma) were carried in an antifade solution containing p-phenylenediamine dihydrochloride (Sigma) to preserve the fluorescein and/or rhodamine fluorescence during prolonged microscopy."
+
+
Bromodeoxyuridine detection Preparations of which the cellular DNA was made single-stranded were rinsed in PBST and incubated with the fluoresceinated anti-BrdU antibody
CYTOGENETIC AND KINETIC TUMOUR ANALYSIS
Table I-The distribution of the number of hybridization spots per nucleus (in per cent) after fluorescent in situ hybridization (FISH) with biotinylated chromosome 1 and 9 specific repetitive DNA probes to freshly isolated interphase nuclei of an in vivo BrdU-labelled bladder tumour. For each of the probes, 100 intact and nonoverlapping nuclei were used for analysis
0 1 2 3 4 > 4 (spots/nucleus) No. 1: No. 9:
1 4825 7 268291 0
1
0
(Becton Dickinson, Mountain View, CA; 1:5 diluted in PBS + 5 per cent non-fat dried milk) for 60 min at 37°C. After rinsing the slides twice in PBST, the nuclei were counterstained with DAPI as described above. A fluoresceinated mouse IgG, antibody was used as a negative control and did not reveal fluorescein-stained cells within the specimen.
Flow, cytomefry The DNA histogram of the AO-stained specimen was obtained as described p r e v i ~ u s l y . ' ~ Normal ~'~ peripheral blood lymphocytes served as a diploid control. The DNA index (DI) for the tumour was defined as the ratio of the mean of the G,/G, tumour peak to the mean of the G,/G, peak of the normal diploid lymphocytes. Cell measurements were carried out immediately after staining using an FC-200 flow cytometer (Ortho Diagnostic Instruments, Westwood, MA). Cell cycle analysis of the DNA histogram was performed using Multicycle (Phoenix Flow Systems, San Diego, CA). RESULTS Before applying fluorescent in situ hybridization (FISH) to the bladder tumour specimen, the accuracy and reproducibility of the double-labelling method had been tested extensively on HL60 cell line cells pulse-labelled in vitro with bromodeoxyuridine (BrdU; data not shown). Then FISH with repetitive DNA probes was used to investigate numerical chromosome aberrations in the bladder cancer cells labelled in vivo with BrdU. For each probe, 100 intact and non-overlapping nuclei were counted by two independent investigators, and the number of bright fluorescing spots per nucleus was scored. The results of the FISH analysis with the chromosome 1 and 9 specific probes are given in Table I. In this bladder tumour, 68 per cent of the
19
cells displayed monosomy 9 and 29 per cent disomy 9 (Fig. l), while the No. 1 probe demonstrated a diploid distribution pattern (Table I). In situ hybridizations with chromosome 7 and 11 specific DNA probes also revealed a diploid distribution like that seen for No. 1 (not shown). These latter hybridizations were performed because aberrations of these chromosomes have been reported in bladder cancer. ' - l o In a second series, detection ofincorporated BrdU was performed with a fluoresceinated anti-BrdU antibody and the nuclei were counterstained with DAPI. We compared standard heat denaturation with an enzymatic procedure, employing Exonuclease 111, which leaves the target DNA permanently single-stranded. For each procedure, 200 cell nuclei were counted by two investigators, and the percentage of BrdU-positive cells was calculated. Standard heat denaturation revealed 5 per cent positive cells, while the enzyme method detected 6 per cent. These numbers did not change significantly after a FISH procedure directly followed by BrdU detection (not shown). The percentage of BrdUpositivecells was in agreement with the percentageof S-phase in the DNA histogram (Fig. 3). Finally, FISH with the No. 1 and 9 probes was combined with BrdU detection (BrdU-FISH). The standard heat denaturation protocol was used for simplicity. The biotinylated probes were detected with rhodamine-labelled avidin; BrdU was visualized with the fluoresceinated antibody; and the total nuclear DNA was counterstained with DAPI. The results after analysing BrdU-positive cells for both probes are given in Table 11.The vast majority of the BrdU-positive bladder cancer cells contained monosomy 9 (75 per cent; Fig. 2), while the chromosome 1 specific probe displayed a diploid pattern (Table 11). DISCUSSION We combined an in situ hybridization procedure (FISH) with simultaneous detection of BrdU incorporated in vivo (BrdU-FISH) in the cells of a human bladder cancer in order to trace possible differences in the growth properties ofcell clones with and without a tumour specific cytogenetic abnormality. The method is fast, since it can be applied to interphase cells, and simple to perform. No culturing of the cells is required, which avoids possible bias due to selection of cells that will grow in vitro. After standard heat denaturation, in situ hybridization, and probe
