Purpose: Lung cancer cells frequently exhibit marked chromosome instability. We postulated that alterations of the double-strand break repair genes (BRCA1, BRCA2, and XRCC5) might be involved in lung cancer.

Patients and Methods: We examined the loss of protein and mRNA expression and the 5′CpG hypermethylation and allelic imbalance of the BRCA1, BRCA2, and XRCC5 genes in 98 non–small cell lung cancer (NSCLC) samples. Anchorage-dependent growth after reexpression of these genes was examined in a lung cancer cell line that originally lacked BRCA1 and BRCA2 expression.

Results: The data indicated that low protein expression of BRCA1 and BRCA2 was frequent in lung adenocarcinomas (42-44%), whereas low XRCC5 protein expression was more prevalent among squamous cell carcinoma (32%). In addition, low BRCA1 expression was significantly associated with low RB expression, especially in lung adenocarcinoma. Concurrent alterations in XRCC5 and p53 were the most frequent profiles in smoking patients. Importantly, low mRNA and protein expressions of BRCA1, BRCA2, and XRCC5 were significantly associated with their promoter hypermethylation. 5-Aza-2′-deoxycytidine treatment of NSCLC cells showed demethylation and reexpression of the BRCA1 and BRCA2 genes and reduced anchorage-independent growth.

Conclusions: Our retrospective study provides compelling evidence that low mRNA and protein expression in the BRCA1/BRCA2 and XRCC5 genes occur in lung adenocarcinoma and squamous cell carcinoma, respectively, and that promoter hypermethylation is the predominant mechanism in deregulation of these genes. Alteration of the double-strand break repair pathway, perhaps by interacting with p53 and RB deregulation, is important in the pathogenesis of a subset of NSCLC.

Lung cancer is one of the most common malignancies, making it very important to identify the associated gene(s) involved. We previously did a genome-wide loss of heterozygosity (LOH) study in 71 primary surgically resected non–small cell lung cancer (NSCLC) tumors in which a high mean LOH frequency was reported, indicating the presence of chromosome instability in NSCLC (1). Chromosome instability may be caused by failure in the repair of DNA double-strand breaks (DSB; ref. 2). Interestingly, the chromosome regions at 2q33-35 and 13q12.3, which included gene loci encoding the XRCC5 (Ku80) and BRCA2 (breast cancer 2) DSB repair proteins, respectively, showed a high frequency of LOH in NSCLC tumors (1). We therefore postulated that alterations of genes in chromosomal stability control pathways, such as DSB repair genes (BRCA1, BRCA2, and XRCC5), may be involved in NSCLC tumorigenesis.

Two distinct and complementary mechanisms, homologous recombination and nonhomologous end joining, are involved in the repair of detrimental DSBs in mammalian cells (2, 3). The breast tumor suppressors BRCA1 and BRCA2 have been shown to interact with DNA damage–induced Rad51 protein to form foci on DSBs (4, 5). Mouse Brca1- and Brca2-deficient embryonic stem cells show a reduced efficiency of homologous recombination (6, 7) and develop spontaneous chromosomal aberrations (8, 9). It has been shown that XRCC5 and XRCC6 encode the 80- and 70-kDa subunits of the Ku70/Ku80 heterodimer, which may be involved in the binding of DSB ends during nonhomologous end joining (10). Primary fibroblasts from Xrcc5−/− mice were found to frequently acquire chromosomal aberrations (11). Growing evidence suggests the potential involvement of defects in the DSB repair system in the chromosome instability of human cancers.

Germ line mutations in the BRCA1 and BRCA2 genes are strongly associated with an elevated risk of breast and ovarian cancers (12). Mutation carriers have a statistically increased risk of several other cancers [i.e., pancreatic and cervical cancers with BRCA1 (13) and prostate, pancreas, stomach, melanoma, lung, and bladder cancers with BRCA2 (14)]. Decreased protein expression of XRCC5 is also reported for melanoma, cervical, and colon cancers (15). Frequent LOH at the BRCA1, BRCA2, and XRCC5 has been observed in several sporadic cancers other than breast and ovary (i.e., tumors of the colon, pancreas, soft tissue, and lung; refs. 1, 16). Recently, promoter hypermethylation of the BRCA1 and BRCA2 has been shown in sporadic breast and ovarian tumors (17, 18). However, no direct evidence regarding the potential involvement of defects in the DSB repair system has yet been obtained in lung cancer. In the data to date, only two studies have reported the mRNA expression (19) and 5′CpG methylation (20) analyses on the BRCA1 gene in lung cancer. No XRCC5 studies have been reported previously in lung cancer.

