Detection of proliferating S-phase brain tumor cells by in situ DNA replication

J Neurosurg 95:833 838, 2001 Detection of proliferating S-phase brain tumor cells by in situ DNA replication ROBERT J. WEIL, M.D., STEVEN A. TOMS, M.D., MAHLON D. JOHNSON, M.D., PH.D., AND AMANDA MEALER, B.S. Departments of Neurosurgery and Pathology, Vanderbilt University School of Medicine, Nashville, Tennessee Object. Current methods used to describe the proliferative status of brain tumors rely on labor-intensive, potentially costly procedures. This article provides a description of a rapid, inexpensive, uncomplicated technique used to identify proliferating cells in tissue obtained at the time of resection. Methods. Touch preparations of 16 fresh astrocytic tumors and four fresh healthy temporal neocortical tissue samples were obtained at the time of surgery. Slides were placed in hypotonic potassium chloride to permeabilize their membranes, incubated in nucleotide precursors, and labeled with bromodeoxyuridine; they were later examined with the aid of a fluorescence microscope. The percentage of tumor cells in the S phase increased in conjunction with the grade of tumor and corresponded with the findings of immunohistochemical staining for the cell-cycle marker MIB-1. These results were confirmed in cell culture by using normal human astrocytes and two glioma cell lines. Slides can be analyzed in as little as 30 minutes after removal of tissue during surgery. Conclusions. In this study the authors describe a simple method by which cells in the S phase of the cell cycle, which are contained in fresh tumor obtained at the time of surgery, can be labeled. This method may prove a useful adjunct to frozen-section analysis and may permit discrimination of neoplastic tissues from other tissues observed in small specimen samples. KEY WORDS glioma S phase proliferation index DNA replication proliferating cell nuclear antigen MIB-1 bromodeoxyuridine touch preparation DESPITE the need, rapid reproducible procedures designed to describe molecular disease of primary brain tumors have been slow to emerge. Traditional pathological examinations have been complemented in the past decade by immunohistochemical analyses and other methods that attempt to link the aneuploidy or proliferative status of tumor cells with their biological and clinical behavior. 3,5,8,10,12 Histological markers of proliferation, such as PCNA, MIB-1, and BrdU labeling, or flow cytometric analysis of karyotypical abnormalities, require time-consuming and expensive procedures. We have developed a rapid, inexpensive, reproducible method to identify S-phase cells specifically in fresh touch preparations obtained at surgery, without prelabeling or Abbreviations used in this paper: AA = anaplastic astrocytoma; ATP = adenosine triphosphate; BrdU = bromodeoxyuridine; CTP = cytidine triphosphate; datp = deoxyadenosine triphosphate; dctp = deoxycytidine triphosphate; dgtp = deoxyguanosine triphosphate; drntp = deoxyribonucleoside triphosphate; dutp = deoxyuridine triphosphate; FISH = fluorescence in situ hybridization; GBM = glioblastoma multiforme; GTP = guanosine triphosphate; PBS = phosphate-buffered saline; PCNA = proliferating cell nuclear antigen; rntp = ribonucleoside triphosphate; UTP = uridine triphosphate. J. Neurosurg. / Volume 95 / November, 2001 other histological manipulation. These simple cell preparations can be permeabilized and incubated in buffer containing rntps and drntps, and can be labeled with BrdU: only cells in the S phase of the cell cycle will synthesize DNA in vitro. Incubations as short as 15 minutes allow detection of S-phase cells. This technique is a rapid and specific tool that may complement both intraoperative and final pathological diagnosis. It overcomes a number of problems encountered with more detailed methods currently used to analyze the proliferative status of tumor cells, and it may prove useful in predicting the biological behavior of brain tumors. Materials and Methods Surgical Specimens Tissue samples obtained at surgery were touched lightly to clean, sterile, silica-coated glass slides. The tissues were then sent for routine pathological examination, and immunohistochemical staining for MIB-1 was performed in the pathology department according to routine protocols. All brain tissue samples were obtained as part of an institutional review board approved protocol. Slides were promptly immersed in 0.075 M KCl at 37 C for at least 5 minutes to permeabilize the cells, after which the cells were labeled according to modifications of the procedure of Mills, et al. 6,9,16 Slides are washed twice in fresh 1 PBS for 1 minute each time at 20 C, and 833