20
H. VAN DEKKEN E T A L .
Fig. I-Photomicrograph demonstrating FISH with the biotinylated chromosome9 specific DNA probe to interphase nuclei of an in vivo BrdU-labelled bladder tumour. Hybridized probe was visualized with fluoresceinated avidin (yellow); total nuclear DNA was counterstained with propidium iodide (red). Solid arrows indicate cell nuclei with monosomy 9; open arrows indicate nuclei with disomy 9 ( x 100 objective)
Fig. 2-FISH with thechromosome9 specific probe to interphase nuclei ofan in V I V O BrdU-labelled bladder tumour. The incorporated BrdU was visualized with a fluoresceinated anti-BrdU antibody (green). Hybridized probe was detected with an avidin rhodamine conjugate (pink); total nuclear DNA was counterstained with DAPI (blue; x 100 objective). Left panel: in siru hybridization showing two nuclei with monosomy 9 (solid arrows) and one nucleus with disomy 9 (open arrow). Right panel: The lower monosomiccell nucleus displays the in vivo incorporated BrdU (arrow)
C Y T O G E N E T I C A N D K I N E T I C T U M O U R ANALYSIS
Table 11-The distribution of the number of hybridization spots per BrdU-positive cell nucleus (in per cent) after FISH with biotinylated chromosome 1 and 9 specific repetitive DNA probes to freshly isolated interphase nuclei of an in vivo BrdU-labelled bladder tumour. For each of the probes, 20 nuclei were used for analysis 0 1 2 3 4 >4(spots/nucleus)
No. I-BrdU+:
No.9-BrdU+:
0 0 8 5 5 10 0 75 25 0 0
0 0
m 4-
C. 3 0 . U 4
0.
I-
7l Fig. 3-Flow cytometric DNA histogram of acridine orange (A0)-stained bladder tumour cells. The position of the G,/G,peak of normal diploid lymphocyte control cells is indicated by an arrow on the X-axis. A G,/G, peak (DI = 1.0) and a G,/M/tetraploid peak (DI =2.1) can be distinguished. Cell cycle analysis revealed 80 per cent G,/G,, 6 per cent S, and 14 per cent G,/M/ tetraploid cells in this DNA histogram
detection, the cellular DNA was still sufficiently accessible to allow staining with the anti-BrdU antibody. There was no significant advantage with the Exonuclease I11 enzymatic protocol, which leaves the target DNA permanently single-stranded. Because heat denaturation is less complex, it was the preferred method. We have not seen any side effects of the intravenous infusions of BrdU in a series of more than 100 patients with colon carcinoma. This is in accordance with a recent in vivo BrdU-labelling study on prostatic cancer, in which the same dose of BrdU was given.20
21
The alpha-satellite chromosome specific DNA probes used in this study bind to late-replicating repetitive sequences. Therefore, nuclei with double the G,/G,FISH spot number can be regarded as G2/ M or tetraploid cells in our analysis. A good correlation was seen between the chromosome 1 (control) probe and the DNA histogram with respect to the percentages ofcells in the respective cell cycle phases (Tables I and 11; Fig. 3 ) . The DNA histogram demonstrated 14 per cent of the cells to be G2/M/ tetraploid (Fig. 3 ) . Thirteen per cent of 'the cells showed more than two FISH spots (7 per cent four spots) when hybridized with the No. 1 probe (Table I). For the chromosome 9 probe 68 per cent showed one spot and 29 per cent of the cells showed two No. 9 spots (Table I). The disomic No. 9 nuclei represented either G,/G, (and S-phase) diploid cells or G,/M/tetraploid tumour cells in which chromosome 9 is underrepresented, i.e., two instead of four spots. Since only 1 per cent of the cells showed three No. 9-FISH spots (0 per cent four spots), although 14 per cent are G,/M/tetraploid by flow cytometry, we conclude that chromosome 9 is underrepresented in the G2/M/tetraploid tumour cell fraction. Histopathology of the tissue specimen showed it to be composed entirely of tumour with no normal bladder tissue present (not shown). The probe spot distributions of the BrdU-positive cells (Table 11) resembled those of the total cell population (Table I). Thus, no growth advantage or disadvantage was associated with cells carrying monosomy 9 in this bladder tumour.
