RASSF1A Promoter Methylation and Kras2 Mutations in Non Small Cell Lung Cancer

Abstract

In the present studies, we investigated the correlation between RASSF1A promoter methylation status and Kras2 mutations in 65 primary non small cell lung cancer (NSCLC) including 33 adenocarcinomas, 12 large cell carcinomas, and 20 squamous cell carcinomas. Mutational analysis of Kras2 showed: 30% (10 of 33) of adenocarcinomas, 25% (3 of 12) of large cell carcinomas, and only 5% (1 of 20) of squamous cell carcinomas contained activated Kras2 mutation at codon 12 or 13. RASSF1A promoter region CpG island methylation was detected in adenocarcinomas (55%), large cell carcinomas (25%), and squamous cell carcinomas (25%). Interestingly, combined RASSF1A methylation and Kras2 mutation data show that only ∼7% adenocarcinomas/large cell carcinomas exhibited both KRASSF1A promoter methylation and Kras2 mutation, whereas 24% adenocarcinomas, 50% large cell carcinomas, and 70% squamous cell carcinomas showed neither Kras2 mutation nor RASSF1A promoter methylation. These results showed that the majority of the primary NSCLCs with Kras2 mutations lack RASSF1A inactivation, and both RASSF1A inactivation and Kras2 mutation events occur frequently in adenocarcinomas and large cell carcinomas. Our results indicate a trend of inverse relationship between Kras2 activation and RASSF1A promoter methylation in the majority of human lung adenocarcinomas and large cell carcinomas.

Introduction

Lung cancer is the leading cause of cancer death in men and women in the US [1]. Molecular changes in protooncogenes and tumor suppressor genes have been detected in all stages of lung tumorigenesis. Activating point mutations in the ras genes have been implicated in the pathogenesis of several human cancers including those of the lung, colon, and pancreas [2]. In particular, non small cell lung cancer (NSCLC), which accounts for 75% of lung cancers, has been shown to harbor Kras2 mutations in 30% to 50% of lung adenocarcinomas. However, Kras2 mutations were rarely detected in lung squamous cell carcinomas [3–7].

RASSF1A, a newly discovered Ras effector, is inactivated in a variety of human cancers including lung, colon, breast, prostate, thyroid, and renal cell carcinomas [8–15]. Allelic loss at 3p21.3 is one of the common events in the pathogenesis of lung cancer and many other cancers [14]. RASSF1 gene at 3p21.3 encodes two major transcripts, RASSF1A and RASSF1C. Both RASSF1A and RASSF1C proteins possess the RAS-binding domain, which binds RAS in a GTP-dependent manner in vivo and in vitro [15]. The binding of RASSF1 to Mst1, in conjunction with the binding of Nore1 (a closely related family member of RASSF1) to Mst1, appears to activate a proapoptotic pathway [15–17]. It has been proposed that Ras can activate this proapoptotic pathway by utilizing the Ras association domain of RASSF1 and Nore1. Of interest to tumorigenesis is the observation that expression of the splice variant RASSF1A is absent in lung cancer lines and in primary tumors [13,14,18,19]. One allele of RASSF1A is lost (the 3p21.3 locus) in a high percentage of all lung cancer subtypes and expression of the other allele is inhibited by promoter methylation [18–21]. The RASSF1A promoter is hypermethylated in 72% to 79% of small cell lung cancers and in 30% to 40% of NSCLCs [19,21]. Also, expression of RASSF1A in lung carcinoma cells was shown to reduce colony formation, suppress anchorage-independent growth, and inhibit tumor formation in nude mice [18]. These data suggest that RASSF1A is a tumor suppressor gene and is inactivated in a high percentage of human lung tumors. Other studies have found that RASS1A is a tumor suppressor in numerous other types of cancers, including breast, ovary [11–14], colon [22], head and neck [23], and prostate [24] cancers.