R. J. Weil, et al. TABLE 1 Labeling of cells in culture during S phase* Confluent Cells 3 H-Synchronized Cells in S Phase Cells in S Phase normal astrocytes 0.3 0.1 10.4 2.7 U373 cells 1.4 0.7 48.9 10.2 U87MG cells 1.2 0.6 62.1 13.8 *Values are expressed as mean percentage standard deviation. Cell culture conditions are described in Materials and Methods. The term 3 H- synchronized cells refers to subconfluent (50 80% confluent) cells synchronized during the S phase by application of thymidine to serum-free medium for 24 hours, as described in Materials and Methods. Sources of Supplies and Equipment The glioma cell lines U373 and U87MG were gifts from D. O Rourke of the University of Pennsylvania (Philadelphia, PA) and W. Cavenee of the Ludwig Institute (San Diego, CA), respectively. Healthy astrocytes were purchased from Clonetics (Walkersville, MD). The Lab-Tek slide chambers used to grow the cells were manufactured by Nalge Nunc International (Naperville, IL). Sucrose, NaCl, spermidine, spermine, bovine serum albumin, K-Hepes, MgCl 2, ATP, GTP, CTP, UTP, datp, dgtp, dctp, thymidine, and BrdU were all acquired from Sigma Chemical Co. (St. Louis, MO). The protease-inhibitor cocktail was obtained from Boehringer Mannheim (Indianapolis, IN) and the fluorescein-11-dutp from Roche Diagnostics (Indianapolis, IN). The fluorescent mounting medium was purchased from Dako, Corp. (Carpinteria, CA). Stat- View statistical software (Version 4.5) was provided by the SAS Institute (Cary, NC). The Axiophot microscope was obtained from Carl Zeiss (Thornwood, NY), the charge-coupled device camera from Photometrics Ltd. (Tucson, AZ), and the IP Lab Image software from Signal Analytics (Vienna, VA). were then transferred in SuNaSp/BSA buffer (0.25 M sucrose, 75 mm NaCl, 0.5 mm spermidine, 0.15 M spermine, with 3% bovine serum albumin). The slides were incubated in this buffer for 1 minute at 20 C, after which they were removed and 50 L of incubation buffer was added to each slide to cover all touch preparations. The incubation buffer consisted of three parts (by volume) of SuNaSp/BSA with protease-inhibitor cocktail to one part nucleotide mix (40 mm K-Hepes, ph 7.8; 4 mm MgCl 2 ; 3 mm ATP; 0.1 mm each of GTP, CTP, and UTP; 25 M datp; 0.1 mm each of dgtp and dctp; and 50 M fluorescein-11-dutp. After incubation newly replicating DNA stained green. The slides were incubated in a humidified chamber at 37 C for 10 minutes (during this and subsequent steps the slides should be shielded from light). The slides were washed twice in 1 PBS at room temperature for 1 minute and then fixed by immersion in 4% paraformaldehyde in PBS (ph 7.4) at 4 C for 5 minutes. Next, the slides were washed twice in 1 PBS at 20 C. They were stained with propidium iodide and RNase A in 1 PBS (1 g/ml each) for 3 minutes at 20 C. Propidium iodide stains all cells red, measures total DNA content, and serves as a counterstain for the fluorescein. 1,9 The slides were washed in deionized water for 2 minutes at 20 C. Following this procedure, the slides were covered with fluorescent mounting medium. Images of the specimens were collected using a microscope and the images were captured with the aid of a charge-coupled device camera by using imaging software; cells that were double labeled appeared yellow. For cell-culture studies, three separate fields with 200 cells each were counted and the mean and standard deviation were determined. Areas containing the greatest amount of S-phase or MIB-1 staining were selected for counting. The Student t-test and chi-square test were applied, as appropriate. 17 Statistical calculations were performed using commercially available computer software. Cell Culture The cells were grown under standard conditions at 37 C in 95% O 2 /5% CO 2. Glioma cell lines U373 and U87MG were used as a comparison. Healthy astrocytes were grown according to the supplier s specifications. Cells were grown to confluence and were then prepared in the manner described earlier. Synchronization of the S Phase Subconfluent cells grown on slide chambers at 37 C were synchronized with 3 mm thymidine for 24 hours in serum-free medium. 1,9 The cells were washed, after which 10 M BrdU was added for 4 hours to label the S-phase cells. The cells were permeabilized and labeled in the manner specified earlier. 6,9,16 Results Previous work has shown that cells frozen during the S phase can synthesize DNA in vitro when thawed and incubated in buffers containing complete nucleotide precursors. 