ACKNOWLEDGEMENT
This work was supported in part by NCI grant RO.I.CA14134. REFERENCES I . Dal Cin P, Sandberg AA, Chromosomal aspects of human oncogenesis. Crit Rev Oncogenesis 1989; 1: 113-126. 2. Cremer T, Tesin D, Hopman A H N , Manuelidis L. Rapid interphase and metaphase assessment of specific chromosomal changes in neuroectodermal tumor cells by in situ hybridization with chemically modified D N A probes. Exp Cell Rrs 1988; 176 199-220. 3. Hopman AHN, Ramaekers FCS, Raap AK, et a/. In riru hybridisation as a tool to study numerical chromosome aberrations in solid bladder turnours. Histochrrnisrry 1988; 8 9 307-316. 4. Devilee P, Thierry RF, Kievits T, ef ul. Detection of chromosome aneuploidy in interphase nuclei from human primary breast tumors using chromosome-specitic repetitive D N A probes. Cuncer Res 19x8; 48:5825-5830. 5. Van Dekken H, Bauman JGJ. A new application of in .vimhybridization: detection of numerical and structural chromosome aberrations
22
6
7
8 9
10 II
12
H. V A N D E K K E N ET A L. with a combination centromeric-telomeric DNA probe. Cyrogenet CellGenet 1988;48: 188-189. Van Dekken H, Hagenbeek A, Bauman JGJ. Detection of host cells following sex-mismatched bone marrow transplantation by Ruorescent in .siiu hybridization with a Y-chromosome specific probe. Leukrmio 1989; 3: 724 728. Cooke HJ, Hindley J, Cloning of human satellite 111 DNA: different components are on different chromosomes. Nucleic Acids Rex 1979; 6 3 177-3 197. Moyiis R K , Albright KL, Bartholdi MF. L'I ui.Human chromosomespecific repetitive DNA sequences: novel markers for genetic analysis. Chromo~omu1987; 95: 375-386. Pinkel D, Gray JW, Trask B, van den Engh G, Fuscoe J, van Dekken H. Cytogenetic analysis by in siru hybridization with fluorescently labeled nucleic acid probes. Cold Spring Harbor Symp Quanr Biol 1986:Sl: 151- 157. Sandberg AA, Turc-Carel C, Gemmill RM. Chromosomes in solid tumors and beyond. Cuncer Res 1988;48: 1049-1059. Nowell PC. Chromosomal and molecular clues to tumor progression. Semin Oncol 1989; 16 11&127. Gray JW, Pinkel D , Trask B, el al. Analytical cytology applied to detection of prognostically important cytogenetic aberrations: current status and future directions. In: Chapman JD, Peters LJ, Withers HR, eds. Prediction ofTumor Treatment Response. New York: Pergamon Press. 1989; I 1 1-126.
13. Raber MN, Barlogie B. DNA Row cytometry of human solid tumors. In: Melamed MR, Lindmo T, Mendelsohn ML, eds. Flow Cytometry and Sorting. 2nd ed. New York: Wiley-Liss, 1990; 745-754. 14. HoahinoT,NagashimaT,MurovicJ,LevinEM, LevinVA, RuppSM. Cell kineticstudiesofiusiru human brain tumors with bromodeoxyuridine. Cytumetrj 1985; 6 627-632. I S . Dolbeare F, Gray JW. Use of restriction endonucleases and exonuclease 111 to expose halogenated pyrimidines for immunochemical staining. Cjiomeiry 1988; 9 631-635. I 6 Kunicka JE, Darzynkiewicz Z, Melamed MR. DNA insiiusensitivity to denaturation: a new parameter for flow cytometry of normal human colonic epithelium and colon carcinoma. Cancer Re3 1987; 47: 3942-3947. 17 Van Dekken H, Pizzolo JG, Kelsen DP, Melamed M R . Targeted cytogenetic analysia ofgastric tumors by in situ hybridization with a set of chromosome specific DNA probes. Cancer 1990; 66 491497. 18. Van Dekken H, Pinkel D, Mullikin J, Gray JW. Enzymatic production of single-stranded DNA as a target for Ruorescence in siru hybridirdtion. Chromosomu 1988; 97: I 5. 19. Johnson G D , de C Nogeird Ardujo G N . A simple method of reducing the fading of immunofluorescence during microscopy J lmmunol Mer/iod.x 198 I , 43: 349 350 20. Nemoto R, Hattori K, Uchida K, er ol. S-phase fraction of human prostate adenocarcinoma studied with in viva bromodeoxyuridine labeling. Cancer 1990; 6 6 509-5 14.