A recent observation indicates the possibility that activation of the Ras gene and hypermethylation of the RASSF1A gene are disjointed events. Van Engeland et al. [22] detected oncogenic Kras2 mutations in 87 of 222 colorectal tumors and RASSF1A promoter methylation in 45 of these tumors. Only 11 of these tumors (5%) exhibited both genetic alterations and 110 (50%) contained either Kras2 mutations or promoter methylation [22]. Similarly, a study of 110 NSCLCs exhibited Kras2 mutations in 13 tumors and RASSF1A promoter methylation in 35 tumors, and only two tumors had the same mutation whereas 46 had either a Kras2 oncogene or RASSF1A promoter methylation [19]. These data suggest that inactivation of RASSF1A occurs predominantly in colorectal tumors and adenocarcinomas of the lung without alteration of the Kras2 gene.

The purpose of this study was to examine the association between Kras2 mutation and RASSF1A methylation in human lung adenocarcinomas, large cell carcinomas, and squamous cell carcinomas, and to determine if there was an inverse correlation between genetic alterations of RASSF1A and Kras2.

Materials and Methods

Tissues and DNA Isolation

A total of 65 pathologically documented lung NSCLCs were used in the present study. These tumors and their paired normal tissues were obtained from the Cooperative Human Tissue Network of the Ohio State University Department of Pathology (Columbus, OH). A pathologist classified all tumors histopathologically. High-molecular-weight DNA was isolated from both tumor and normal tissues according to published protocols [25].

Analysis of Kras2 Mutations by Polymerase Chain Reaction (PCR) Direct Sequencing

PCR amplifications of Kras2 exon 1 from lung tumors were carried out as described previously [26]. The sequences of PCR primers of Kras2 exon 1 were Kras2-1F: 5′-TTTTTATTATAAGGCCTGCT-3′ and Kras2-1R: 5′-GTCCACAAAATGATTCTGAA-3′. The 114-bp PCR products were eluted by using QIAquick gel extraction kit (Qiagen, Valencia, CA). The 12th and 13th codon mutations were analyzed on the ABI PRISM 3700 DNA analyzer (Perkin-Elmer/Applied Biosystems, Foster City, CA).

Methylation Analyses

The methylation status of the RASSF1A promoter region was determined by chemical modification of genomic DNA with sodium bisulfite and methylation-specific PCR (MSP) was performed. Bisulfite treatment converts cytosine bases to uracil bases but has no effect on methylcytosine bases. The 144-bp fragment of the RASSF1A promoter region was amplified using: F: 5′-GTT TAG TTT GGA TTT TGG GGG AG-3′; and R: 5′-CCC RCA ACT CAA TAA ACT CAA ACT C-3′. The reaction was incubated at 95°C for 1 minute, 56°C for 1 minute, 72°C for 1 minute, and 35 cycles. The PCR product was used as a template for the MSP reaction. Primers for the unmethylated reaction were: F: 5′-GGG GTT TGT TTT GTG GTT TTG TTT-3′; and R: AAC ATA ACC CAA TTA AAC CCA TAC TTC A-3′. Primers for the methylated reaction were: F: 5′-GGG TTC GTT TTG TGG TTT CGT TC-3′; and R: 5′-TAA CCC GAT TAA ACC CGT ACT TCG-3′. The annealing temperature was 60°C. DNA from normal lung tissue was used as the control for unmethylated RASSF1A, and normal lung DNA treated with Sss1 methyltransferases was used as the control for methylated RASSF1A. H₂O was used as negative control. Twenty microliters of each PCR reaction was loaded onto a 6% nondenaturing polyacrylamide gel, stained with ethidium bromide, and pictured under UV light.

Results

Methylation Status of the RASSF1A

A total of 65 primary NSCLCs including 33 adenocarcinomas, 12 large cell carcinomas, and 20 squamous cell carcinomas were analyzed for RASSF1A promoter methylation. As shown in Table 1, 18 of 33 (55%) adenocarcinomas, 3 of 12 (25%) large cell carcinomas, and 5 of 20 (25%) squamous cell carcinomas were detected with RASSF1A promoter methylation. Figure 1A illustrates the representative MSP analysis on pairs of tumors and normal tissues.

Table 1.

Correlation Between the Kras2 Activation and RASSF1A Methylation in NSCLCs.