To investigate the clinical association of DSB repair alteration in NSCLC tumorigenesis and to elucidate the possible mechanisms involved, we did gene/protein alteration studies, including LOH and promoter hypermethylation, and loss of RNA and protein expression on the DSB repair genes (BRCA1, BRCA2, and XRCC5) in tumors from 98 NSCLC patients. The clinicopathologic data of patients and anchorage-independent growth in model cells were also examined to determine the associations of DSB repair alteration with the clinical and biological behavior of NSCLC tumorigenesis.

Sample preparation and clinical characterization of patients. Tissues were collected after obtaining appropriate institutional review board permission and informed consent from the recruited patients. Surgically resected tumor samples from 98 patients with NSCLC were collected between 1999 and 2004. Of these patients, 52 had adenocarcinomas, 37 had squamous cell carcinomas (SCC), 6 had large cell carcinomas, and 3 had nondifferential NSCLC. The histology of the tumor types and their stages were determined according to the WHO classification and the tumor-node-metastasis system, respectively. Information on the age, sex, and smoking history of the patients was obtained from hospital records.

Surgically resected tumor samples were immediately snap-frozen and subsequently stored in liquid nitrogen. For the LOH and methylation assay, genomic DNA from matched pairs of primary tumors and nearby normal lung tissues was prepared using proteinase K digestion and phenol-chloroform extraction and then ethanol precipitation. For the RNA expression assay, total RNA was prepared from tumors and normal lung tissues, using Trizol reagent (Invitrogen, Carlsbad, CA). cDNA was synthesized using SuperScript reverse transcriptase (Invitrogen).

Analysis of protein expression: immunohistochemistry assay. Paraffin blocks of tumors were cut into 5-μm slices and then processed using standard deparaffinization and rehydration techniques. For the BRCA2 analysis, proteinase K buffer was used for the antigen retrieval step. Monoclonal antibodies for BRCA1-8F7 (1:500; GeneTex, San Antonio, TX), anti-BRCA2 (1:100; Santa Cruz Biotechnology, Santa Cruz, CA), XRCC5-Ab7 (1:200; Neomarker, Fremont, CA), p53 (1:200; DAKO AB, Glostrup, Denmark), and RB (1:200; Neomarker) antibodies were used as the primary antibodies to detect the protein expression. The valuation of the immunohistochemistry was conducted blindly without knowledge of the clinical and pathologic characteristics of the cases. The stains were graded low or negative expression when <25% tumor cells were positive using adequate staining in surrounding normal stromal and epithelial cells (2123).

Analysis of mRNA expression: semiquantitative multiplex reverse transcription-PCR assay.BRCA1, BRCA2, and XRCC5 mRNA expression were assayed in a multiplex reverse transcription-PCR (RT-PCR) analysis using the β-actin gene as an internal control. The primer sequences for BRCA1, BRCA2, and β-actin genes were as described in refs. 19, 24, 25. Special sets of primers were used to detect the alternative splicing for the BRCA1 and BRCA2 genes (24, 25). Primers for the XRCC5 reaction were 5′-GCTTTTCCTCATATCAAGCATAACT-3′ (forward) and 5′-AATCTCTGAAATCGAGGATTTGG-3′ (reverse). The number of cycles and primers and the amount of cDNA used were determined to provide quantitative amplification during multiplex RT-PCR. To quantify the relative levels of gene expression in the multiplex RT-PCR assay, the value for the internal standard (β-actin) in each test tube was used as the baseline value for gene expression in that sample, and a relative value was calculated for each target gene transcript amplified from each tumor and matched normal sample. Tumor cells that exhibited mRNA expression below 50% of that of normal cells were deemed to have an abnormal pattern (26).