164: 17-22 (199 1)
SIMULTANEOUS DETECTION O F FLUORESCENT IN SITU HYBRIDIZATION AND IN VIVO INCORPORATED BrdU IN A HUMAN BLADDER TUMOUR HERMAN VAN DEKKEN, EDWARD W. SCHERVISH, JOHN G. PIZZOLO, WILLIAM R. FAIR AND MYRON R. MELAMED
Department of Pathology and Urology Service, Department of Surgery. Memorial Sloan-Kettering Cancer Center, New York, N Y 10021, U.S.A. Received 4 September 1990 Accepted 14 November 1990
SUMMARY We have used fluorescent in situ hybridization and simultaneous in vivo bromodeoxyuridine labelling of a solid bladder cancer to examine tumour cell subsets for possible proliferative growth differences. In this dual-labelled preparation, most tumour cell nuclei exhibited monosomy 9, consistent with reported karyotypes of bladder cancer.' Incorporated bromodeoxyuridine was visualized with a fluoresceinated antibody in 5-6 per cent of the tumour cells, concordant with S-phase estimates by cell cycle analysis of the flow cytometric DNA histogram. A majority of the bromodeoxyuridine-positive cells also carried the monosomy 9 chromosome abnormality. This is the first report to demonstrate the feasibility of combined in situ hybridization and detection of bromodeoxyuridine incorporated in vivo in human tumour cells in order to provide information on the growth rate of specific subsets of tumour cells identified by chromosomal constitution. KEY WORDS-h
situ hybridization, in vivo BrdU labelling, fluorescence, double-labelling, monosomy 9, bladder cancer, subset analysis.
INTRODUCTION In recent years, fluorescent in situ hybridization (FISH) has emerged as a fast and powerful tool to study numerical chromosome changes in interphase nuclei of tumour cells and cell For this purpose labelled chromosome specific alpha-satellite DNA probes were applied.',' These probes bind to the (peri)centromeric repetitive sequences found on human chromosomes.8~9Owing to the repetitive nature of this DNA, the probes bind to a large target, which is easily visualized in the interphase nucleus. Although much is known about the specific chromosome aberrations in haematologic cancers, there is still little information concerning solid turnours.'.''''' The reason for this is the difficulty Addressee for correspondence: Herman van Dekken, Department of Pathology. University Hospital Nijmegen, Geert Grooteplein Zuid 24,6525 GA Nijmegen, The Netherlands.
0022-3417/91/01001746 $05.00 0 1991 byJohn Wiley&Sons, Ltd.
in preparing high-quality metaphase spreads for cytogenetic analysis. Flow cytometry, which has yielded important data on DNA ploidy measurements, cannot disclose the primary cytogenetic events. I 2,1 We have been interested in developing a method to investigate the relationship between tumour karyotype and tumour proliferation and progression, in particular with respect to tumour cell clones that carry tumour specific chromosome abnormalities. As a target for study, we have chosen bladder cancer, since a numerical chromosome aberration (monosomy 9) has been found repeatedly in this tumour'"O (Schervish et al., in preparation). By treating the tumour cells with bromodeoxyuridine (BrdU), information can be derived about the fraction of growing (i.e., DNA-synthesizing) cells within the tumour. In thispaperwedescribea method forstudying the proliferative differences in human bladder tumour I 4 3 l 5