Tumor Type	Patient Number	Kras2 Mutation		RASSF1A Methylation	Tumors Contain Kras2 Mutations and RASSF1A Methylation	Tumors Contain Neither Kras2 Mutations nor RASSF1A Methylation2
		Codon 12	Codon 13	Total
Adenocarcinoma	33	8/33 (24%)	2/33 (6%)	10/33 (30%)	18/33 (55%)	3/33 (9%)	8/33 (24%)
Large cell carcinoma	12	3/12 (25%)	0/12 (0%)	3/12 (25%)	3/12 (25%)	0/12 (0%)	6/12 (50%)
Squamous cell carcinoma	20	1/20 (5%)	0/20 (0%)	1/20 (5%)	5/20 (25%)	0/5 (0%)	14/20 (70%)

Figure 1.

Kras2 mutation analysis and MSP analysis of RASSF1A in NSCLCs. (A) Representative autoradiograph showing RASSF1A promoter hypermethylation in tumors. Tumor identification numbers were indicated above each autoradiograph; the presence of a visible PCR product in those lanes marked M indicated tumor presence of methylated RASSF1A alleles, whereas the presence of a visible PCR product in those lanes marked U indicated tumor presence of unmethylated RASSF1A alleles. Normal lung tissue (NL) and in vitro methylated DNA (MC) were used as negative and positive controls for RASSF1A promoter methylation. H₂O was also included in each reaction as negative control. (B) Representative example of 12th and 13th codon mutations in Kras2. Lung tumor without Kras2 mutation is shown on the left. Lung tumor with Kras2 12th codon mutation (GGT→GAT transition) is shown in the middle. Lung tumor with Kras2 13th codon mutation (GGC→TGC transition) is shown on the right.

Kras2 Gene Mutation

Using PCR direct sequencing analysis, Kras2 gene 12th and 13th codon mutations were determined in all the NSCLC DNA. In agreement with previous reports, our data showed that 10 of 33 (30%) adenocarcinomas, 3 of 12 (25%) large cell carcinomas, and only 1 of 20 (5%) squamous cell carcinomas contained the Kras2 gene mutation at codon 12 (GGT→TGT, CGT, GAT, GCT, and GTT transition) or codon 13 (GGC→GAC, TGC) (Figure 1B, Table 1).

Correlations Between Kras2 Mutation and RASSF1A Inactivation

By combing RASSF1A promoter methylation analysis with Kras2 mutation data, we found that only 3 of 33 (9%) of the samples contained both a Kras2 mutation and RASSF1A promoter methylation. None of the large cell carcinomas (zero of three) or squamous cell carcinomas (zero of five) with RASSF1A promoter methylation contained a Kras2 mutation (Table 1). Our results indicate that the majority of the adenocarcinomas and large cell carcinomas with Kras2 mutations lack RASSF1A promoter methylation. In addition, 8 of 33 (24%) adenocarcinomas, 6 of 12 (50%) large cell carcinomas, and 14 of 20 (70%) squamous cell carcinomas contain neither Kras2 mutation nor RASSF1A methylation (Table 1). By integrating previous studies [19,22], a comparison of RASSF1A methylation and Kras2 mutation data was complied and summarized in 6Table 2. These data show that only ∼7% adenocarcinomas/large cell carcinomas, 2% stage I lung adenocarcinomas, and 5% colorectal tumors exhibited both RASSF1A promoter methylation and Kras2 mutations.

Table 2.

Kras2 Mutation and RASSF1A Methylation in Human Lung and Colon Cancers.

Genetic Changes	Lung Adenocarcinomas and Large Cell Carcinomas*	Stage I Lung Adenocarcinomas†	Colon Cancer‡
Kras2 mutation and RASSF1A methylation	3 (7%)	2 (2%)	11 (5%)
Kras2 mutation only	10 (22%)	11 (10%)	76 (34%)
RASSF1A methylation only	18 (40%)	33 (30%)	34 (15%)
Neither Kras2 mutation or RASSF1A methylation	14 (31%)	64 (58%)	101 (46%)
Total tumors examined	45	110	222

^*The present study.

^†Tomizawa et al. [19].

^‡Van Engeland et al. [22].

