- Email: [email protected]
Telomere Length on Bladder Washing Samples from Patients with Bladder Cancer Correlates with Tumor Characteristics
European Urology
European Urology 48 (2005) 432–437
Bladder Cancer
Telomere Length on Bladder Washing Samples from Patients with Bladder Cancer Correlates withTumor Characteristics Flow Cytometry Method for Quantitative Fluorescence In Situ Hybridization (Flow-FISH Technique) Jesus Fernandez-Gomeza,*, Safwan Escaf Barmadaha, David Gosalbezb, Oscar Rodriguez-Fabaa, Antonio Jalona, Roberto Gonzaleza, Teresa Garcia Mirallesb, Ana Calasc a
Department of Urology, Hospital Central of Asturias, University of Oviedo, C/Celestino Villamil s/n. 33006-Oviedo, Asturias, Spain Department of Pathology, Hospital Central of Asturias, University of Oviedo, C/Celestino Villamil s/n. 33006-Oviedo, Asturias, Spain c Cytomertry Department, University of Oviedo, C/Celestino Villamil s/n. 33006-Oviedo, Asturias, Spain b
Accepted 26 April 2005 Available online 24 May 2005
Abstract Objective: The purpose of the present study was to evaluate the length of telomeres in patients with bladder cancer using a quantitative flow cytometry (flow-FISH) technique. Methods: Bladder washing samples from 51 patients with bladder cancer were obtained immediately before transurethral resection. The average length of telomere repeats was measured by flow-FISH, as previously reported. Results were expressed in molecular equivalents of soluble fluorochrome (MESF) units. Results: Bladder washing specimens provided adequate cell numbers for flow-FISH in 49 cases. The TEL means were 1014.71, 2343.36, 5567 and 18267.57 for Ta, T1, T2 and T3/4 tumors, respectively. Regarding grade it was obtained a mean MESF value of 1379.46, 3391.29 and 15925.11 for G1, G2 and G3, respectively. ANOVA demonstrated statistically significant differences in stage (p: 0.014) and tumor grades (p: 0.012). In relation to ploidy, we found a mean MESF value of 2701.37 and 16085.44 MESF units for diploid and aneuploid cells, respectively. Significant difference (p: 0.003) was observed between both groups. Conclusion: To date, this is the first report wherein telomere length was measured using flow-FISH method in exfoliated cells in urine from patients with bladder cancer. Further investigations are required to demonstrate whether flow-FISH technique might be considered as a tumor marker of bladder cancer. # 2005 Elsevier B.V. All rights reserved. Keywords: Bladder cancer; Telomere; Flow cytometry; FISH technique
1. Introduction Telomeres have important functions in the stability and replication of chromosomes [1,2]. These functions are mediated by highly conserved repetitive DNA sequence (TTAGGG in the human) and associated proteins formed at the ends of eukaryotic chromosomes [3]. The number of telomeric repeats in human somatic * Corresponding author. E-mail address: [email protected] (J. Fernandez-Gomez).
cells decreases with cell divisions and with age [4]. Telomere shortening may act as a mitotic clock in normal somatic cells [5]. High levels of the enzyme telomerase (capable of elongating telomeres) have been found in tumor cells [6]. The gold standard for measurement of telomere length is Southern blotting which, although accurate and reproducible, is relatively cumbersome and requires a large number of cells [7]. Fluorescence in situ hybridization (FISH) of telomere repeats has also been used to calculate telomere length [4]. Flow cytometry is a well
0302-2838/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.eururo.2005.04.030
J. Fernandez-Gomez et al. / European Urology 48 (2005) 432–437
433
established method for rapid detection of fluorescence signals which can be applied to FISH studies with chromosome-specific probes, giving a quantitative value [8]. The purpose of the present study was to evaluate the length of telomeres in patients with bladder cancer using a quantitative flow cytometry (flow-FISH) technique.
2. Material and methods In a previous study we had evaluated the ability of flow-FISH analysis in 8 voided urine samples and 7 bladder washings from patients with bladder cancer, but voided urine provided insufficient cell numbers for telomere length measurement. Therefore, flowFISH analysis was performed in 60 bladder washing samples from patients either with bladder cancer (n: 51) or with non-neoplastic urological conditions (n: 9) (control group). Bladder washing samples from patients with bladder cancer were obtained immediately before transurethral resection. Following centrifugation at 300 rpm for 10 minutes, bladder washings were stored at 70 8C. The surgical pathology diagnosis of the resections was classified according to TNM pT status, tumor grade and ploidy (Table 1). The average length of telomere repeats was measured by flowFISH as previously reported [7–9], using a commercial kit (Telomere PNA kit/FITC for flow cytometry by DAKO). The core component of each Kit is the fluorescein-labeled-PNA (peptide nucleic acid) probe that is hybridized with telomere restriction fragments (TRF) and examined by fluorescence microscopy. A corresponding tube without PNA probe was included with each sample as a negative control. Briefly, cells frozen were rapidly thawed, washed and resuspended in PBS. The cells were stained with Trypan blue and counted using a Neubauer count chamber (approximately, 105 cells/ml were used per FISH assay). Fresh cell suspensions were collected in new Eppendorf tubes, phosphate-buffered saline (PBS) was added and, after centrifugation at 500 rpm for 5 minutes, supernatants were removed. The pellets were resuspended in 300 ml hybridization mixture with and without PNA probe (negative control). After incubation for 10 minutes the tubes were vortexed and placed in a water bath at 87 8C for 10 minutes with continuous shaking to denature the DNA. The tubes were placed in the dark at room temperature overnight, thereafter the cells were
Fig. 1. Examples of positive FISH (1A, patient 11+) or negative hydridization results (1B, patient B1602-).
resuspended in 1 ml 1/10 PBS, vortexed and placed in a water bath at 40 8C for 10 minutes to remove excess, unbound PNA probe. After posthybridization washes and centrifugation at 500 rpm for 5 minutes, the nuclei were counterstained with 300 ml of Propidium iodide (PI), vortexed and then kept in the dark at 4 8C for 4– 24 hours. Positive hybridization was detected using a fluorescence microscope (Leica DMP XA 100) and special filters for FITC (fluorescein isothiocyanate) and PI (Fig. 1).