Methylation-specific PCR assay and bisulfite genomic sequencing. The methylation status in the promoter region of the BRCA1, BRCA2, and XRCC5 genes was determined by chemical treatment with sodium bisulfite and subsequent methylation-specific PCR (MSP) analysis. MSP analyses were conducted using the following primers for the −1429 to −1314 region in BRCA1 promoter: unmethylated BRCA1, 5′-TGAGAGATGTTTGGTTTTTTT-3′ (forward) and 5′-CACACTTTTCCTTTTCCACT-3′ (reverse); methylated BRCA1, 5′-CGAGAGACGTTTGGTTTTTTT-3′ (forward) and 5′-CGCGCTTTTCCGTTACCACG-3′ (reverse). A nested-MSP was used for the −293 to −145 region in the XRCC5 promoter: unmethylated XRCC5-outer, 5′-TGATTATGGTGGAATGGAG-3′ (forward) and 5′-ACCACTTCATTTCCTACTAAACC-3′ (reverse); unmethylated XRCC5-inner, 5′-TGTATGTTTGGTGGGAATTTGTG-3′ (forward) and 5′-CATACACAAATTCCCACCAA-3′ (reverse); methylated XRCC5-outer, 5′-TCGATTACGGCGGAAT-3′ (forward) and 5′-CCGTACCCCGAAACTC-3′ (reverse); methylated XRCC5-inner, 5′-CGTATGTTCGGCGGGAATTTGCG-3′ (forward) and 5′-CATACGCAAATTCCCGCCGA-3′ (reverse). The MSP for the BRCA2 gene were described by Hilton et al. (27). Positive control samples with unmethylated lymphocyte DNA and SssI methyltransferase–treated methylated DNA were also included for each set of PCR. Because clinical tissue samples should be considered heterogeneous, both U and M amplicons can be expected. To increase the specificity of the assay, the hypermethylation genes were defined as the amplification of more M products compared with U products from the tumor sample (28). In addition, the PCR products of 11 tumor available samples were sequenced by forward and reverse primers for BRCA1 (−1429 to −1314, a total of eight CpGs detected), BRCA2 (−133 to +98, a total of 26 CpGs detected), and XRCC5 (−293 to −145, a total of 23 CpGs detected) in an ABI Prism 377 DNA Sequencer (PE Applied Biosystems, Foster City, CA).

LOH analysis. PCR reactions were conducted using fluorescently labeled and unlabeled primers for the microsatellite marker(s) located in the gene loci. They were D17S1322 and D17S1323 (5′ and 3′ to the BRCA1 locus, respectively), D13S1695 and D13S1699 (5′ and 3′ to the BRCA2 locus, respectively), and D2S2382 and D2S164 (intragenic in and 3′ to the XRCC5 locus, respectively). PCR products were mixed with fluorescent molecular weight markers for subsequent electrophoresis in a MegaBACE 1000 automatic sequencer (Amersham Pharmacia, Piscataway, NJ). Allele sizes were determined using Genetic Profiler Analysis version 2.0 software. The allelic ratio was calculated as (T1/T2) / (N1/N2), the ratio of area values for tumor (T) versus normal (N) alleles. LOH was defined as an allelic ratio >2.0 or <0.5 (1).

5-Aza-2′-deoxycytidine treatment of NSCLC cells. The human NSCLC cell line CL1-5-F4 was derived from Taiwanese patients with lung adenocarcinoma (kindly provided by Dr. P-C. Yang, Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan). Cells were plated at 105 per 100-mm culture dish on the day before treatment. The cultures were treated for two to three doubling times with 1 μmol/L of 5-aza-2′-deoxycytidine (5-Aza-dC). The 5-Aza-dC–containing medium was changed for each cell doubling during the treatment. On the day after the third doubling, the cells were harvested for an analysis of their methylation status using a MSP assay and of their production of mRNA using a RT-PCR assay.

Anchorage-independent growth assay. Cells (1 × 104) were seeded in the upper layer with 0.4% agarose gel, 1 μmol/L of 5-Aza-dC, and 10% serum containing medium. The lower layer contained 0.6% agarose gel and 1 μmol/L of 5-Aza-dC. Cells were fed with medium containing 1 μmol/L of 5-Aza-dC and 10% serum for 10 days. After 10 days of treatment, the cell mass growth over 50 μm was counted.

Statistical analysis. Pearson's χ2 test was used to compare the frequency of BRCA1, BRCA2, and XRCC5 alterations in NSCLC patients with different characteristics, including age, sex, and smoking status, and various clinicopathologic variables, such as tumor type and tumor stage. The smokers included both current smokers and ex smokers. The comparison of tumor type focused mainly on adenocarcinoma versus SCC. For the staging of the primary tumor, we pooled stages I and II for the early-stage category and stages III and IV for the late-stage category.

Gene and protein alterations of BRCA1, BRCA2, and XRCC5 in primary NSCLC tumors

Loss of protein expression and its correlation with clinicopathologic variables and p53/RB alterations. Immunohistochemical staining of BRCA1, BRCA2, and XRCC5 proteins was done on 98 tumor samples. The data indicated that 37%, 34%, and 21% of tumors showed an absence or low expression of BRCA1, BRCA2, and XRCC5 proteins, respectively (Fig. 1B, D, and H; Table 1). In addition, we found that low expression of either BRCA1 or BRCA2 proteins was in 57% of tumors analyzed (Table 1).

Fig. 1.