Discussion

In the present study, we systematically evaluated the correlation between activating Kras2 mutations and RASSF1A promoter methylation in 33 human lung adenocarcinomas, 12 large cell carcinomas, and 20 squamous cell carcinomas. We have shown that 30% of the adenocarcinomas, 25% of large cell carcinomas, and only 5% of the squamous cell carcinomas contained activating Kras2 mutation at codon 12 or 13. In addition, we analyzed the methylation status of the promoter region of the RASSF1A gene. Our results showed that RASSF1A inactivation was observed in 55% of adenocarcinomas, 25% large cell carcinomas, and 25% squamous cell carcinomas. When Kras2 mutations and RASSF1A methylation analysis data were combined, we found that only 9% of lung adenocarcinomas and none of large and squamous cell carcinomas contains both a Kras2 mutation and RASSF1A methylation. These data, together with previous study [19], indicated that the majority of lung adenocarcinomas and large cell carcinomas with Kras2 mutations lack RASSF1A promoter methylation. Our results also showed that 24% adenocarcinomas, 50% large cell carcinomas, and 70% squamous cell carcinomas contain neither a Kras2 mutation nor RASSF1A methylation. These data provide evidence that some other alternative pathways might be involved in human lung tumorigenesis other than Kras2 mutation and RASSF1A hypermethylation.

As shown in Table 2, our results are highly consistent with two recently published studies. Van Engeland et al. [22] evaluated Kras2 mutations and RASSF1A promoter methylation in colorectal cancer. They found that the majority of colorectal cancers with Kras2 mutation lack RASSF1A promoter methylation. Another report by Tomizawa et al. [19] found that only 2% of lung adenocarcinomas with RASSF1A hypermethylation showed Kras2-activating mutation [19]. They showed that RASSF1A methylation occurs more frequently in poorly differentiated tumors than in well-differentiated tumors, suggesting that RASSF1A plays an important role in the progression of lung adenocarcinoma and that RASSF1A hypermethylation is a marker for prognosis of patients with stage I lung adenocarcinoma [19].

The Ras effector, RASSF1A, suppressed the tumorigenicity of lung cancer cells in a manner similar to wild-type Ras, suggesting that RASSF1A could be an effector of wildtype Ras genes, which were recently found to be potent tumor suppressors [27,28]. Thus, loss of RASSF1A expression by methylation in human cancer would not require the event of Ras activation. Interestingly, Kras2 mutations are found in less than 1% of SCLCs, whereas inactivation of RASSF1A is close to 80% or 100% by hypermethylation and loss of heterozygosity (LOH) [9]. Therefore, these results raise the possibility that the wild-type Ras-RASSF1 signaling pathway suppresses lung cancer cell proliferation and malignant transformation, and loss of this controlling function by either Kras2 point mutation/LOH or RASSF1A promoter methylation/LOH may result in tumorigenesis. Because RASSF1A has the identical RAS associate domain as RASSF1C, RASSF1C likely binds RAS in a GTP-dependent manner. Therefore, it is possible that RASSF1A shares the same binding manner as RASSF1C [15]. Feig and Buchsbaum [16] and Khokhlatchev et al. [17] suggested that Ras could both promote apoptosis through its association with a Nore1-RASSF1-Mst1 complex and suppress apoptosis through its activation of PI3 kinase.

Inverse correlation between tumor suppressor genes involved in the same signaling pathways has been previously noted (e.g., inactivation of both p16 and Rb is rarely observed in the same tumor). The identification of RASSF1A as a tumor suppressor gene and a downstream effector of Ras in a proapoptotic pathway suggests that Ras can promote both apoptosis and cell survival (by the previously described PI3-K pathway). With this information on the correlation between genetic alterations of RASSF1A and Kras2 during tumorigenesis, future studies should examine the ability of RASSF1A to function as a tumor suppressor gene in experimental lung tumorigenesis and the mechanism through which this novel tumor suppressor gene mediates its effect.

Abbreviations

LOH

loss of heterozygosity

NSCLC

non small cell lung cancer

PCR

polymerase chain reaction

MSP

methylation-specific PCR