Table 1 TEL mean (MESF units), intervals and number of cases according to stage and grade (n: number of cases; d: number of diploid cases) Grade
G1 G2 G3 Total a
Stage
Mean N (d) Mean N (d) Mean N (d) Mean N (d)
Superficial tumors
Muscle-invasive tumors
Total
Ta
T1
T2
T3-4
889.2 (318–1800) 10a (10) 1106.2 (892–1291) 4 (4) – – 1014.71 (318–1800) 14 (14)
2717.3 (1820–3378) 3 (3) 2241.36 (1599–3222) 11 (11) – – 2343.35 (1599–3378) 14 (14)
– – 5462.27 (3113–8700) 11 (9) 5951 (5157–6376) 3 (2) 5567 (3113–8700) 14 (11)
– – 2400 1 (0) 20912 (5502–89214) 6 (1) 18268 (2400–89214) 7 (1)
2 false negative results occurred in patients with superficial pTaG1 diploid tumors.
1379.4 (318–3378) 13a (13) 3391.29 (892–8700) 27 (24) 15925.1 (5157–89214) 9 (3) 5159.67 (318–89214) 49a (40)
434
J. Fernandez-Gomez et al. / European Urology 48 (2005) 432–437
Fig. 2. Telomere fluorescence in the tumor cells following hybridization with FITC and calibration curves of molecular equivalents of soluble fluorchrome (MESF) units.
Finally, detection analysis of ploidy and telomere fluorescence was achieved through the use of a cytometer (Cytoron AbsoluteOrtho Diagnostic System). Excitation of the fluorochrome was performed by a 488 nm Argon laser. After flow cytometer acquisition, an electronic gate around the G0/G1 cell cycle phase was drawn to analyse only G0/G1 cell population, displaying a histogram for fluorescence gated on this population. For each sample, duplicates were analysed recording the mean fluorescence intensity for both PNA probe and control samples. Moreover, if the sample contained aneuploid cells evaluation was performed in both diploid and aneuploid G0/G1 cell populations. The telomere length (TEL) was defined as the mean fluorescence signal in PNA probe samples after subtraction of the control samples fluorescence signal. Telomere length results were expressed in molecular equivalents of soluble fluorochrome (MESF) units, after calibration with a suspension of 6 labeled microbead populations (Pharmigen, San Diego, CA). Each set of beads has a fixed amount of bound fluorescence directly corresponding to a certain MESF value, from which standard curves can be created to convert the fluorescent intensity of sampled cells into quantifiable units (MESF) (Fig. 2). The data obtained were analyzed using SPSS software. Hypothesis of equation of means was confirmed by criterion of Mann-Whitney test (2 groups) or Kruskal-Wallis test (>2 groups). Results were considered significant at the value p < 0.05.
samples from patients with bladder cancer had mean MESF (M) value of 5159.67 (318-89,214) (n: 49). When we studied the stage of the tumors we found a mean MESF value of 1014.71, 2343.36, 5567 and 18267.57 for Ta, T1, T2 and T3/4 tumors, respectively. ANOVA demonstrated statistically significant differences (p: 0.014) between stage groups in TELs which were lower in patients with superficial tumors (Table 1). Regarding grade (Table 1) it was obtained a mean MESF value of 1379.46, 3391.29 and 15925.11 for G1, G2 and G3, respectively. Analysis of variance demonstrated significant differences (p: 0.012) in the TEL means between tumor grades. We combined stage and grade of the tumors to analyse telomere length in each group. The mean TELs were 1014.7, 2343.35, 5207.08 and 15925.11 for TaG1/2, T1G1/2, T2-4G2 and T2-4G3, respectively (T1G3 tumors were not included in our series) (Fig. 3). Also, ANOVA showed significant differences between groups (p: 0.025). The mean TELs were 2701.37 and 16085.44 MESF units for diploid and aneuploid cells, respectively. Significant difference (p: 0.003) was observed between both groups. In diploid tumors we observed a progressive elevation in TEL means according to grade and stagegrade combination (superficial aneuploid tumors were not included in our series). Aneuploid tumors (all of them muscle-invasive neoplasms) had higher TELs means than diploid tumors (Fig. 3). Multivariate regres-
3. Results Automatization of above described technique allowed us to analyse 10 samples a day for TEL measurement. Unfortunately, bladder washing specimens provided unsatisfactory cell counts in patients without bladder cancer, thus TEL measurements in the control group were not calculated. Of the 51 patients with bladder cancer, flow cytometry demonstrated that 42 samples had diploid cells and 9 had aneuploid cells. Bladder washing specimens provided adequate cell numbers for flow-FISH in 49 cases (two false negative results occurred in patients with superficial pTaG1 tumors). Cells in bladder washing
Fig. 3. MESF values are summarized within clusters defined by ploidy status. Each box within a cluster represents a category of combination of stage and grade. All aneuploid tumors were muscle invasive neoplasms. Extreme aneuploid MESF value (89,214) does not appear on plot (T2-4G3 tumor).
J. Fernandez-Gomez et al. / European Urology 48 (2005) 432–437
sion analysis, including ploidy, stage and grade, demonstrated that ploidy was the only independent factor (p: 0.003) linked to telomere length results. When we focused on diploid tumors regarding stage and grade, stage was an independent predictive factor of telomere length.
