518 H-ras Gene Mutations in Salivary Gland Mucoepidermoid Carcinomas Jinyoung Yoo, M.D., Ph.D.1 Robert A. Robinson, M.D., BACKGROUND. The authors’ recent investigation of salivary gland tumors in ras 2 Ph.D. 1 Department of Pathology, Catholic University, St. Vincent Hospital, Suwon, South Korea. 2 Department of Pathology, University of Iowa, Iowa City, Iowa. gene alteration has suggested that K-ras activation may not play a role in their oncogenesis but H-ras may, especially in mucoepidermoid carcinomas. A study was undertaken to assess the overall incidence of mutated H-ras genes in mucoepidermoid carcinomas and to discover its potential correlation with clinicopathologic parameters. METHODS. Fifty samples from patients with salivary gland mucoepidermoid carcinoma were analyzed for point mutations at codons 12, 13, and 61 of the H-ras gene using the polymerase chain reaction followed by automated direct sequencing methodology. RESULTS. Mutated H-ras genes were detected in 9 patients, for an overall incidence of 18% (9 of 50 patients). All but 1 of the mutations occurred at codon 12: a GGC-to-GTC transversion in 8 patients and a GGC-to-GAC transition in 1 patient, resulting in the amino acid substitution of valine and aspartic acid, respectively, for glycine. One of the samples showed concurrent mutations at codons 12 (GGC-toGTC) and 13 (GGT-to-GGA). None of the samples demonstrated mutations involving codon 61. The H-ras mutations were observed in 5% (1of 21), 17% (2 of 12), and 35% (6 of 17) of low, intermediate, and high grade lesions, respectively. CONCLUSIONS. These data suggest involvement of H-ras activation in conjunction with other yet-unknown events in the development and/or progression of mucoepidermoid carcinomas. It is noteworthy that a stepwise increase in the frequency of H-ras mutations strongly correlates with tumor grade (P ⫽ 0.017). Molecular analysis of this gene alteration may provide assistance in the determination of tumor grade and differentiation. Cancer 2000;88:518 –23. © 2000 American Cancer Society. KEYWORDS: H-ras, mutation, mucoepidermoid carcinoma, salivary gland tumor, sequencing. as genes encode for a p21ras that is known to play an important role in the regulation of normal signal transduction. Activation of the ras genes, usually occurring by point mutation within specific codons of the H-ras, K-ras, and N-ras genes, creates functional reduction of intrinsic GTPase activity of the proteins, leading to increased cell proliferation.1,2 Mutated ras genes have been identified in a variety of human cancers,1– 4 although the frequency varies widely depending on the tumor type. N-ras alteration is predominantly found in hematopoietic malignancies5 and K-ras mutations are largely detected in pancreatic, colorectal, and lung adenocarcinomas,6 –11 whereas H-ras gene mutations are rather rare but often described in urothelial neoplasms.12 There has been little interest in the molecular biologic mechanisms underlying the oncogenesis of salivary gland tumors. The stud- R The authors are grateful to Christine Bromley, Mary Maher-Sturm, and Janis Rodgers for their technical assistance, and Sonya Park for her valuable help in the preparation of the manuscript. Address for reprints: Robert A. Robinson, M.D., Ph.D., Department of Pathology, University of Iowa Hospitals and Clinics, 200 Hawkins Drive 5232A RCP, Iowa City, IA 52242-1009. Received May 24, 1999; revision received September 20, 1999; accepted November 3, 1999. © 2000 American Cancer Society H-ras in Mucoepidermoid Carcinomas/Yoo et al. ies to date are quite limited, and only a few reports of isolated cases of the alterations in ras, RB1, and p53 genes in some salivary