18
H. VAN DEKKEN E T A L .
cells with and without monosomy 9. FISH with chromosome 1 and 9 specific DNA probes, labelled with biotin, were used to assess the presence of the chromosome aberration in interphase nuclei. The DNA-synthesizing cells of the tumour were labelled in vivo,just prior to surgery, by BrdU administration anddetected later bymeansofafluoresceinated antiBrdU antibody. By combining these two methods (BrdU-FISH) in a dual-parameter detection system, we were able to investigate the relationship between growth rate and the presence of monosomy 9 in interphase cells of this solid cancer. MATERIALS AND METHODS Sample preparation BrdU (200 mg/m2) was administered to the patient intravenously in 150 ml of D5W (5 per cent dextrose in H 2 0 ) over 30 min, about 1 h prior to cystoprostatectomy. This patient was one of the first in an approved clinical trial involving in vivo BrdU administration to patients with bladder cancer. Histopathology of the surgical specimen showed in situ and deeply invasive transitional cell carcinoma of the right lateral wall and base of the bladder, extending through the bladder wall and into ducts and stroma of the prostate. Multifocal carcinoma in situ was found elsewhere in the bladder on the anterior wall, left base, and dome, with superficial invasion in the dome. There was metastatic carcinoma in pelvic lymph nodes bilaterally. The specimen described in this study was taken from the invasive carcinoma of the right lateral wall of the bladder. A nuclear suspension from this fresh surgical specimen was obtained by mechanical disaggregation as described b e f ~ r e . ' ~One . ' ~ million nuclei were used for DNA flow cytometry; the rest were fixed twice in ethanol-acetic acid (3: I ; v/v) for 15 min on ice and stored in the fixative at - 20°C. Prior to in situ hybridization (ISH) and/or BrdU detection, the cells were dropped onto ethanol-cleaned microscope glass slides and air-dried at room temperature for 4-5 h. Probes andprobe labelling The chromosome 1 specific repetitive alphasatellite DNA probe was kindly supplied by Dr H. Cook (MRC, Edinburgh, U.K.).7 The alphasatellite DNA probe for chromosome 9 was a gift from Dr R. K. Moyzis (Los Alamos National Laboratory, Los Alamos, NM).8 These probes, in plasmids, were labelled with biotin- 16-dUTP
(Boehringer Mannheim, Indianapolis, IN) by nick translation (BRL, Gaithersburg, MD) according to the manufacturers' directions. All probes were stored at - 20°C. Pretreatment and in situ hybridization Before making the cellular DNA single-stranded, the slides were incubated with RNase A (Boehringer Mannheim; 200 pg/ml in 2 x SSC, pH 7; 1 x SSC is 0.15 M NaCI+O.OI 5 M sodium citrate) for 30 min at 37°C. Further pretreatment included protein removal and post-fixation with f ~ r m a l i n . 'Pro~ duction of single-stranded target DNA was accomplished either with heat d e n a t u r a t i ~ n or ~~~~'~ with an enzymatic procedure employing Exonuclease I11 (BRL).'* The slides were then dehydrated in an ethanol series. Subsequently, in situ hybridization was performed essentially as de~cribed.',~,'~ After TSH, the slides were washed under stringent conditions and then immersed in 4 x SSC+O.l per cent Tween 20, pH 7.5 (4SSCT). D N A probe detection Detection of the biotinylated probes was accomplished with fluorescein- (FISH only) or rhodamine-labelled avidin (BrdU-FISH) as described They were incubated with the fluorescent avidin conjugate (5 pg/ml in 4SSCT 5 per cent non-fat dried milk) and washed in 4SSCT. The probe-linked fluorescence was amplified, if necessary, by incubation with biotin-labelled goatanti-avidin (5 pg/ml in 4SSCT+ 5 per cent non-fat dried milk), washed in 4SSCT, and again incubated for 20 min with the avidin conjugate (fluorescein avidin DCS, rhodamine avidin D, and anti-avidin from Vector Laboratories, Burlingame, CA). For FISH only, the DNA of the cell nuclei was counterstained with propidium iodide (5 pglml) to allow simultaneous observation of total DNA and hybridized probe. For BrdU-FISH, the slides were immersed in PBS 0.05 per cent Tween 20 (PBST), labelled with the anti-BrdU antibody (see below), and counterstained with 4',6-diamidino-2-phenylindole (DAPI; 2pg/ml). The DNA stains (both from Sigma) were carried in an antifade solution containing p-phenylenediamine dihydrochloride (Sigma) to preserve the fluorescein and/or rhodamine fluorescence during prolonged microscopy."