4. Discussion More than 70% of superficial bladder cancers recur after resection and approximately 15% progress to invasive cancer. Almost 25% of patients with newly diagnosed bladder cancer have muscle-invasive disease, the vast majority (85–92%) already with advanced invasion at the time of diagnosis [10]. Therefore, an accurate prognostic marker of bladder cancer has been required to allow the stratification of patients with either superficial or muscle-invasive tumors into those who may benefit from aggressive treatment and those who may be managed with conservative measures [11,12]. Telomeres are specialized structures at the ends of the eukaryotic linear chromosomes which prevent endto-end fusion of chromosomes and serve for maintaining genomic stability [13,14]. It has been shown that telomeres shorten 50 to 150 base pairs per cell division in most somatic tissues, limiting replicative potential of normal somatic cells [15], which, in general, do not express telomerase [16,17], In contrast, telomeres do not progressively shorten in germ line tissues and tumors, using various mechanisms [17,18]. In this way, telomere alterations are observed during carcinogenesis, and telomere shortening appears to be one of the earliest and most prevalent genetic alterations in various human malignancies [19–26]. After tumor initiation, the vast majority of human cancers stabilizes their telomeres either by activating the enzyme telomerase or, in a minority of cases, by an alternative pathway termed alternate lengthening of telomeres [27–29]. The most commonly used tool to estimate telomere length has traditionally been Southern analysis [4,15], which although accurate and reproducible, is relatively cumbersome and requires a relative large number of cells. Recently, a novel technique for examinating telomere length in different cell and tissue types has been described, flow-FISH (flow cytometry method for quantitative FISH), a powerful tool that produces reliable, reproducible and accurate results [7]. There are very few studies that focus on the telomere length of bladder cancer cells in later phases of the tumor process [30–32]. Most of them have been performed experimentally in different bladder cancer cell lines
435
[31,32]. To date, this is the first report wherein telomere length was measured using flow-FISH method in exfoliated cells in urine from patients with bladder cancer. Our study investigated a possible relationship between the telomere length determined in urine bladder washing specimens and traditional prognostic factor of bladder cancer (stage, grade and ploidy). Analysis of spontaneous voided urine and controls (individuals without bladder cancer) were not possible because of the low number of exfoliated cells obtained for telomere length measurement in these patients. Although small number of cells did not allow to calculate telomere length in two patients with superficial low grade tumors (TaG1), our technique demonstrated high sensitivity for intermediate and high grade. Meeker et al., using a recently developed FISH technique for direct telomere length assessment in formalin-fixed human tissue specimens, reported that 29% of precursor lesions of cancer, particularly bladder lesions, displayed marked telomere heterogeneity. These authors found abnormally short or abnormally long telomeres in bladder specimens [20]. In 1996, Kamata et al. investigated tumor tissues obtained from 21 patients together with apparently normal adjacent urothelium as controls. The mean telomere length was determined using Southern blot analysis and telomerase activity was semiquantified using polymerase chain reaction based assay. These authors reported a reduction of telomere length in tumor tissue with respect to normal urothelium with no relationship to telomerase activity. Moreover, they found that superficial bladder tumors (Ta/T1) had significantly shorter telomeres than invasive (T1b or higher) transitional carcinomas [30]. In our study, some overlap occurred in TEL results between superficial and invasive tumors and between different grades, but we found that telomere length in bladder washing samples showed a tendency to be longer in more aggressive tumors. Telomere means were significantly longer in patients with aneuploid tumors than in those with diploid tumors. Also, we found higher TELs as the stage and grade were more advanced in diploid tumors. Multivariate regression analysis demonstrated that ploidy was the only independent factor linked to telomere length results, and when we focused on diploid tumors, stage was the only predictive factor. Nevertheless, this is a preliminary study about a new method to determine telomere length on bladder washings and probably there is some bias in our series data (T1G3 not included, all aneuploid cases corresponded to muscle invasive tumors. . .). From our point of view, further investigations should be planned with appropriate number of selected cases (T1G3, aneuploid superficial tumors, diploid muscle-invasive
436
J. Fernandez-Gomez et al. / European Urology 48 (2005) 432–437
neoplasms. . .) to perform multivariate analysis with the objective of assessing whether TEL measured by flowFISH technique can be considered an independent prognostic factor of bladder cancer. In this way, it should bear in mind the importance of telomere alterations in carcinogenesis. Recently, has been reported that telomere dysfunction induces chromosomal instability [33] and that functional role of aneuploidy in cancer might be attributed to telomere elongation [34]. Additionally, future studies should investigate methods to improve Flow-FISH accuracy. Interestingly, some authors have proposed multiparameter approaches to select epithelial (cytokeratin staining) or proliferating cells (monoclonal antibodies) in urine specimens, thus improving prognostic significance of traditional flow cytometry [35,36]. Likewise, a comparative study with other method for telomere length determination as well as a connection between telomere length in tumor tissue and in urine or telomerase activity might be analyzed.
Although it has been demonstrated that telomerase may be detected in bladder cancer tissue and in the urine of patients with bladder cancer [37,38], controversial results have been reported about the correlation between telomere length and telomerase activity in hematologic tumors [39]. In the same way, Kamata et al. have not found a relationship between telomere length and telomerase activity in bladder cancer tissue [30]. Our flow-FISH procedure was relatively expensive, but technical protocols of flow-FISH have been recently revised. In this way, new flow-FISH methods are rapid, cost-effective and may be clinically applicable to the examination of a large number of samples [7,9,40,41]. In summary, telomere analysis may give insight into the biologic processes underlying the initiation and development of transitional cell carcinoma. Possibly, in the future optimized flow-FISH technique could serve as an accurate tumor marker of bladder cancer.
References [1] Blackburn EH. No end in sight. Cell 1994;77:621–3. [2] Sandell LL, Zakian VA. Lost of a yeast telomere: Arrest, recovery and chromosome loss. Cell 1993;75:729–39. [3] Moyzis RK, Buckingham JM, Carm LS, Dani M, Deaven LL, Jones MD, et al. A highly repetitive sequence (TTAGGG)n, present at the telomeres of human chromosomes. Proc Natl Acad Sci USA 1988; 85:6622–6. [4] Lansdorp PM, Verwoerd NP, van de Rijke FM, Dragowska V, Little MT, Dirks RW, et al. Heterogeneity in telomere length of human chromosomes. Human Mol Genet 1996;5(5):685–91. [5] Harkey CB, Vaziry H, Counter CM, Allsopp RC. The telomere hypothesis of cellular aging. Exp Gerontol 1992;27:375–82. [6] Kim W, Piatyszek MA, Prowse KR, Harley CB, West MD, Ho PLC, et al. Specific association of human telomerase activity with immortal cells and cancer. Science 1994;266:2011–5. [7] Lauzon W, Sanchez-Dardon J, Cameron DW, Badley AD. Flow cytometric measurement of telomere length. Cytometry 2000;42:159–64. [8] Hultdin M, Gro¨ nlund E, Norrback KF, Eriksson-Lindstro¨ m E, Just T, Roos G. Telomere analysis by florescence in situ hybridation and flow cytometry. Nucleic Acid Research 1998;26(16):3651–6. [9] Rufer N, Dragowska W, Thornbury G, Roosnek E, Lansdorp PM. Telomere length dynamics in human lymphocyte subpopulations measured by flow cytometry. Nat Biotechnol 1998;16(8):743–7. [10] Messing EM. Urothelial tumors of the urinary tract. Bladder cancer. In: Walsh PC, Retik AB, Vaughan ED, Wein AJ, editors. Campbell’s Urology. 8th ed. Philadelphia: Saunders; 2002. p. 2732–65. [11] Lokeshwar VB, Soloway MS. Current bladder tumor tests: does their projected utility fulfill clinical necessity? J Urol 2001;165(4):1067–77. [12] Stein JP, Grossfeld GD, Ginsberg DA, Esrig D, Areeman JA, Figueroa AJ, et al. Prognostic markers in bladder cancer: a contemporary review of the literature. J Urol 1998;160(3 Pt 1):645–59. [13] Lee HW, Blasco MA, Gottlieb GJ, Horner II JW., Greider CW, DePinho RA. Essential role of mouse telomerase in highly proliferative organs. Nature 1998;392(6676):569–74. [14] Meyerson M, Counter CM, Eaton EN, Ellisen LW, Steiner P, Caddle SD, et al. hEST2, the putative human telomerase catalytic subunit gene, is up-regulated in tumor cells and during immortalization. Cell 1997;90(4):785–95.