6,7,9,16 We set out to adapt this method to detect S-phase cells in touch preparations of brain tumors obtained at surgery by virtue of their DNA replication in situ. First, normal astrocytes and glioma cell lines grown to confluence on coverslips were permeabilized and labeled. In normal astrocytes, which grow slowly, there was a very small fraction of cells in the S phase compared with the U373 and U87MG cells ( 0.3% compared with approximately 1 2%, respectively; Table 1). Synchronization of serum-starved cells with thymidine increased counts of cells in the S phase, as has previously been shown with HeLa cells. 6,9,16 Touch preparations were obtained from samples of tumor and healthy temporal lobe at surgery and were labeled (Fig. 1). The slides were compared with representative pathological sections processed for routine histological examination. Ten to thirty percent of cells in viable fractions of GBM samples (obtained from eight patients) were labeled (Fig. 2). Areas of tissue obtained from enhancing portions of the GBMs stained vigorously. This contrasts with an area of combined tumor and radiation necrosis in a patient with a recurrent GBM, which only displayed approximately 5% labeling. In this specimen, evidence of significant DNA damage was seen in response to propidium iodide staining (Fig. 2D). Samples obtained from five AAs displayed 5 to 10% staining in all areas of viable tumor (Fig. 2), whereas three low-grade astrocytomas exhibited fewer cells ( 3%) staining positively for the S phase (Fig. 3). Specimens of normal temporal lobe obtained in four patients displayed no S-phase labeling in either gray or white matter (Fig. 3). Differences in S-phase labeling between tumor types were statistically significant (Table 2), which is consistent with advancement to higher tumor grade. For four patients with AAs and six with GBMs, in whom immunohistochemical analysis was also performed, general levels of MIB-1 staining and S-phase labeling indices corresponded for each tissue or tumor type (Fig. 2 and Table 2), although there was more variability observed when the MIB-1 method was used. At a minimum, we evaluated three separate areas of tumor that had the highest levels of staining and calculated and compared S-phase labeling and MIB-1 staining. For the intratumoral group of GBMs, S-phase labeling was more consistent than MIB-1 staining (p 0.05, one-group chi- 834 J. Neurosurg. / Volume 95 / November, 2001

Detection of proliferating S-phase brain tumor cells FIG. 1. Schematic drawing showing the method used for detection of in situ DNA replication in touch preparations of fresh brain tissue. Small (0.5 1 cm 2 ) pieces of fresh tissue are lightly applied to silica-coated glass slides to impress a single layer of cells. The slide is immersed in hypotonic solution and then flooded with incubation mix (blue area). The slide is fixed, stained, and visualized with the aid of a fluorescence microscope (yellow area) to assess cells in the S phase. square test) when S-phase staining was compared with MIB-1 staining. The S-phase labeling also produced less staining variability between GBMs and AAs (p 0.05, two-group chi-square test). Discussion Cell division requires accurate and timely reproduction and segregation of chromosomes. 11,14 16 Regulation of this process is accomplished by an orderly series of events known as the cell cycle. 11,15 Two major events define the cell cycle: the reproduction of the chromosomes through DNA synthesis (S phase), and chromosome segregation during mitosis (M phase). 8,11,14,15 The S phase is regulated by a number of proteins that are essential to control DNA synthesis. 4,8,11,15 A nexus of cell-cycle regulators and signal transduction pathways has been identified, and a variety of methods has been developed to study these regulators for diagnostic purposes. At present, four basic methods for studying cell-cycle kinetics in tumor specimens are available: 1) thymidine incorporation followed by autoradiography; 2) BrdU injection in vivo followed by immunohistochemical analysis; 3) flow cytometry of fresh tissues; and 4) immunohistochemical analysis of paraffin-embedded tissues J. Neurosurg. / Volume 95 / November, 2001 performed using cell cycle specific antibodies. 