+
+
Bromodeoxyuridine detection Preparations of which the cellular DNA was made single-stranded were rinsed in PBST and incubated with the fluoresceinated anti-BrdU antibody
CYTOGENETIC AND KINETIC TUMOUR ANALYSIS
Table I-The distribution of the number of hybridization spots per nucleus (in per cent) after fluorescent in situ hybridization (FISH) with biotinylated chromosome 1 and 9 specific repetitive DNA probes to freshly isolated interphase nuclei of an in vivo BrdU-labelled bladder tumour. For each of the probes, 100 intact and nonoverlapping nuclei were used for analysis
0 1 2 3 4 > 4 (spots/nucleus) No. 1: No. 9:
1 4825 7 268291 0
1
0
(Becton Dickinson, Mountain View, CA; 1:5 diluted in PBS + 5 per cent non-fat dried milk) for 60 min at 37°C. After rinsing the slides twice in PBST, the nuclei were counterstained with DAPI as described above. A fluoresceinated mouse IgG, antibody was used as a negative control and did not reveal fluorescein-stained cells within the specimen.
Flow, cytomefry The DNA histogram of the AO-stained specimen was obtained as described p r e v i ~ u s l y . ' ~ Normal ~'~ peripheral blood lymphocytes served as a diploid control. The DNA index (DI) for the tumour was defined as the ratio of the mean of the G,/G, tumour peak to the mean of the G,/G, peak of the normal diploid lymphocytes. Cell measurements were carried out immediately after staining using an FC-200 flow cytometer (Ortho Diagnostic Instruments, Westwood, MA). Cell cycle analysis of the DNA histogram was performed using Multicycle (Phoenix Flow Systems, San Diego, CA). RESULTS Before applying fluorescent in situ hybridization (FISH) to the bladder tumour specimen, the accuracy and reproducibility of the double-labelling method had been tested extensively on HL60 cell line cells pulse-labelled in vitro with bromodeoxyuridine (BrdU; data not shown). Then FISH with repetitive DNA probes was used to investigate numerical chromosome aberrations in the bladder cancer cells labelled in vivo with BrdU. For each probe, 100 intact and non-overlapping nuclei were counted by two independent investigators, and the number of bright fluorescing spots per nucleus was scored. The results of the FISH analysis with the chromosome 1 and 9 specific probes are given in Table I. In this bladder tumour, 68 per cent of the
19
cells displayed monosomy 9 and 29 per cent disomy 9 (Fig. l), while the No. 1 probe demonstrated a diploid distribution pattern (Table I). In situ hybridizations with chromosome 7 and 11 specific DNA probes also revealed a diploid distribution like that seen for No. 1 (not shown). These latter hybridizations were performed because aberrations of these chromosomes have been reported in bladder cancer. ' - l o In a second series, detection ofincorporated BrdU was performed with a fluoresceinated anti-BrdU antibody and the nuclei were counterstained with DAPI. We compared standard heat denaturation with an enzymatic procedure, employing Exonuclease 111, which leaves the target DNA permanently single-stranded. For each procedure, 200 cell nuclei were counted by two investigators, and the percentage of BrdU-positive cells was calculated. Standard heat denaturation revealed 5 per cent positive cells, while the enzyme method detected 6 per cent. These numbers did not change significantly after a FISH procedure directly followed by BrdU detection (not shown). The percentage of BrdUpositivecells was in agreement with the percentageof S-phase in the DNA histogram (Fig. 3). Finally, FISH with the No. 1 and 9 probes was combined with BrdU detection (BrdU-FISH). The standard heat denaturation protocol was used for simplicity. The biotinylated probes were detected with rhodamine-labelled avidin; BrdU was visualized with the fluoresceinated antibody; and the total nuclear DNA was counterstained with DAPI. The results after analysing BrdU-positive cells for both probes are given in Table 11.The vast majority of the BrdU-positive bladder cancer cells contained monosomy 9 (75 per cent; Fig. 2), while the chromosome 1 specific probe displayed a diploid pattern (Table 11). DISCUSSION We combined an in situ hybridization procedure (FISH) with simultaneous detection of BrdU incorporated in vivo (BrdU-FISH) in the cells of a human bladder cancer in order to trace possible differences in the growth properties ofcell clones with and without a tumour specific cytogenetic abnormality. The method is fast, since it can be applied to interphase cells, and simple to perform. No culturing of the cells is required, which avoids possible bias due to selection of cells that will grow in vitro. After standard heat denaturation, in situ hybridization, and probe