[15] De Lange T. Telomeres and senescence: ending the debate. Science 1998;279(5349):334–5. [16] Hastie ND, Dempster M, Dunlop MG, Thompson AM, Green DK, Allshire RC. Telomere reduction in human colorectal carcinoma and with ageing. Nature 1990;346(6287):866–8. [17] Ohki R, Tsurimoto T, Ishikawa F. In vitro reconstitution of the end replication problem. Mol Cell Biol 2001;21(17):5753–66. [18] Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000; 100(1):57–70. [19] Sharpless NE, DePinho RA. Telomeres stem cells, senescence, and cancer. J Clin Invest 2004;113(2):160–8. [20] Meeker AK, Argani P. Telomere shortening occurs early during breast tumorigenesis: a cause of chromosome destabilization underlying malignant transformation? J Mammary Gland Biol Neoplasia 2004; 9(3):285–96. [21] Meeker AK, Hicks JL, Platz EA, March GE, Bennett CJ, Delannoy MJ, et al. Telomere shortening is an early somatic DNA alteration in human prostate tumorigenesis. Cancer Res 2002;62(22):6405–9. [22] Al-Maghrabi J, Vorobyova L, Toi A, Chapman W, Zielenska M, Squire JA. Identification of numerical chromosomal changes detected by interphase fluorescence in situ hybridization in high-grade prostate intraepithelial neoplasia as a predictor of carcinoma. Arch Pathol Lab Med 2002;126(2):165–9. [23] van Heek NT, Meeker AK, Kern SE, Yeo CJ, Lillemoe KD, Cameron JL, et al. Telomere shortening is nearly universal in pancreatic intraepithelial neoplasia. Am J Pathol 2002;161(5):1541–7. [24] Vukovic B, Park PC, Al-Maghrabi J, Beheshti B, Sweet J, Evans A, et al. Evidence of multifocality of telomere erosion in high-grade prostatic intraepithelial neoplasia (HPIN) and concurrent carcinoma. Oncogene 2003;22(13):1978–87. [25] Plentz RR, Wiemann SU, Flemming P, Meier PN, Kubicka S, Kreipe H, et al. Telomere shortening of epithelial cells characterises the adenoma-carcinoma transition of human colorectal cancer. Gut 2003;52(9):1304–7. [26] Satyanarayana A, Manns MP, Rudolph KL. Telomeres and telomerase: a dual role in hepatocarcinogenesis. Hepatology 2004;40(2):276–83. [27] Meeker AK, De Marzo AM. Recent advances in telomere biology: implications for human cancer. Curr Opin Oncol 2004;16(1):32–8.
J. Fernandez-Gomez et al. / European Urology 48 (2005) 432–437 [28] Lengauer C, Kinzler KW, Vogelstein B. Genetic instabilities in human cancers. Nature 1998;396(6712):643–9. [29] Reddel RR. Alternative lengthening of telomeres, telomerase, and cancer. Cancer Lett 2003;194(2):155–62. [30] Kamata S, Kageyama Y, Yonese J, Oshima H. Significant telomere reduction in human superficial transitional cell carcinoma. Br J Urol 1996;78(5):704–8. [31] Kageyama Y, Kamata S, Yonese J, Oshima H. Telomere length and telomerase activity in bladder and prostate cancer cell lines. Int J Urol 1997;4(4):407–10. [32] Suleman S. Telomere length analysis as a novel diagnostic test for bladder cancer enquiries. J Interdisciplinary Stud High School Students 2003;1(1):1–5. [33] Mathieu N, Pirzio L, Freulet-Marriere MA, Desmaze C, Sabatier L. Telomeres and chromosomal instability. Cell Mol Life Sci 2004; 61(6):641–56. [34] Gagos S, Irminger-Finger I. Chromosome instability in neoplasia: chaotic roots to continuous growth. Int J Biochem Cell Biol 2005; 37(5):1014–33. [35] Hijazi A, Devonec M, Bouvier R, Revillard JP. Flow cytometry study of cytokeratin 18 expression according to tumor grade and deoxyribonucleic acid content in human bladder tumors. J Urol 1989;141(3):522–6.
437
[36] Fradet Y, Tardif M, Bourget L, Robert J. Clinical cancer progression in urinary bladder tumors evaluated by multiparameter flow cytometry with monoclonal antibodies. Laval University Urology Group Cancer Res 1990;50(2):432–7. [37] Erdem E, Dikmen G, Atsu N, Dogan P, Ozen H. Telomerase activity in diagnosis of bladder cancer. Scand J Urol Nephrol 2003;37(3): 205–9. [38] Longchampt E, Lebret T, Molinie V, Bieche I, Botto H, Lidereau R. Detection of telomerase status by semiquantitative and in situ assays, and by real-time reverse transcription-polymerase chain reaction (telomerase reverse transcriptase) assay in bladder carcinomas. BJU Int 2003;91(6):567–72. [39] Januszkiewicz D, Wysoki J, Lewandowski K, Pernak M, Nowicka K, Rembowska J, et al. Lack of correlation between telomere length and telomerase activity and expression in leukemic cells. Int J Mol Med 2003;12(6):935–8. [40] Shieh J, To C, Carramao J, Nishimura N, Maruta Y, Hashimoto Y, et al. High-throughput array production using precision glass syringes. Biotechniques 2002;32(6):1360–2. [41] Baerlocher GM, Lansdorp PM. Telomere length measurements in leukocyte subsets by automated multicolor flow-FISH. Cytometry A 2003;55(1):1–6.