gland tumors have been published.13,14 Of all malignant salivary gland tumors, mucoepidermoid carcinoma is the most common, accounting for up to 9% of these lesions.15 A previous investigation (submitted for publication) of ras activation in salivary gland tumors demonstrated the relevance of H-ras mutation to mucoepidermoid carcinoma, which prompted us to investigate the mutational status of this gene in a large series of the tumors to determine whether mutational activation of H-ras is an important molecular event in the development and progression of mucoepidermoid carcinomas. In the current study, 50 DNA samples of mucoepidermoid carcinomas in the salivary glands were analyzed to evaluate the frequency and pattern of alterations of the H-ras gene and to discover a possible correlation between the mutation and clinical or morphologic parameters. DNA was isolated from archival paraffin-embedded tissues. After amplification of the first and second exons of the H-ras gene, containing codons 12/13 and 61, respectively, by using the polymerase chain reaction (PCR), mutational activation was investigated with automated direct sequencing analysis. MATERIALS AND METHODS Patients and Tissue Specimens Fifty cases of mucoepidermoid carcinoma were collected; these included 11 mucoepidermoid carcinomas described in a previous investigation (submitted for publication), all of which were also investigated for the presence of H-ras mutation at codon 61 in the current study. All specimens were obtained from patients undergoing either excisions or biopsies for their tumor at the University of Iowa Hospitals and Clinics. None of them had received therapy before surgery. Clinical and follow-up information was obtained from the patients’ charts and supplemented with data from the tumor registry. The important clinicopathologic characteristics of the patients are summarized in Table 1. There were 18 female and 32 male patients; their ages at the time of diagnosis were between 10 and 85 years. The various sites included parotid glands in 19 cases; palate in 7 cases; submandibular glands in 6 cases; maxillary sinus in 5 cases; floor of the mouth in 4 cases; retromolar trigone in 2 cases; buccal mucosa in 2 cases; and tongue, inferior temporal fossa, lingual plate, and parapharyngeal and pterygomandibular areas in 1 case each. The tumor sizes ranged from 0.6 to 8.0 cm in greatest dimension. 519 The tissue samples were fixed in 10% buffered formalin. After routine embedding, light microscopy led to the final diagnosis. Histologic review was performed by one of the authors (J.Y.) to confirm the diagnosis. Low grade tumors showed abundant cystic components, very rare mitoses, frequent mucin-containing cells, and minimal atypia in the squamous cells. Intermediate grade tumors had a nearly solid growth pattern, occasional mitoses, and occasional mucin-containing cells. No keratinization was seen. High grade tumors had marked nuclear atypia of the squamous cells, but mucin-containing cells could also be found. The tumors consisted of 21 low grade, 12 intermediate grade, and 17 high grade lesions. Only sections containing more than 50% tumor tissue without necrosis and hemorrhage were selected for study. DNA Extraction Microdissection and DNA extraction were performed as previously described, with minor modifications.16 Briefly, 5 sections (10 thick) were cut from the selected paraffin blocks and stained with hematoxylin and eosin. All of the slides (lacking coverslips) were used for microdissection. Precisely identified tumor tissue was obtained with care by use of a needle to assure that more than 75% of the recovered cells were tumor, as opposed to unremarkable connective