High Frequency of K-Ras Mutations in Human Colorectal Hyperplastic Polyps Gut: First Published As 10.1136/Gut.40.5.660 on 1 May 1997
Total Page:16
File Type:pdf, Size:1020Kb
660 Gut 1997; 40: 660-663 High frequency of K-ras mutations in human colorectal hyperplastic polyps Gut: first published as 10.1136/gut.40.5.660 on 1 May 1997. Downloaded from K Otori, Y Oda, K Sugiyama, T Hasebe, K Mukai, T Fujii, H Tajiri, S Yoshida, S Fukushima, H Esumi Abstract plastic polyps, we evaluated K-ras mutations Background-Hyperplastic polyps are and nuclear accumulation of p53 protein, common benign colorectal polyps, and are which are common in colorectal adenomas and thought to have little association with ma- adenocarcinomas. lignant tumours in the colorectum. How- ever, several reports suggest that some hyperplastic polyps may develop into Methods colorectal neoplasms. Aim-To clarify genetic alterations in PATIENTS AND SPECIMENS colorectal hyperplastic polyps. Twenty eight colorectal epithelial polyps were Methods-Twenty eight colorectal polyps resected from patients undergoing colon- having serrated components were re- oscopy during routine clinical practice at the sected from patients endoscopically. The National Cancer Center Hospital East from K-ras gene mutations in codons 12 and 13 1992 to 1995. All specimens were fixed with were analysed by PCR-RFLP. Intra- 10% buffered formalin. The specimens were nuclear p53 protein was immunostained stained with Carszzi's haematoxylin for light by the avidin-biotin complex method. stereomicroscopy. The pit patterns were classi- Results-A mutation of the K-ras gene fied according to Kudo's classification,'2 and was detected in nine (47%) of 19 hyper- then embedded in paraffin wax. Each specimen plastic polyps, and five (56%) of nine was cut into 3 ,u.m sections, and stained with adenomas. p53 protein nuclear accumula- haematoxylin and eosin. Histological diag- tion was detected immunohistochemically noses were made using the World Health in two (22%) of nine adenomas, but not in Organisation classification.'3 any ofthe 19 hyperplastic polyps. http://gut.bmj.com/ Conclusion-Some hyperplastic polyps may be true neoplastic lesions, and could PREPARATION OF DNA SAMPLES be precursors ofmalignant neoplasia. Based on the histopathological findings in each Investigative (Gut 1997; 40: 660-663) case, polyp tissue was cut from three serial 5 Treatment Division K Otori ,um thick sections, and after deparaffinisation K Sugiyama Keywords: hyperplastic polyps, K-ras, adenoma, colon, with xylene and ethanol, the genomic DNA was H Esumi colorectal carcinogenesis. extracted from the polyp tissue by serial on September 29, 2021 by guest. Protected copyright. with proteinase K and RNase A.'4 To Pathology Division, digestions National Cancer prevent contamination, we used a new needle Center Research Colorectal hyperplastic polyps are not gener- to microdissect every time. A sample of 1-2 ,ul Institute, East ally regarded as neoplastic or premalignant of the DNA solution was used as a template for T Hasebe K Mukai lesions in the colon and rectum, and are not the polymerase chain reaction (PCR). thought to have any malignant potential. How- Department of ever, coexisting adenomatous areas and hyper- Internal Medicine, National Cancer plasia in the same polyp have been reported.'-"1 ANALYSIS OF K-ras MUTATION Center Hospital East, In addition, hyperplastic polyps are more Mutations in K-ras (codons 12 and 13) were 6-5-1 Kashiwanoha, common in colons harbouring adenomas or screened by mismatched primer mediated PCR Kashiwa, Chiba, followed 277, Japan carcinomas. This previous evidence suggests amplification for 40 cycles, by Y Oda that hyperplastic polyps are related to restriction fragment length polymorphism T Fujii colorectal neoplasms, and in certain cases, (RFLP) analysis using specific restriction endo- H Tajiri polyps may progress to cancer, following nucleases (PCR-RFLP) detected as described S Yoshida the so-called hyperplasia-adenoma-adeno- previously. 14 For codon 12, PCR products First Department of carcinoma sequence.5-7 By contrast, the encoding the wild type and mutant sequences Pathology, Osaka City adenoma-adenocarcinoma sequence is the were distinguished as 114 base pair (bp) and University Medical School, 1-4-54, major pathway of colorectal carcinogenesis, 143 bp fragments, respectively, by digestion Asahi-machi, and is thought to be associated with the ac- with the restriction enzyme MvaI (Takara, Abeno-ku, Osaka, cumulation of mutations or deletions both in Kyoto, Japan). For codon 13, the wild type was 545, Japan S Fukushima oncogenes and in tumour suppressor genes." cleaved into 125 bp and 32 bp fragments by of various and the restriction enzyme BglI (Takara, Kyoto, Correspondence to: Some mutations oncogenes Dr Hiroyasu Esumi, tumour suppressor genes may have already Japan), whereas the PCR product