European Urology 48 (2005) 432–437
Bladder Cancer
Telomere Length on Bladder Washing Samples from Patients with Bladder Cancer Correlates withTumor Characteristics Flow Cytometry Method for Quantitative Fluorescence In Situ Hybridization (Flow-FISH Technique) Jesus Fernandez-Gomeza,*, Safwan Escaf Barmadaha, David Gosalbezb, Oscar Rodriguez-Fabaa, Antonio Jalona, Roberto Gonzaleza, Teresa Garcia Mirallesb, Ana Calasc a
Department of Urology, Hospital Central of Asturias, University of Oviedo, C/Celestino Villamil s/n. 33006-Oviedo, Asturias, Spain Department of Pathology, Hospital Central of Asturias, University of Oviedo, C/Celestino Villamil s/n. 33006-Oviedo, Asturias, Spain c Cytomertry Department, University of Oviedo, C/Celestino Villamil s/n. 33006-Oviedo, Asturias, Spain b
Accepted 26 April 2005 Available online 24 May 2005
Abstract Objective: The purpose of the present study was to evaluate the length of telomeres in patients with bladder cancer using a quantitative flow cytometry (flow-FISH) technique. Methods: Bladder washing samples from 51 patients with bladder cancer were obtained immediately before transurethral resection. The average length of telomere repeats was measured by flow-FISH, as previously reported. Results were expressed in molecular equivalents of soluble fluorochrome (MESF) units. Results: Bladder washing specimens provided adequate cell numbers for flow-FISH in 49 cases. The TEL means were 1014.71, 2343.36, 5567 and 18267.57 for Ta, T1, T2 and T3/4 tumors, respectively. Regarding grade it was obtained a mean MESF value of 1379.46, 3391.29 and 15925.11 for G1, G2 and G3, respectively. ANOVA demonstrated statistically significant differences in stage (p: 0.014) and tumor grades (p: 0.012). In relation to ploidy, we found a mean MESF value of 2701.37 and 16085.44 MESF units for diploid and aneuploid cells, respectively. Significant difference (p: 0.003) was observed between both groups. Conclusion: To date, this is the first report wherein telomere length was measured using flow-FISH method in exfoliated cells in urine from patients with bladder cancer. Further investigations are required to demonstrate whether flow-FISH technique might be considered as a tumor marker of bladder cancer. # 2005 Elsevier B.V. All rights reserved. Keywords: Bladder cancer; Telomere; Flow cytometry; FISH technique
1. Introduction Telomeres have important functions in the stability and replication of chromosomes [1,2]. These functions are mediated by highly conserved repetitive DNA sequence (TTAGGG in the human) and associated proteins formed at the ends of eukaryotic chromosomes [3]. The number of telomeric repeats in human somatic * Corresponding author. E-mail address: [email protected] (J. Fernandez-Gomez).
cells decreases with cell divisions and with age [4]. Telomere shortening may act as a mitotic clock in normal somatic cells [5]. High levels of the enzyme telomerase (capable of elongating telomeres) have been found in tumor cells [6]. The gold standard for measurement of telomere length is Southern blotting which, although accurate and reproducible, is relatively cumbersome and requires a large number of cells [7]. Fluorescence in situ hybridization (FISH) of telomere repeats has also been used to calculate telomere length [4]. Flow cytometry is a well
0302-2838/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.eururo.2005.04.030
J. Fernandez-Gomez et al. / European Urology 48 (2005) 432–437
433
established method for rapid detection of fluorescence signals which can be applied to FISH studies with chromosome-specific probes, giving a quantitative value [8]. The purpose of the present study was to evaluate the length of telomeres in patients with bladder cancer using a quantitative flow cytometry (flow-FISH) technique.
2. Material and methods In a previous study we had evaluated the ability of flow-FISH analysis in 8 voided urine samples and 7 bladder washings from patients with bladder cancer, but voided urine provided insufficient cell numbers for telomere length measurement. Therefore, flowFISH analysis was performed in 60 bladder washing samples from patients either with bladder cancer (n: 51) or with non-neoplastic urological conditions (n: 9) (control group). Bladder washing samples from patients with bladder cancer were obtained immediately before transurethral resection. Following centrifugation at 300 rpm for 10 minutes, bladder washings were stored at 70 8C. The surgical pathology diagnosis of the resections was classified according to TNM pT status, tumor grade and ploidy (Table 1). The average length of telomere repeats was measured by flowFISH as previously reported [7–9], using a commercial kit (Telomere PNA kit/FITC for flow cytometry by DAKO). The core component of each Kit is the fluorescein-labeled-PNA (peptide nucleic acid) probe that is hybridized with telomere restriction fragments (TRF) and examined by fluorescence microscopy. A corresponding tube without PNA probe was included with each sample as a negative control. Briefly, cells frozen were rapidly thawed, washed and resuspended in PBS. The cells were stained with Trypan blue and counted using a Neubauer count chamber (approximately, 105 cells/ml were used per FISH assay). Fresh cell suspensions were collected in new Eppendorf tubes, phosphate-buffered saline (PBS) was added and, after centrifugation at 500 rpm for 5 minutes, supernatants were removed. The pellets were resuspended in 300 ml hybridization mixture with and without PNA probe (negative control). After incubation for 10 minutes the tubes were vortexed and placed in a water bath at 87 8C for 10 minutes with continuous shaking to denature the DNA. The tubes were placed in the dark at room temperature overnight, thereafter the cells were
Fig. 1. Examples of positive FISH (1A, patient 11+) or negative hydridization results (1B, patient B1602-).
resuspended in 1 ml 1/10 PBS, vortexed and placed in a water bath at 40 8C for 10 minutes to remove excess, unbound PNA probe. After posthybridization washes and centrifugation at 500 rpm for 5 minutes, the nuclei were counterstained with 300 ml of Propidium iodide (PI), vortexed and then kept in the dark at 4 8C for 4– 24 hours. Positive hybridization was detected using a fluorescence microscope (Leica DMP XA 100) and special filters for FITC (fluorescein isothiocyanate) and PI (Fig. 1).