tissue elements, necrotic debris, or inflammatory or hemorrhagic cell populations. No attempt was made to separate different components of the tumor. The tissue was transferred to a sterile Eppendorf tube, suspended in 100 1 of lysis buffer (50 mM Tris, l mM ethylenediaminetetraacetic acid, 0.5% Tween 20, containing proteinase K), and incubated at 55 °C overnight. Nucleic acid was purified by phenolchloroform extraction and precipitated with ethanol. Quantitation was carried out by ultraviolet (UV) absorption. Polymerase Chain Reaction A 121-base pair (bp) DNA fragment of the first exon and a 219-bp DNA fragment of the second exon of the H-ras gene, containing codons 12/13 and 61, respectively, were amplified by using as primers 5⬘-CTG-AGG-AGC-GATGAC-GGA-ATA-TAA-GC-3⬘ (sense) (Glen Research, Sterling, VA) and 5⬘-CTC-TAT-AGT-GGG-GTC-GTA-TTCGTC-CA-3⬘ (antisense), which flank codons 12 and 13, and 5⬘-TGA-GCC-CTG-TCC-TCC-TGC-AGG-ATT-C-3⬘ (sense) and 5⬘-GCC-AGC-CTC-ACG GGG-TTC-ACCTGT-A-3⬘ (antisense), which flank codon 61. The composition of the PCR reaction mixture (100⬃1) was as follows: 0.5 g of genomic DNA in 20 mM TrisHCl (pH 8.4), 50 mM KCl, 2.0 mM MgCl2, 0.2 mM each of deoxyribonucleoside triphosphate, 0.5 M each primer, and 2.5 units Platinum Taq DNA polymerase (Life Technologies, 520 CANCER February 1, 2000 / Volume 88 / Number 3 TABLE 1 List of Patients and H-ras Status Age/gender Site/size (cm) Gradea H-ras Codon Nucleotide Amino acid 54/F 71/M Lt. parapharyngeal/3.1 Lt. maxillary sinus/5.4 55/M 52/M 73/M 64/M 42/M 44/F 45/M 51/M 72/M 81/M 84/M 39/M 81/M 80/F 64/M 11/M 74/M 15/M 37/M 52/M 40/F 48/M 10/F 84/F 43/M 34/M 62/M 35/F 34/F 10/F 34/M 78/F 39/M 25/F 59/F 45/F 60/M 85/M 62/F 69/F 78/M 49/M 33/F 51/F 84/M 43/M 74/F 48/M Lt. floor of mouth/4.0 Rt. maxillary sinus/3.3 Rt. parotid/5.0 Lt. parotid/3.0 Rt. submandibular gland/4.5 Lt. parotid/3.0 Lt. parotid/2.8 Rt. retromolar trigone/1.5 Lt. parotid/5.0 Rt. parotid/2.2 Rt parotid/6.0 Tongue/1.0 Lt. submandibular gland/3.9 Rt. palate/4.0 Rt. parotid/1.8 Rt. palate/1.5 Rt. parotid/4.0 Lt. palate/2.0 Lt. parotid/8.0 Lt. parotid/1.0 Rt. retromolar trigone/0.9 Rt. buccal mucosa/1.0 Lt. palate/2.0 Rt. maxillary sinus/1.5 Rt. lingual plate/3.0 Rt. parotid/1.2 Lt. submandibular gland/3.0 Lt. parotid/4.5 Lt. parotid/0.6 Lt. palate/2.9 Rt. maxillary sinus/3.9 Rt. parotid/3.1 Palate/2.5 Lt. pterigomandibular/2.7 Rt. floor of mouth/2.0 Lt. parotid/0.6 Inf. temporal fossa/2.5 Lt. palate/0.8 Lt. submandibular gland/2.0 Rt. parotid/2.0 Lt. maxillary sinus/2.1 Floor of mouth/3.0 Submandibular gland/2.5 Lt. parotid/0.9 Floor of mouth/5.0 Rt. submandibular gland/2.9 Lt. parotid/4.0 Buccal mucosa/2.2 3 3 3 3 3 3 2 2 1 3 1 3 3 3 1 3 1 1 1 3 1 2 1 1 1 1 2 2 1 3 1 1 2 2 1 1 2 3 1 3 1 2 1 2 3 1 1 3 2 2 3 Mutant Mutant Mutant Mutant Mutant Mutant Mutant Mutant Mutant Mutant Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type Wild type 12.2 12.2 13.3 12.2 12.2 12.2 12.2 12.2 12.2 12.2 GGC 3 GTC GGC 3 GTC GGT 3 GGA GGC 3 GTC GGC 3 GTC GGC 3 GTC GGC 3 GTC GGC 3 GAC GGC 3 GTC GGC 3 GTC Gly 3 Val Gly 3 Val Gly 3 Gly Gly 3 Val Gly 3 Val Gly 3 Val Gly 3 Val Gly 3 Asp Gly 3 Val Gly 3 Val b a b 1: low; 2: intermediate; 3: high grade. Patient had two mutations. H-ras in Mucoepidermoid Carcinomas/Yoo et al. Gaithersburg, MD). Amplification was then performed on an MJ Research PTC-100 thermal cycler (Watertown, MA) for 35 cycles, which consisted of an initial denaturation at 94 °C for 1 minute; a cycle of 94 °C, 55 °C, and 72 °C for 1 minute each; and 72 °C for 5 minutes on the last cycle. Negative controls without DNA template were run in parallel to exclude contamination of reagents. As a positive control, we used T24 bladder