Acta Histochem. Cytochem. Vol.31 No.3 197-202 (1998) Detection of the p53 Gene Deletion by Dual Color Fluorescence In situ Hybridization in Squamous Cell Carcinoma of the Skin Morimasa Matsuta1, Mayumi Matsuta2, Hidehiko Suzuki3, Toshihide Akasaka2 and Teruo Kagabu1 Departmentsof1 Obstetrics& Gynecology,2Dermatology, 3PlasticSurgery, Iwate Medical University, 19-1 Uchimaru, Morioka020-8505 Receivedfor publicationJanuary 5, 1998and in revisedform May11, 1998 Dual color fluorescence in situ hybridization 22% to 59% (33.9•}10.9%). The ratio be- (FISH) analysis was applied to 22 squamous tween the copy number of chromosome 17 cell carcinomas (SCC) of the skin in order to centromeres and p53 in the predominant detect deletion of the p53 tumor suppressor population of the tumor nuclei with p53 gene gene in interphase nuclei using a cosmid deletion was either 2/1 or 1/1. Moreover, probe for p53 and a centromeric probe for the major subfraction was composed of chromosome 17. Fewer p53 signals than chromosome 17 disomic cells in all cases. 17 centromeric signals and zero or one cos- This suggests that either p53 gene deletion of mid signals were observed in 5.6•}4.9% one allele in disomic cells or the loss of an (mean•}SD) of controls. Based on this, 20% entire chromosome 17 in aneusomic cells is deletion was established as the cutoff line to the main mechanism of p53 gene deletion in define deletion of the p53 gene. Twenty of SCC of the skin. Immunohistochemical p53 the 22 SCC cases demonstrated p53 gene expression was frequently associated with deletion according to this criterion. The the p53 gene deletion detected by FISH. percentage of cells with deletion ranged from Key words: Dual color fluorescence in situ hybridization, p53 deletion, Squamous cell car- cinoma, Skin I. Introduction as a surrogate of p53 mutation. Furthermore over-ex- pression of the p53 protein has been clinicopathologically Tumor suppressor gene p53 is located on chromo- correlated to various tumors [1, 7, 16]. Mutation and some 17p13.1. This gene is involved in transcriptional deletion of the p53 alleles are observed in various solid regulation and codes various cyclin dependent kinase inhi- tumors, including colon cancer, lung cancer, ovarian can- bitors [2, 4]. Many studies have implied that the p53 pro- cer, esophageal cancer, and squamous cell carcinoma of tein plays an important role in arresting cell growth in the the skin [6, 15, 17, 20, 23]. The incidence of these p53 1 phase of the cell cycle in response to DNA damage G[4, mutations and deletions is greater than 50% for each of 23], and that the biological function of p53 assures the these tumors, suggesting that alterations of the p53 gene appropriate internal environment for cell replication and play an important role in the development of tumors. p53 division [11]. The tumor suppressor genes are generally mutations can be detected by polymerase chain reaction believed to be inactivated by mutation of one allele and the single strand conformation polymorphism (SSCP-PCR) . deletion of the second. This theory, known as the •g2 hits The loss of heterozygosity (LOH) at 17p has been detected theory•h, was first proposed by Knudson [9]. Mutation by restriction fragment length polymorphism (RFLP) . and deletion of the p53 gene leads to amino acid changes, A non-isotopic in situ hybridization method has re- and results in the over-expression of mutant-type p53 pro- cently been developed called fluorescence in situ hybridi- tein due to its prolonged half-life. It is widely accepted zation (FISH) technique [3]. This technique enables that immunohistochemical p53 protein expression is used several new areas of cytogenetic investigation by allowing visual determination of the presence and normality of Correspondence to: Dr. Morimasa Matsuta, Department of Ob- specific genetic sequences in single metaphase or inter- stetrics & Gynecology, Iwate Medical University , 19-1 Uchimaru, phase cells [8, 17, 19, 20, 23]. Amplified and deleted loci Morioka 020-8505, Japan. on the target chromosomes can be visualized microscopi- 197 198 Matsuta et al. cally in single cells by FISH using nucleic acid probes pH 7.0), 2ƒÊl of salmon testis DNA (Sigma) and 1ƒÊl of specific to these loci. Since deletion of the p53 gene is biotinized probe specific for a centromeric repetitive DNA theoretically detectable by FISH in interphase nuclei, sequence on chromosome 17 (D17Z1, Oncor). This mix- allelic loss of the p53 gene is in principle detectable by FISH ture was denatured in a water bath at 73•Ž for 5min and if the loss occurs by deletion [17]. A few studies have immediately placed on ice. examined interphase cytogenetics by dual color hybridi- After mixing the p53 and centromeric probes, 10ƒÊl of zation using chromosome 17 pericentromeric- and p53 the hybridization mixture were added for each square region-specific DNA probes. Those