1 3,8,10,19 The first three methods have several drawbacks, including variable uptake of tracers such as BrdU; cost of supplies, equipment, and personnel; and the length of time required, usually a minimum of 1 to 2 days. Immunohistochemical methods have been more widely adopted, not only because of their lower cost, but also because they can be performed serially or as part of a larger panel of antibodies. Numerous studies have shown the utility of implementing several protein markers of cell proliferation to help diagnose and predict the outcome of brain tumor (for a review, see the article by Matsutani 8 ). One such enzyme, PCNA, is expressed principally during the S phase of the cell cycle. The PCNA binds DNA polymerase and appears to play a critical role in the initiation of cell proliferation. 13 Another proliferation marker, Ki-67, which is an antigen corresponding to a nuclear nonhistone protein expressed in proliferating cells, has a similar utility; a variant, MIB-1, is a monoclonal antibody used on formalinfixed, paraffin-embedded tissues. 5,8,12 These markers are also used to label cells in the M, G 1, and G 2 phases, however, and may be deregulated in highly malignant tumors. Delay in tissue fixation also alters the results. Thus, expression levels of MIB-1 or PCNA do not exhibit a linear relationship with the number of cells found in the S phase, 835

R. J. Weil, et al. FIG. 2. Photomicrographs demonstrating different methods to determine S-phase staining of malignant gliomas. A and B: Hematoxylin and eosin stained sections of an AA (A) and a GBM with necrosis (B). C and D: The S-phase staining (positive cells stain yellow) is approximately 5 to 10% in the AA (C) and nearly 30% in the GBM (D); in the latter, near an area of necrosis seen in B (right side of panel), positively stained cells become more rare. E and F: The results of MIB-1 staining of the same tumors seen in A and B, respectively. Nuclei of positive cells stain brown. Original magnifications 40 (A), 100 (B), 200 in (C, D, and E), and 400 (F). and may underestimate the true proliferative potential of a tumor. 8,10 To overcome some of these obstacles, we have developed a rapid and simple method used to identify S-phase cells in fresh touch preparations obtained at the time of brain tumor resection. Tissue is lightly touched to sterile, silica-coated glass slides, which are then permeabilized and incubated with nucleotide precursors. 6,7,9,16 The re- 836 J. Neurosurg. / Volume 95 / November, 2001

Detection of proliferating S-phase brain tumor cells FIG. 3. Photomicrographs showing the results of S-phase staining of healthy brain tissue and a low-grade astrocytoma. A: Healthy temporal lobe. B: Low-grade astrocytoma. C: Typical view of healthy temporal lobe tissue: there is no overlap of fluorescein-stained nuclear fragments (green background) and propidium iodide (red) staining. D: One positive cell staining yellow ( 3% of cells) in a low-grade astrocytoma. The (arrow) indicates a false-positive stain in which one green nucleus abuts a second, red nucleus; there is no yellow overlap. Original magnifications 40 (A), 100 (B), and 200 (C and D). J. Neurosurg. / Volume 95 / November, 2001 maining tissue is sent for regular pathological examination. An incubation time as short as 15 minutes permits detection of labeled DNA precursors. Cells obtained ex vivo from fresh brain specimens remain viable and capable of initiating DNA replication. When incubated with buffer containing rntps and drntps, cells in the S phase continue to synthesize DNA in situ. As is the case with other pathological or labeling techniques, there is some mutability between separate regions of tumor. This is related to intrinsic tumor factors such as degree of cellularity, proliferation of stromal and enodothelial components, or areas of calcification obscuring the underlying tumor component; tumor or radiation-induced necrosis; and tissue impression and processing techniques. 18 Nevertheless, it appears that the S-phase method leads to more uniform labeling both within and between tumors. It may be that the absence of processing artifacts in fresh-touch preparations minimizes the variability in S-phase staining compared with MIB-1 labeling. The S-phase labeling method does not require additional histological processing. The results can be archived on a computerized system, and the slides can be used for additional analyses with other molecular diagnostic studies, such as FISH, because this in situ labeling techniques does not appear to inhibit later FISH probe hybridization (data not shown). The method we describe is simple, rapid, and reproducible. It utilizes reagents and equipment commonly found in most modern pathology departments. It is superior to any current method of in vivo labeling in humans because it does not require preoperative injection of precursors such as BrdU. The slides may be analyzed in as little as 30 minutes after removal of the tissue sample, which is comparable to the length of time required for frozensection analysis. Rapid identification of S-phase cells may 837