20
H. VAN DEKKEN E T A L .
Fig. I-Photomicrograph demonstrating FISH with the biotinylated chromosome9 specific DNA probe to interphase nuclei of an in vivo BrdU-labelled bladder tumour. Hybridized probe was visualized with fluoresceinated avidin (yellow); total nuclear DNA was counterstained with propidium iodide (red). Solid arrows indicate cell nuclei with monosomy 9; open arrows indicate nuclei with disomy 9 ( x 100 objective)
Fig. 2-FISH with thechromosome9 specific probe to interphase nuclei ofan in V I V O BrdU-labelled bladder tumour. The incorporated BrdU was visualized with a fluoresceinated anti-BrdU antibody (green). Hybridized probe was detected with an avidin rhodamine conjugate (pink); total nuclear DNA was counterstained with DAPI (blue; x 100 objective). Left panel: in siru hybridization showing two nuclei with monosomy 9 (solid arrows) and one nucleus with disomy 9 (open arrow). Right panel: The lower monosomiccell nucleus displays the in vivo incorporated BrdU (arrow)
C Y T O G E N E T I C A N D K I N E T I C T U M O U R ANALYSIS
Table 11-The distribution of the number of hybridization spots per BrdU-positive cell nucleus (in per cent) after FISH with biotinylated chromosome 1 and 9 specific repetitive DNA probes to freshly isolated interphase nuclei of an in vivo BrdU-labelled bladder tumour. For each of the probes, 20 nuclei were used for analysis 0 1 2 3 4 >4(spots/nucleus)
No. I-BrdU+:
No.9-BrdU+:
0 0 8 5 5 10 0 75 25 0 0
0 0
m 4-
C. 3 0 . U 4
0.
I-
7l Fig. 3-Flow cytometric DNA histogram of acridine orange (A0)-stained bladder tumour cells. The position of the G,/G,peak of normal diploid lymphocyte control cells is indicated by an arrow on the X-axis. A G,/G, peak (DI = 1.0) and a G,/M/tetraploid peak (DI =2.1) can be distinguished. Cell cycle analysis revealed 80 per cent G,/G,, 6 per cent S, and 14 per cent G,/M/ tetraploid cells in this DNA histogram
detection, the cellular DNA was still sufficiently accessible to allow staining with the anti-BrdU antibody. There was no significant advantage with the Exonuclease I11 enzymatic protocol, which leaves the target DNA permanently single-stranded. Because heat denaturation is less complex, it was the preferred method. We have not seen any side effects of the intravenous infusions of BrdU in a series of more than 100 patients with colon carcinoma. This is in accordance with a recent in vivo BrdU-labelling study on prostatic cancer, in which the same dose of BrdU was given.20
21
The alpha-satellite chromosome specific DNA probes used in this study bind to late-replicating repetitive sequences. Therefore, nuclei with double the G,/G,FISH spot number can be regarded as G2/ M or tetraploid cells in our analysis. A good correlation was seen between the chromosome 1 (control) probe and the DNA histogram with respect to the percentages ofcells in the respective cell cycle phases (Tables I and 11; Fig. 3 ) . The DNA histogram demonstrated 14 per cent of the cells to be G2/M/ tetraploid (Fig. 3 ) . Thirteen per cent of 'the cells showed more than two FISH spots (7 per cent four spots) when hybridized with the No. 1 probe (Table I). For the chromosome 9 probe 68 per cent showed one spot and 29 per cent of the cells showed two No. 9 spots (Table I). The disomic No. 9 nuclei represented either G,/G, (and S-phase) diploid cells or G,/M/tetraploid tumour cells in which chromosome 9 is underrepresented, i.e., two instead of four spots. Since only 1 per cent of the cells showed three No. 9-FISH spots (0 per cent four spots), although 14 per cent are G,/M/tetraploid by flow cytometry, we conclude that chromosome 9 is underrepresented in the G2/M/tetraploid tumour cell fraction. Histopathology of the tissue specimen showed it to be composed entirely of tumour with no normal bladder tissue present (not shown). The probe spot distributions of the BrdU-positive cells (Table 11) resembled those of the total cell population (Table I). Thus, no growth advantage or disadvantage was associated with cells carrying monosomy 9 in this bladder tumour.