having the Investigative Treatment mutant could not be digested by this Division, National Cancer occurred, even in hyperplastic polyp-like early sequence Center Research Institute, adenomas in the adenoma-adenocarcinoma enzyme. Human placental DNA (Sigma, St East, 6-5-1 Kashiwanoha, sequence." To date, however, no mutation of Louis, MO, USA) was used as a wild type Kashiwa, Chiba, 277, Japan. has been in control. And mutant DNA in codon 12 derived Accepted for publication these genes reported hyperplastic 31 October 1996 polyps. To clarify the genetic changes in hyper- from the human colon adenocarcinoma cell line K-ras mutations in hyperplastic polyps 661 SW480 (American Type Culture Collection, Results Rockville, MD, USA) and DNA samples The Table shows the clinicopathological extracted from colorectal carcinomas, which characteristics of 28 colorectal polyps. The were known to harbour a heterozygous point mean age of patients (21 men and seven mutation at K-ras codon 13 (GGC to GAC; women) was 61 years (ranging from 43 to 81 Gut: first published as 10.1136/gut.40.5.660 on 1 May 1997. Downloaded from data not shown) were used as mutant controls. years). The incidence of polyps was higher in We always used aerosol resistant tips for PCR the rectum than in the colon. Histologically, 19 to prevent contamination. The mixtures with of 28 colorectal polyps were diagnosed as several ratios of mutant (SW480 DNA) to wild hyperplastic polyps, and nine of 28 as type (human placental) DNA were amplified by adenomas. Some hyperplastic polyps showed PCR to determine the sensitivity for detecting slightly elongated and stratified nuclei mutated K-ras gene. Mutation in K-ras is (Figure), but they could not be diagnosed as detectable by this method when mutated DNA serrated adenomas. Seventeen hyperplastic comprised more than 10% of genomic DNA. polyps showed small or large asteroid pits. However, two hyperplastic polyps did not fit into Kudo's classification, and were tentatively DETECTION OF p53 PROTEIN called "cerebriform". Adenomas showed All polyps were assessed for nuclear ac- various pit patterns: asteroid, oval, gyrous-like cumulation of p53 protein. Sections were de- pit, and a composite type (Table). All but the paraffinised, heated in a hot water bath at 90°C adenoma in case 1 were no more than 10 mm for 20 minutes in phosphate buffered saline in diameter. The size of the adenomas (8.2 (PBS), and then stained using a 1 :1000 (SD) 6 5 mm) was significantly larger than that dilution of rabbit polyclonal antibody to of hyperplastic polyps (5*0 (1*6 mm); p<005). human p53 protein, RSP53 (Nichirei, Tokyo, On PCR-RFLP analysis, K-ras mutations in Japan), by the avidin-biotin-peroxidase com- codon 12 were detected in eight (42%) of 19 plex method. Human lung carcinoma tissue, hyperplastic polyps and five (56%) of nine known to contain p53 protein, was used as a adenomas; a K-ras mutation in codon 13 was positive control (data not shown). When all detected in one (5%) of 19 hyperplastic polyps; epithelial cells of one crypt or more than 20% and no mutation in codon 13 was detected in of 500 epithelial cells in a section were stained adenomas. The total frequency of K-ras with the antibody, the polyp was scored as p53 mutations was nine (47%) of 19 hyperplastic positive. polyps and five (56%) ofnine adenomas. Accu- mulation of p53 protein in the nuclei was found in two (22%) of nine adenomas (cases STATISTICAL ANALYSES 1 and 4); this was not found in any of the 19 http://gut.bmj.com/ Statistical analysis was performed using hyperplastic polyps. APC gene mutations were computer assisted Student's t test with Stat searched for in the other seven colorectal View software for the Macintosh computer, hyperplastic polyps, but none was found. version 4-0 (Abacus Concepts, Berkeley, CA, USA). A value of p<005 was considered significant. Discussion It is generally accepted that hyperplatic polyps on September 29, 2021 by guest. Protected copyright. are benign lesions without malignant potential Clinicopathological data and do not develop into colorectal neoplasms. Our study is the first to identify genetic alter- Site of Size Type K-ras mutation Case Age Sex polyp (mm) ofpit (codon) ations (K-ras mutations) in hyperplastic polyps of the colon and rectum. Hyperplastic polyp: Similarly, K-ras 1 64 M A 6 a 12 mutation occurs in the hyperplastic foci of the 2 61 F A 8 a 12 pancreatic duct and hyperplastic aberrant crypt 3 57 M R 5 c - 4 71 F A 6 a 12 foci (ACF) of the colorectal epithelium, which 5 51 F A 5 a are thought to be early precursor lesions of 6 43 M R 4 a 12 7 67 M R 2 a pancreatic and colorectal cancer.15-19 K-ras 8 81 F A 4 a 12 gene mutations may be responsible for the 9 67 M S 4 a 12 10 43 M T 8 a 13 hyperplastic change, and may not contribute to 11 41 M R 5 a immediate carcinogenesis of the pancreas and 12 71 M A 4 a 13 60 M R 6 a colorectum.