Table 1 TEL mean (MESF units), intervals and number of cases according to stage and grade (n: number of cases; d: number of diploid cases) Grade
G1 G2 G3 Total a
Stage
Mean N (d) Mean N (d) Mean N (d) Mean N (d)
Superficial tumors
Muscle-invasive tumors
Total
Ta
T1
T2
T3-4
889.2 (318–1800) 10a (10) 1106.2 (892–1291) 4 (4) – – 1014.71 (318–1800) 14 (14)
2717.3 (1820–3378) 3 (3) 2241.36 (1599–3222) 11 (11) – – 2343.35 (1599–3378) 14 (14)
– – 5462.27 (3113–8700) 11 (9) 5951 (5157–6376) 3 (2) 5567 (3113–8700) 14 (11)
– – 2400 1 (0) 20912 (5502–89214) 6 (1) 18268 (2400–89214) 7 (1)
2 false negative results occurred in patients with superficial pTaG1 diploid tumors.
1379.4 (318–3378) 13a (13) 3391.29 (892–8700) 27 (24) 15925.1 (5157–89214) 9 (3) 5159.67 (318–89214) 49a (40)
434
J. Fernandez-Gomez et al. / European Urology 48 (2005) 432–437
Fig. 2. Telomere fluorescence in the tumor cells following hybridization with FITC and calibration curves of molecular equivalents of soluble fluorchrome (MESF) units.
Finally, detection analysis of ploidy and telomere fluorescence was achieved through the use of a cytometer (Cytoron AbsoluteOrtho Diagnostic System). Excitation of the fluorochrome was performed by a 488 nm Argon laser. After flow cytometer acquisition, an electronic gate around the G0/G1 cell cycle phase was drawn to analyse only G0/G1 cell population, displaying a histogram for fluorescence gated on this population. For each sample, duplicates were analysed recording the mean fluorescence intensity for both PNA probe and control samples. Moreover, if the sample contained aneuploid cells evaluation was performed in both diploid and aneuploid G0/G1 cell populations. The telomere length (TEL) was defined as the mean fluorescence signal in PNA probe samples after subtraction of the control samples fluorescence signal. Telomere length results were expressed in molecular equivalents of soluble fluorochrome (MESF) units, after calibration with a suspension of 6 labeled microbead populations (Pharmigen, San Diego, CA). Each set of beads has a fixed amount of bound fluorescence directly corresponding to a certain MESF value, from which standard curves can be created to convert the fluorescent intensity of sampled cells into quantifiable units (MESF) (Fig. 2). The data obtained were analyzed using SPSS software. Hypothesis of equation of means was confirmed by criterion of Mann-Whitney test (2 groups) or Kruskal-Wallis test (>2 groups). Results were considered significant at the value p < 0.05.
samples from patients with bladder cancer had mean MESF (M) value of 5159.67 (318-89,214) (n: 49). When we studied the stage of the tumors we found a mean MESF value of 1014.71, 2343.36, 5567 and 18267.57 for Ta, T1, T2 and T3/4 tumors, respectively. ANOVA demonstrated statistically significant differences (p: 0.014) between stage groups in TELs which were lower in patients with superficial tumors (Table 1). Regarding grade (Table 1) it was obtained a mean MESF value of 1379.46, 3391.29 and 15925.11 for G1, G2 and G3, respectively. Analysis of variance demonstrated significant differences (p: 0.012) in the TEL means between tumor grades. We combined stage and grade of the tumors to analyse telomere length in each group. The mean TELs were 1014.7, 2343.35, 5207.08 and 15925.11 for TaG1/2, T1G1/2, T2-4G2 and T2-4G3, respectively (T1G3 tumors were not included in our series) (Fig. 3). Also, ANOVA showed significant differences between groups (p: 0.025). The mean TELs were 2701.37 and 16085.44 MESF units for diploid and aneuploid cells, respectively. Significant difference (p: 0.003) was observed between both groups. In diploid tumors we observed a progressive elevation in TEL means according to grade and stagegrade combination (superficial aneuploid tumors were not included in our series). Aneuploid tumors (all of them muscle-invasive neoplasms) had higher TELs means than diploid tumors (Fig. 3). Multivariate regres-
3. Results Automatization of above described technique allowed us to analyse 10 samples a day for TEL measurement. Unfortunately, bladder washing specimens provided unsatisfactory cell counts in patients without bladder cancer, thus TEL measurements in the control group were not calculated. Of the 51 patients with bladder cancer, flow cytometry demonstrated that 42 samples had diploid cells and 9 had aneuploid cells. Bladder washing specimens provided adequate cell numbers for flow-FISH in 49 cases (two false negative results occurred in patients with superficial pTaG1 tumors). Cells in bladder washing
Fig. 3. MESF values are summarized within clusters defined by ploidy status. Each box within a cluster represents a category of combination of stage and grade. All aneuploid tumors were muscle invasive neoplasms. Extreme aneuploid MESF value (89,214) does not appear on plot (T2-4G3 tumor).
J. Fernandez-Gomez et al. / European Urology 48 (2005) 432–437
sion analysis, including ploidy, stage and grade, demonstrated that ploidy was the only independent factor (p: 0.003) linked to telomere length results. When we focused on diploid tumors regarding stage and grade, stage was an independent predictive factor of telomere length.