carcinoma cell lines, which are known to have H-ras codon 12 mutation. The PCR reactions were routinely checked for the appropriately amplified DNA on an ethidium bromide– stained agarose gel under UV light transillumination. Automated Direct Sequencing The PCR products of 50 samples were purified with a QIAquick PCR purification kit (Qiagen, Valencia, CA) and then sequenced with a 373A DNA sequencer (Applied Biosystems, Foster City, CA) by using dye-primer conditions recommended by Applied Biosystems. Both strands were sequenced for each DNA analyzed, and genomic DNA from control samples was sequenced in parallel to confirm the mutations. Statistical Methods We used the Mantel–Haenszel chi-square test to establish correlation between the presence of H-ras mutation and tumor grade (Mantel, 1966). RESULTS The H-ras mutations at codons 12 and 13 as well as at codon 61 were evaluated in 50 DNA samples of mucoepidermoid carcinomas. The prevalence of mutation was 18% (in 9 of 50 samples). Single mutation was detected in eight tumors and double mutations in one. All but one of the mutations were in the second base of codon 12. Eight samples had a G-to-T transversion, substituting valine for glycine (Fig. 1), whereas one showed a G-to-A transition, replacing glycine with aspartate. One patient (Case 4) exhibited concurrent mutations involving codons 12 and 13: a GGC-to-GTC transversion at codon 12 and a GGT-to-GGA heterozygous transversion (no change in amino acid) at codon 13. None of the patients had H-ras mutations at codon 61. There was no association of the mutations with the patients’ age, gender, tumor size, or survival. However, the number of cases with mutations increased with increasing severity of histologic change. Of the H-ras mutant carcinomas, 1 was low grade (1 of 21, 5%), 2 intermediate (2 of 12; 17%), and 6 high (6 of 17; 35%) (Fig. 2). The low grade tumors had a 50:50 distribution of squamous and nonsquamous cells (mucin-producing and intermediate cells). The intermediate and high grade tumors had few (⬍10%) mucin- 521 containing cells. The correlation between the incidence of H-ras mutation and tumor grade was found to be statistically significant (P ⫽ 0.017). Clinicopathologic characteristics of the tumors are given in Table 1, along with mutation status. DISCUSSION Alterations of oncogenes and/or tumor suppressor genes have frequently been implicated in the oncogenesis of various human malignancies. However, little is known about the molecular event in salivary gland tumors. Studies of these tumors in ras gene alteration have only included scattered small series of isolated cases. No K-ras mutations were demonstrated in cases with adenoid cystic carcinoma, pleomorphic adenoma, and pleomorphic carcinoma,14,17 whereas mutated H-ras genes were present in 35% of the pleomorphic adenomas.13 In a previous study (submitted for publication), salivary gland carcinomas were investigated for the presence of point mutations in H-ras and K-ras genes, and the frequencies of the H-ras and K-ras activation were 21% and 8%, respectively. Of the 11 mucoepidermoid carcinomas included, there were detectable ras mutations in 5 tumors (45%) and K-ras activation in 1 (9%). The differing results of previously reported studies are probably due to the wide range of investigated cases, the selection of lesion types, and technical variations. In our analysis of 50 mucoepidermoid carcinomas, we found H-ras gene mutations in 9 (18%). This observation suggests that H-ras alterations may play some role in pathways leading to the development of such lesions. H-ras alterations were restricted to codons 12 and 13. All but one of the mutations were present in the