ACKNOWLEDGEMENT
This work was supported in part by NCI grant RO.I.CA14134. REFERENCES I . Dal Cin P, Sandberg AA, Chromosomal aspects of human oncogenesis. Crit Rev Oncogenesis 1989; 1: 113-126. 2. Cremer T, Tesin D, Hopman A H N , Manuelidis L. Rapid interphase and metaphase assessment of specific chromosomal changes in neuroectodermal tumor cells by in situ hybridization with chemically modified D N A probes. Exp Cell Rrs 1988; 176 199-220. 3. Hopman AHN, Ramaekers FCS, Raap AK, et a/. In riru hybridisation as a tool to study numerical chromosome aberrations in solid bladder turnours. Histochrrnisrry 1988; 8 9 307-316. 4. Devilee P, Thierry RF, Kievits T, ef ul. Detection of chromosome aneuploidy in interphase nuclei from human primary breast tumors using chromosome-specitic repetitive D N A probes. Cuncer Res 19x8; 48:5825-5830. 5. Van Dekken H, Bauman JGJ. A new application of in .vimhybridization: detection of numerical and structural chromosome aberrations
22
6
7
8 9
10 II
12
H. V A N D E K K E N ET A L. with a combination centromeric-telomeric DNA probe. Cyrogenet CellGenet 1988;48: 188-189. Van Dekken H, Hagenbeek A, Bauman JGJ. Detection of host cells following sex-mismatched bone marrow transplantation by Ruorescent in .siiu hybridization with a Y-chromosome specific probe. Leukrmio 1989; 3: 724 728. Cooke HJ, Hindley J, Cloning of human satellite 111 DNA: different components are on different chromosomes. Nucleic Acids Rex 1979; 6 3 177-3 197. Moyiis R K , Albright KL, Bartholdi MF. L'I ui.Human chromosomespecific repetitive DNA sequences: novel markers for genetic analysis. Chromo~omu1987; 95: 375-386. Pinkel D, Gray JW, Trask B, van den Engh G, Fuscoe J, van Dekken H. Cytogenetic analysis by in siru hybridization with fluorescently labeled nucleic acid probes. Cold Spring Harbor Symp Quanr Biol 1986:Sl: 151- 157. Sandberg AA, Turc-Carel C, Gemmill RM. Chromosomes in solid tumors and beyond. Cuncer Res 1988;48: 1049-1059. Nowell PC. Chromosomal and molecular clues to tumor progression. Semin Oncol 1989; 16 11&127. Gray JW, Pinkel D , Trask B, el al. Analytical cytology applied to detection of prognostically important cytogenetic aberrations: current status and future directions. In: Chapman JD, Peters LJ, Withers HR, eds. Prediction ofTumor Treatment Response. New York: Pergamon Press. 1989; I 1 1-126.
13. Raber MN, Barlogie B. DNA Row cytometry of human solid tumors. In: Melamed MR, Lindmo T, Mendelsohn ML, eds. Flow Cytometry and Sorting. 2nd ed. New York: Wiley-Liss, 1990; 745-754. 14. HoahinoT,NagashimaT,MurovicJ,LevinEM, LevinVA, RuppSM. Cell kineticstudiesofiusiru human brain tumors with bromodeoxyuridine. Cytumetrj 1985; 6 627-632. I S . Dolbeare F, Gray JW. Use of restriction endonucleases and exonuclease 111 to expose halogenated pyrimidines for immunochemical staining. Cjiomeiry 1988; 9 631-635. I 6 Kunicka JE, Darzynkiewicz Z, Melamed MR. DNA insiiusensitivity to denaturation: a new parameter for flow cytometry of normal human colonic epithelium and colon carcinoma. Cancer Re3 1987; 47: 3942-3947. 17 Van Dekken H, Pizzolo JG, Kelsen DP, Melamed M R . Targeted cytogenetic analysia ofgastric tumors by in situ hybridization with a set of chromosome specific DNA probes. Cancer 1990; 66 491497. 18. Van Dekken H, Pinkel D, Mullikin J, Gray JW. Enzymatic production of single-stranded DNA as a target for Ruorescence in siru hybridirdtion. Chromosomu 1988; 97: I 5. 19. Johnson G D , de C Nogeird Ardujo G N . A simple method of reducing the fading of immunofluorescence during microscopy J lmmunol Mer/iod.x 198 I , 43: 349 350 20. Nemoto R, Hattori K, Uchida K, er ol. S-phase fraction of human prostate adenocarcinoma studied with in viva bromodeoxyuridine labeling. Cancer 1990; 6 6 509-5 14.