4. Discussion More than 70% of superficial bladder cancers recur after resection and approximately 15% progress to invasive cancer. Almost 25% of patients with newly diagnosed bladder cancer have muscle-invasive disease, the vast majority (85–92%) already with advanced invasion at the time of diagnosis [10]. Therefore, an accurate prognostic marker of bladder cancer has been required to allow the stratification of patients with either superficial or muscle-invasive tumors into those who may benefit from aggressive treatment and those who may be managed with conservative measures [11,12]. Telomeres are specialized structures at the ends of the eukaryotic linear chromosomes which prevent endto-end fusion of chromosomes and serve for maintaining genomic stability [13,14]. It has been shown that telomeres shorten 50 to 150 base pairs per cell division in most somatic tissues, limiting replicative potential of normal somatic cells [15], which, in general, do not express telomerase [16,17], In contrast, telomeres do not progressively shorten in germ line tissues and tumors, using various mechanisms [17,18]. In this way, telomere alterations are observed during carcinogenesis, and telomere shortening appears to be one of the earliest and most prevalent genetic alterations in various human malignancies [19–26]. After tumor initiation, the vast majority of human cancers stabilizes their telomeres either by activating the enzyme telomerase or, in a minority of cases, by an alternative pathway termed alternate lengthening of telomeres [27–29]. The most commonly used tool to estimate telomere length has traditionally been Southern analysis [4,15], which although accurate and reproducible, is relatively cumbersome and requires a relative large number of cells. Recently, a novel technique for examinating telomere length in different cell and tissue types has been described, flow-FISH (flow cytometry method for quantitative FISH), a powerful tool that produces reliable, reproducible and accurate results [7]. There are very few studies that focus on the telomere length of bladder cancer cells in later phases of the tumor process [30–32]. Most of them have been performed experimentally in different bladder cancer cell lines
435
[31,32]. To date, this is the first report wherein telomere length was measured using flow-FISH method in exfoliated cells in urine from patients with bladder cancer. Our study investigated a possible relationship between the telomere length determined in urine bladder washing specimens and traditional prognostic factor of bladder cancer (stage, grade and ploidy). Analysis of spontaneous voided urine and controls (individuals without bladder cancer) were not possible because of the low number of exfoliated cells obtained for telomere length measurement in these patients. Although small number of cells did not allow to calculate telomere length in two patients with superficial low grade tumors (TaG1), our technique demonstrated high sensitivity for intermediate and high grade. Meeker et al., using a recently developed FISH technique for direct telomere length assessment in formalin-fixed human tissue specimens, reported that 29% of precursor lesions of cancer, particularly bladder lesions, displayed marked telomere heterogeneity. These authors found abnormally short or abnormally long telomeres in bladder specimens [20]. In 1996, Kamata et al. investigated tumor tissues obtained from 21 patients together with apparently normal adjacent urothelium as controls. The mean telomere length was determined using Southern blot analysis and telomerase activity was semiquantified using polymerase chain reaction based assay. These authors reported a reduction of telomere length in tumor tissue with respect to normal urothelium with no relationship to telomerase activity. Moreover, they found that superficial bladder tumors (Ta/T1) had significantly shorter telomeres than invasive (T1b or higher) transitional carcinomas [30]. In our study, some overlap occurred in TEL results between superficial and invasive tumors and between different grades, but we found that telomere length in bladder washing samples showed a tendency to be longer in more aggressive tumors. Telomere means were significantly longer in patients with aneuploid tumors than in those with diploid tumors. Also, we found higher TELs as the stage and grade were more advanced in diploid tumors. Multivariate regression analysis demonstrated that ploidy was the only independent factor linked to telomere length results, and when we focused on diploid tumors, stage was the only predictive factor. Nevertheless, this is a preliminary study about a new method to determine telomere length on bladder washings and probably there is some bias in our series data (T1G3 not included, all aneuploid cases corresponded to muscle invasive tumors. . .). From our point of view, further investigations should be planned with appropriate number of selected cases (T1G3, aneuploid superficial tumors, diploid muscle-invasive
436
J. Fernandez-Gomez et al. / European Urology 48 (2005) 432–437
neoplasms. . .) to perform multivariate analysis with the objective of assessing whether TEL measured by flowFISH technique can be considered an independent prognostic factor of bladder cancer. In this way, it should bear in mind the importance of telomere alterations in carcinogenesis. Recently, has been reported that telomere dysfunction induces chromosomal instability [33] and that functional role of aneuploidy in cancer might be attributed to telomere elongation [34]. Additionally, future studies should investigate methods to improve Flow-FISH accuracy. Interestingly, some authors have proposed multiparameter approaches to select epithelial (cytokeratin staining) or proliferating cells (monoclonal antibodies) in urine specimens, thus improving prognostic significance of traditional flow cytometry [35,36]. Likewise, a comparative study with other method for telomere length determination as well as a connection between telomere length in tumor tissue and in urine or telomerase activity might be analyzed.
Although it has been demonstrated that telomerase may be detected in bladder cancer tissue and in the urine of patients with bladder cancer [37,38], controversial results have been reported about the correlation between telomere length and telomerase activity in hematologic tumors [39]. In the same way, Kamata et al. have not found a relationship between telomere length and telomerase activity in bladder cancer tissue [30]. Our flow-FISH procedure was relatively expensive, but technical protocols of flow-FISH have been recently revised. In this way, new flow-FISH methods are rapid, cost-effective and may be clinically applicable to the examination of a large number of samples [7,9,40,41]. In summary, telomere analysis may give insight into the biologic processes underlying the initiation and development of transitional cell carcinoma. Possibly, in the future optimized flow-FISH technique could serve as an accurate tumor marker of bladder cancer.
References [1] Blackburn EH. No end in sight. Cell 1994;77:621–3. [2] Sandell LL, Zakian VA. Lost of a yeast telomere: Arrest, recovery and chromosome loss. Cell 1993;75:729–39. [3] Moyzis RK, Buckingham JM, Carm LS, Dani M, Deaven LL, Jones MD, et al. A highly repetitive sequence (TTAGGG)n, present at the telomeres of human chromosomes. Proc Natl Acad Sci USA 1988; 85:6622–6. [4] Lansdorp PM, Verwoerd NP, van de Rijke FM, Dragowska V, Little MT, Dirks RW, et al. Heterogeneity in telomere length of human chromosomes. Human Mol Genet 1996;5(5):685–91. [5] Harkey CB, Vaziry H, Counter CM, Allsopp RC. The telomere hypothesis of cellular aging. Exp Gerontol 1992;27:375–82. [6] Kim W, Piatyszek MA, Prowse KR, Harley CB, West MD, Ho PLC, et al. Specific association of human telomerase activity with immortal cells and cancer. Science 1994;266:2011–5. [7] Lauzon W, Sanchez-Dardon J, Cameron DW, Badley AD. Flow cytometric measurement of telomere length. Cytometry 2000;42:159–64. [8] Hultdin M, Gro¨ nlund E, Norrback KF, Eriksson-Lindstro¨ m E, Just T, Roos G. Telomere analysis by florescence in situ hybridation and flow cytometry. Nucleic Acid Research 1998;26(16):3651–6. [9] Rufer N, Dragowska W, Thornbury G, Roosnek E, Lansdorp PM. Telomere length dynamics in human lymphocyte subpopulations measured by flow cytometry. Nat Biotechnol 1998;16(8):743–7. [10] Messing EM. Urothelial tumors of the urinary tract. Bladder cancer. In: Walsh PC, Retik AB, Vaughan ED, Wein AJ, editors. Campbell’s Urology. 8th ed. Philadelphia: Saunders; 2002. p. 2732–65. [11] Lokeshwar VB, Soloway MS. Current bladder tumor tests: does their projected utility fulfill clinical necessity? J Urol 2001;165(4):1067–77. [12] Stein JP, Grossfeld GD, Ginsberg DA, Esrig D, Areeman JA, Figueroa AJ, et al. Prognostic markers in bladder cancer: a contemporary review of the literature. J Urol 1998;160(3 Pt 1):645–59. [13] Lee HW, Blasco MA, Gottlieb GJ, Horner II JW., Greider CW, DePinho RA. Essential role of mouse telomerase in highly proliferative organs. Nature 1998;392(6676):569–74. [14] Meyerson M, Counter CM, Eaton EN, Ellisen LW, Steiner P, Caddle SD, et al. hEST2, the putative human telomerase catalytic subunit gene, is up-regulated in tumor cells and during immortalization. Cell 1997;90(4):785–95.