second base of codon 12. With one exception (Case 34), the identical mutations were noted, resulting in the same amino acid change (glycine to valine). This preferential member of the ras gene family and the predominant type of nucleotide transition at a specific site might indicate an unusual sensitivity of this codon in mucoepidermoid carcinoma cells for tumorigenesis. The precise mechanism(s) leading to this disclosed genetic alteration, however, remain(s) to be disclosed. The incidence of H-ras gene activation in mucoepidermoid carcinomas was significantly lower in this study (18%) than in a previous series (45%) the authors performed. This discrepancy could be the result of the number of cases analyzed in the two studies and also possibly due to the larger proportion of nonmutated low grade tumors analyzed in this study. The current investigation includes more samples of mucoepidermoid carcinomas (11 vs. 50) and might by chance have detected more mutation negative tu- 522 CANCER February 1, 2000 / Volume 88 / Number 3 mors. Furthermore, more patients with lower grade tumor were included (1 vs. 21 low grade, 1 vs. 12 intermediate grade, and 9 vs. 17 high grade lesions). It was not possible to establish any correlation between the mutation status and tumor grade because of the limited sample size of our earlier study. In this extended analysis, tumors with H-ras mutations had poorer differentiation; six were high grade lesions, two intermediate, and one low. These results suggest a strong correlation between the presence of H-ras gene mutation and tumor grade in mucoepidermoid carcinomas (P ⫽ 0.017). Molecular analysis of this gene alteration may find application as a modifying role in staging. It has been reported that certain amino acid substitutions indicate a poorer prognosis, whereas hydrophobic amino acid replacements are associated with a better prognosis.18 In patients with colorectal carcinoma, K-ras 12 mutation, particularly G-to-T change (valine), was FIGURE 1. H-ras mutation using automated DNA sequencing is demonstrated. This figure, representing Patient 10, represents sequencing of the reverse strand: (A) wild-type sequence, and (B) a point mutation in the second base of codon 12 (GGC-to-GTC) of the forward strand as indicated by a C-to-A transversion (arrow) in the reverse strand. FIGURE 2. The incidence of H-ras mutations in mucoepidermoid carcinomas of different tumor grade is shown. reported to be associated with increased risk of relapse and death.19 In this study, however, the presence of H-ras 12 mutation resulting in substitution of valine did not increase the risk of recurrence. Of interest, one sample carried the sequence GGA in place of the wild-type GGC at codon 13, in addition to the H-ras 12 mutation. Even though concurrent mutations might increase the potential for progression, no difference was observed in the prognosis of patients with a single mutation versus the one whose tumor contained double mutations. The significance of concurrent mutations thus cannot be established at present. Mucoepidermoid carcinoma is composed of a dual population of cells, that is, the mucous cell and the squamous cell. In lesions of low grade, the majority of the cell population is mucous cells, whereas a predominant squamous element characterizes the high grade form. It may be that the molecular pathobiology differs between these two cell populations. However, no attempt was made in our studies to separate these two components of the tumor. It will be of interest to investigate the DNA samples from each population of lesions of different tumor grade to determine which cell component is responsible for the ras gene alteration. In summary, H-ras