results have suggest- centimeter of the preparation slides. The slides were then ed that the dominant mechanism involved in p53 allelic covered by a coverslip and sealed with rubber cement. loss in breast and bladder cancer is a physical deletion at Hybridization was performed at 37•Ž overnight. After 17p [17, 23]. incubation, the slides were washed in 50% forma- The present study investigates the interphase-cyto- mide/2•~SSC at 45•Ž three times each for 5 min, and then genetic application of dual color FISH to evaluate deletion in 2•~SSC at room temperature (RT) twice each for of the p53 gene in squamous cell carcinoma(SCC) of the 10min. Finally, the slides were washed in PN buffer skin in order to better understand the role of p53 deletion (0.1M NaH2PO4, pH 8, and 0.1% Nonidet P-40) at RT in this tumor. for 2min. Detection of hybridization signals II Materials and Methods The hybridized probes were detected immuno- Slides preparation histochemically. Each slide was treated for 10min with Tissue samples were obtained at surgery or by biopsy PNM (5% CarnationTM dry milk, 0.02% Na-azide in PN from twenty-two cases of SCC of the skin. Staging and buffer) and then for 30min with 1ƒÊl of Texas Red avidin grading of the tumors were performed using criteria (Vector Laboratories, Burlingame, CA) in 1ml of PBS in published by Lever et al. [13]. Touch preparations were order to detect centromeric signals. The slide was then prepared by touching the freshly cut surfaces of tumors to washed three times in PBS at RT for 3min and once in glass slides. The slides were fixed in 1% paraformalde- PNM for 10min. p53 signals were detected simultane- hyde/phosphate buffered saline, pH 7.2 (PBS) for 30 min. ously. The slide was incubated for 40min in 10ƒÊl/ml After washing in PBS, the slides were treated with pro- anti-digoxigenin-fluorescein isothiocyanate (FITC) (Boeh- teinease K (0.6ƒÊg/ml) in Tris EDTA. Following the ringer Mannheim Corporation, Indianapolis, IN) diluted wash in PBS, the slides were denatured in 70% at 5% in PNM, followed by three PN washes (3min). formamide/2•~SSC (1•~SSC is 0.15M NaCl/0.015M The counterstain was treated with 4•f-6-diamino-2-phenyl- sodium citrate, pH 7.0) for 3 min at 73•Ž. Next, the slides indole (DAPI) in antifade (100mg p-phenylene diamine were dehydrated through an ethanol series (70%, 90%, dihydrochloride/PBS, pH 8.0). 100%) and then air dried. All chemicals were purchased from Sigma Chemicals (St. Lois, MO) unless otherwise Scoring of FISH signals specified. Large cells (nuclei) and those forming groups or Hybridization efficiency was measured by the same sheets were selected for analysis, because the cells were methods in 10 lymph nodes and 5 normal skin tissues that considered to be epithelial in origin. Dissociated small were not exposed to sun of 5 volunteers to provide con- cells (nuclei) and lymphocyte-like cells were disregarded. trols. Lymphocytes on the tumor sample slides served as Copy numbers of p53 and the centromere signals of 100 internal controls. nuclei per case were counted using a Nikon fluorescent Due to instability in the cosmid signal intensity, microscope. The ratio of chromosome 17 centromeric presumably due to variability in the hybridization efficien- copy numbers to those of p53 (chromosome 17/p53) in cy, a higher concentration of proteinase K treatment (up to each nucleus hybridized with a set of p53 and chromosome 4ƒÊg/ml) was used in some cases in order to increase the 17 centromere probes was calculated. The centromeric intensity of the p53 cosmid probe signal. signals were counted using criteria previously published by Hopman et al. [5]: i.e., 1) nuclei should not be overlapped Probe preparation and hybridization 2) signal intensity should be more or less of the same Hybridizations were performed as described by homogeneous staining intensity 3) minor hybridization Kallioniemi et al. [8, 17] with slight modifications. spots, which can be recognized by their smaller size and p53 cosmid probe: Ten ƒÊl of digoxigenin-labeled p53 lower intensity, should be excluded 4) signal may only be cosmid probe (Oncor, Gaithersburg, MD: supplied in for- counted when completely separated from each other and mamide solution) were preincubated in a water bath at 5) paired (split phenomenon in G2 phase) or closely op- 37•Ž for 5min. posed spots count as one signal. Following the criteria Centromeric probe of chromosome 17: The hybridi- proposed by Sauter et al. [23], the percentage of cells with zation mixture consisted of 7ƒÊl of master mixture (5.5 ml deletions detected by FISH was defined as the fraction of formamide, 1g dextran sulfate, and 0.5ml 20•~SSC, all investigated nuclei that had fewer p53 signals than 17 Detection of p53 Deletion by FISH in SCC 199 centromeric signals and that had at least one centromeric goat serum/PBS for 20min at RT to block nonspecific signal and one cosmid signal.