R. J. Weil, et al. TABLE 2 Labeling of tumors and brain tissues during S phase* No. of S-Phase MIB-1 Staining Tissue or Tumor Patients Labeling (no. of samples) normal temporal lobe 4 0.0 0.1 ND low-grade astrocytoma 3 1.9 1.7 1.4 1.1 (2) AA 5 7.6 3.9 7.3 5.6 (4) GBM 8 34.7 6.1 26.8 14.0 (6) * ND = not determined. Values are expressed as mean percentages standard deviation. p 0.05 when results of S-phase staining of AAs are compared with staining of low-grade astrocytomas (Student t-test). p 0.03 when results of S-phase staining of GBMs are compared with staining of AAs (Student t-test). provide significant details regarding tumor biology during surgery. It is possible that such studies may aid in the differentiation of reactive or inflammatory lesions from neoplastic ones, particularly in difficult cases in which the histological picture is unclear. Acknowledgment The helpful comments provided by P. A. Weil are appreciated. References 1. Ausubel FM, Brent R, Kingston RE, et al (eds): Short Protocols in Molecular Biology, ed 3. New York: John Wiley & Sons, 1995 2. Freeman A, Morris LS, Mills AD, et al: Minichromosome maintenance proteins as biological markers of dysplasia and malignancy. Clin Cancer Res 5:2121 2132, 1999 3. Garcia RL, Coltera MD, Gown AM: Analysis of proliferative grade using anti-pcna/cyclin monoclonal antibodies in fixed, embedded tissues. Comparison with flow cytometric analysis. Am J Pathol 134:733 739, 1989 4. Gavin KA, Kidaka M, Stillman B: Conserved initiator proteins in eukaryotes. Science 270:1667 1671, 1995 5. Kirkegaard LJ, DeRose PB, Yao B, et al: Image cytometric measurement of nuclear proliferation markers (MIB-1, PCNA) in astrocytomas. Prognostic significance. Am J Clin Pathol 109:69 74, 1998 6. Krude T, Jackman M, Pines J, et al: Cyclin-Cdk-dependent initiation of DNA replication in a human cell-free system. Cell 88:109 119, 1997 7. Leno GH, Munshi R: Initiation of DNA replication in nuclei from quiescent cells requires permeabilization of the nuclear membrane. J Cell Biol 127:5 14, 1994 8. Matsutani M: Cell kinetics, in Berger MS, Wilson CB (eds): The Gliomas. Philadelphia: WB Saunders, 1999, pp 204 209 9. Mills AD, Coleman N, Morris LS, et al: Detection of S-phase cells in tissue sections by in siu DNA replication. Nat Cell Biol 2:244 245, 2000 10. Nishizaki T, Orita T, Furutani Y, et al: Flow-cytometric DNA analysis and immunohistochemical measurement of Ki-67 and BudR labeling indices in human brain tumors. J Neurosurg 70:379 384, 1989 11. Nurse P: Ordering S phase and M phase in the cell cycle. Cell 79:547 550, 1994 12. Perry A, Jenkins RB, O Fallon JR, et al: Clinicopathologic study of 85 similarly treated patients with anaplastic astrocytic tumors. An analysis of DNA content (ploidy), cellular proliferation, and p53 expression. Cancer 86:672 683, 1999 13. Prelich G, Kostura M, Marshak DR, et al: The cell-cycle regulated proliferating cell nuclear antigen is required for SV40 DNA replication in vitro. Nature 326:471 475, 1987 14. Smith JS, Boeke JD: Transcription. Is S phase important for transcriptional silencing? Science 291:608 609, 2001 15. Stillman B: Cell cycle control of DNA replication. Science 274: 1659 1664, 1996 16. Stoeber K, Mills AD, Kubota Y, et al: Cdc6 protein causes premature entry into S phase in a mammalian cell-free system. EMBO J 17:7219 7229, 1998 17. Swinscow TDV: Statistics at Square One. London: British Medical Association, 1980 18. Weil RJ, Wu YY, Vortmeyer AO, et al: Telomerase activity in microdissected human gliomas. Mod Pathol 12:41 46, 1999 19. Williams GH, Romanowski P, Morris L, et al: Improved cervical smear assessment using antibodies against proteins that regulate DNA replication. Proc Natl Acad Sci USA 95: 14932 14937, 1998 Manuscript received March 2, 2001. Accepted in final form July 17, 2001. This study was supported by the Vanderbilt Physician Scientist Development award and the Vanderbilt-Ingram Cancer Center to Dr. Weil. Address reprint requests to: Robert J. Weil, M.D., Department of Neurosurgery, T-4224 Medical Center North, Vanderbilt University Medical Center, Nashville, Tennessee 37232 2380. email: robert.weil@surgery.mc.vanderbilt.edu. 838 J. Neurosurg. / Volume 95 / November, 2001