mutations, predominantly at codon 12, were detected in 9 tumors (18%), 8 of which had a single mutation and 1 of which contained double mutations. Mutations were observed more often in higher grade lesions. Based on these and our previous results, we conclude that mutational activation of the H-ras gene may contribute to the development of this particular salivary gland tumor. It is also noteworthy that the presence of H-ras mutations strongly correlates with tumor grade. Mutation detection may be of help in the determination of tumor grade and differentiation. H-ras in Mucoepidermoid Carcinomas/Yoo et al. REFERENCES 1. Bos JL. The ras gene family and human carcinogenesis. Mutat Res1988;195:255–71. 2. Barbacid M. ras genes. Ann Rev Biochem 1987;56:779 – 827. 3. Bishop JM. The molecular genetics of cancer. Science 1987; 235:305–11. 4. Sikora K. The molecules of cancer. The Watson Smith Lecture, 1990. J R Coll Phys Lond 1990;24:196 –205. 5. Farr CJ, Saiki RK, Erlich HA, McCormick F, Marshall CJ. Analysis of RAS gene mutations in acute myeloid leukemia by polymerase chain reaction and oligonucleotide probes. Proc Natl Acad Sci U S A 1988;85:1629 –33. 6. Grunewald K, Lyons J, Frohlich A, Feichtinger H, Weger RA, Schwab G, et al. High frequency of Ki-ras codon 12 mutations in pancreatic adenocarcinomas. Int J Cancer 1989,43: 1037– 41. 7. Tada M, Omata M, Ohto M. Clinical application of ras gene mutations for diagnosis of pancreatic adenocarcinoma. Gastroenterology 1991;100:233– 8. 8. Apple SK, Hecht JR, Novak JM. Polymerase chain reactionbased K-ras mutation detection of pancreatic adenocarcinoma in routine cytology smears. Am J Clin Pathol 1996; 105:321– 6. 9. Span M, Moerkerk PTM, De Goeij AFPM, Arends JW. A detailed analysis of K-ras point mutations in relation to tumor progression and survival in colorectal cancer patients. Int J Cancer (Pred Oncol) 1996;69:241–5. 10. Cerottini JP, Caplin S, Saraga E, Givel JC, Benhattar J. The type of K-ras mutation determines prognosis in colorectal cancer. Am J Surg 1998;175:198 –202. 11. Leohavong P, Zhu D, Melacrinos AC, DeMichele MAA, Wey- 12. 13. 14. 15. 16. 17. 18. 19. 523 ant RJ, Luketich JD, et al. Detection of low fraction K-ras mutations in primary lung tumors using a sensitive method. Int J Cancer (Pred Oncol) 1997;74:162–70. Fujita J, Kraus MH, Onove H, Srivastava SK, Ebi Y, Kitamura Y, et al. Activated H-ras oncogene in human kidney tumors. Cancer Res 1988;48:5251–5. Milasin J, Pujic N, Dedovic N, Gavric M, Vrapic V, Petrovic V, et al. H-ras gene mutations in salivary gland pleomorphic adenomas. Int J Oral Maxillofac Surg 1993;22:359 – 61. Yamamoto Y, Virmani AK, Wistuba II, McIntire D, Vuitch F, Albores-Saavedra J, et al. Loss of heterozygosity and microsatellite alterations inpS3 and RB genes in adenoid cystic carcinoma of the salivary glands. Hum Pathol 1996;27:1204 – 10. Lucas RB. Pathology of tumors of the oral tissues. New York: Churchill Livingstone, 1984. Yoo J, Lee HK, Kang CS, Park WS, Lee JY, Shim SI. p53 gene mutations and p53 protein expression in human soft tissue sarcomas. Arch Pathol Lab Med 1997;121:395–9. Yamamoto Y, Kishimoto Y, Virmani AK, Smith A, Vuitch F, Albores-Saavedra J, et al. Mutations associated with carcinomas arising from pleomorphic adenomas of the salivary glands. Hum Pathol 1996;27:782– 6. Siegfried JM, Gillespie AT, Mera R, Carsey TJ, Kechavong P, Testa JR, et al. Prognostic value of specific KRAS mutations in lung adenocarcinomas. Cancer Epidemiol Biomarkers Prev 1997;6:841–7. Andreyev HJ, Norman AR, Cunningham D, Oates JR, Clarke PA. Kirsten ras mutations in patients with colorectal cancer: the multicenter “RASCAL” study. J Natl Cancer Inst 1998;90: 675– 84.