[15] De Lange T. Telomeres and senescence: ending the debate. Science 1998;279(5349):334–5. [16] Hastie ND, Dempster M, Dunlop MG, Thompson AM, Green DK, Allshire RC. Telomere reduction in human colorectal carcinoma and with ageing. Nature 1990;346(6287):866–8. [17] Ohki R, Tsurimoto T, Ishikawa F. In vitro reconstitution of the end replication problem. Mol Cell Biol 2001;21(17):5753–66. [18] Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000; 100(1):57–70. [19] Sharpless NE, DePinho RA. Telomeres stem cells, senescence, and cancer. J Clin Invest 2004;113(2):160–8. [20] Meeker AK, Argani P. Telomere shortening occurs early during breast tumorigenesis: a cause of chromosome destabilization underlying malignant transformation? J Mammary Gland Biol Neoplasia 2004; 9(3):285–96. [21] Meeker AK, Hicks JL, Platz EA, March GE, Bennett CJ, Delannoy MJ, et al. Telomere shortening is an early somatic DNA alteration in human prostate tumorigenesis. Cancer Res 2002;62(22):6405–9. [22] Al-Maghrabi J, Vorobyova L, Toi A, Chapman W, Zielenska M, Squire JA. Identification of numerical chromosomal changes detected by interphase fluorescence in situ hybridization in high-grade prostate intraepithelial neoplasia as a predictor of carcinoma. Arch Pathol Lab Med 2002;126(2):165–9. [23] van Heek NT, Meeker AK, Kern SE, Yeo CJ, Lillemoe KD, Cameron JL, et al. Telomere shortening is nearly universal in pancreatic intraepithelial neoplasia. Am J Pathol 2002;161(5):1541–7. [24] Vukovic B, Park PC, Al-Maghrabi J, Beheshti B, Sweet J, Evans A, et al. Evidence of multifocality of telomere erosion in high-grade prostatic intraepithelial neoplasia (HPIN) and concurrent carcinoma. Oncogene 2003;22(13):1978–87. [25] Plentz RR, Wiemann SU, Flemming P, Meier PN, Kubicka S, Kreipe H, et al. Telomere shortening of epithelial cells characterises the adenoma-carcinoma transition of human colorectal cancer. Gut 2003;52(9):1304–7. [26] Satyanarayana A, Manns MP, Rudolph KL. Telomeres and telomerase: a dual role in hepatocarcinogenesis. Hepatology 2004;40(2):276–83. [27] Meeker AK, De Marzo AM. Recent advances in telomere biology: implications for human cancer. Curr Opin Oncol 2004;16(1):32–8.
J. Fernandez-Gomez et al. / European Urology 48 (2005) 432–437 [28] Lengauer C, Kinzler KW, Vogelstein B. Genetic instabilities in human cancers. Nature 1998;396(6712):643–9. [29] Reddel RR. Alternative lengthening of telomeres, telomerase, and cancer. Cancer Lett 2003;194(2):155–62. [30] Kamata S, Kageyama Y, Yonese J, Oshima H. Significant telomere reduction in human superficial transitional cell carcinoma. Br J Urol 1996;78(5):704–8. [31] Kageyama Y, Kamata S, Yonese J, Oshima H. Telomere length and telomerase activity in bladder and prostate cancer cell lines. Int J Urol 1997;4(4):407–10. [32] Suleman S. Telomere length analysis as a novel diagnostic test for bladder cancer enquiries. J Interdisciplinary Stud High School Students 2003;1(1):1–5. [33] Mathieu N, Pirzio L, Freulet-Marriere MA, Desmaze C, Sabatier L. Telomeres and chromosomal instability. Cell Mol Life Sci 2004; 61(6):641–56. [34] Gagos S, Irminger-Finger I. Chromosome instability in neoplasia: chaotic roots to continuous growth. Int J Biochem Cell Biol 2005; 37(5):1014–33. [35] Hijazi A, Devonec M, Bouvier R, Revillard JP. Flow cytometry study of cytokeratin 18 expression according to tumor grade and deoxyribonucleic acid content in human bladder tumors. J Urol 1989;141(3):522–6.
437
[36] Fradet Y, Tardif M, Bourget L, Robert J. Clinical cancer progression in urinary bladder tumors evaluated by multiparameter flow cytometry with monoclonal antibodies. Laval University Urology Group Cancer Res 1990;50(2):432–7. [37] Erdem E, Dikmen G, Atsu N, Dogan P, Ozen H. Telomerase activity in diagnosis of bladder cancer. Scand J Urol Nephrol 2003;37(3): 205–9. [38] Longchampt E, Lebret T, Molinie V, Bieche I, Botto H, Lidereau R. Detection of telomerase status by semiquantitative and in situ assays, and by real-time reverse transcription-polymerase chain reaction (telomerase reverse transcriptase) assay in bladder carcinomas. BJU Int 2003;91(6):567–72. [39] Januszkiewicz D, Wysoki J, Lewandowski K, Pernak M, Nowicka K, Rembowska J, et al. Lack of correlation between telomere length and telomerase activity and expression in leukemic cells. Int J Mol Med 2003;12(6):935–8. [40] Shieh J, To C, Carramao J, Nishimura N, Maruta Y, Hashimoto Y, et al. High-throughput array production using precision glass syringes. Biotechniques 2002;32(6):1360–2. [41] Baerlocher GM, Lansdorp PM. Telomere length measurements in leukocyte subsets by automated multicolor flow-FISH. Cytometry A 2003;55(1):1–6.