Immunohistochemical analysis of BRCA1 (A and B), BRCA2 (C and D), RB (E and F), XRCC5 (G and H), and p53 (I and J) proteins in formalin-fixed, paraffin-embedded sections of resected tumor specimens of NSCLC. Positive immunoreactivity is visible as a red precipitate. A concordantly negative immunoreactivity of BRCA1 and BRCA2 was found in an adenocarcinoma patient (B and D). Accordingly, low RB expression (F) was also found in this patient. Loss of both XRCC5 and overexpression of p53 (H and J) in an SCC patient. Original magnification, ×200.

Fig. 1.

Immunohistochemical analysis of BRCA1 (A and B), BRCA2 (C and D), RB (E and F), XRCC5 (G and H), and p53 (I and J) proteins in formalin-fixed, paraffin-embedded sections of resected tumor specimens of NSCLC. Positive immunoreactivity is visible as a red precipitate. A concordantly negative immunoreactivity of BRCA1 and BRCA2 was found in an adenocarcinoma patient (B and D). Accordingly, low RB expression (F) was also found in this patient. Loss of both XRCC5 and overexpression of p53 (H and J) in an SCC patient. Original magnification, ×200.

Close modal
Table 1.

Correlation analyses of BRCA1/BRCA2/XRCC5 protein alteration in relation to clinicopathologic variables of 98 NSCLC patients

CharacteristicsBRCA1
BRCA2
BRCA1 or BRCA2
XRCC5
Total, n*Low protein, n (%)Total, n*Low protein, n (%)Total, n*Low protein, n (%)Total, n*Low protein, n (%)
Overall 98 36 (37) 98 33 (34) 98 56 (57) 98 21 (21) 
Clinicopathologic variables         
Smoking         
    Yes 56 21 (38) 56 19 (34) 56 32 (57) 56 13 (23) 
    No 30 10 (33) 30 13 (43) 30 18 (60) 30 3 (10) 
Tumor type         
    Adenocarcinoma 52 22 (42) 52 23 (44) 0.054 52 36 (69) 0.014 52 8 (15) 
    SCC 37 11 (30) 37 9 (24) 37 16 (43) 37 12 (32) 0.058 
Tumor stage         
    I + II 50 16 (32) 50 16 (32) 50 27 (54) 50 13 (26) 
    III + IV 44 19 (43) 44 17 (39) 44 28 (64) 44 7 (16) 
CharacteristicsBRCA1
BRCA2
BRCA1 or BRCA2
XRCC5
Total, n*Low protein, n (%)Total, n*Low protein, n (%)Total, n*Low protein, n (%)Total, n*Low protein, n (%)
Overall 98 36 (37) 98 33 (34) 98 56 (57) 98 21 (21) 
Clinicopathologic variables         
Smoking         
    Yes 56 21 (38) 56 19 (34) 56 32 (57) 56 13 (23) 
    No 30 10 (33) 30 13 (43) 30 18 (60) 30 3 (10) 
Tumor type         
    Adenocarcinoma 52 22 (42) 52 23 (44) 0.054 52 36 (69) 0.014 52 8 (15) 
    SCC 37 11 (30) 37 9 (24) 37 16 (43) 37 12 (32) 0.058 
Tumor stage         
    I + II 50 16 (32) 50 16 (32) 50 27 (54) 50 13 (26) 
    III + IV 44 19 (43) 44 17 (39) 44 28 (64) 44 7 (16) 

NOTE: Ps with significance are shown as superscripts.

*

Total number of samples in some categories is less than the overall number analyzed because clinical data was not available for these samples.

To examine whether there were associations of alterations at the BRCA1, BRCA2, and XRCC5 proteins with the clinical characteristics of patients, the occurrence of alteration at each protein was compared with the patients' clinicopathologic variables, including sex, age, smoking habit, tumor type, and tumor stage (Table 1). χ2 analysis showed that no significant correlation of BRCA1, BRCA2, and XRCC5 protein expressions was detected in these clinicopathologic variables. However, low BRCA1 or BRCA2 protein expression occurred primarily in patients suffering from adenocarcinoma types of lung cancer (P = 0.014). By contrast, low protein expression of the XRCC5 was nearly significantly restricted to patients suffering from SCC of lung cancer (P = 0.058).

In this cohort of patients, 81 patients had been analyzed for the protein expression of p53 and RB in our lab previously (29, 30). Therefore, the alterations of BRCA1, BRCA2, and XRCC5 proteins were tested for their association with the alterations in 53 and RB to reveal the existence of a correlation of DSB repair with p53/RB pathway (Table 2). The data indicated that a loss of both BRCA1 and RB protein expression was more in adenocarcinoma patients than that in SCC patients (P = 0.020). In addition, there was a borderline significant trend for low XRCC5 protein and p53 overexpression in smoking patients compared with that in nonsmoking patients (P = 0.084).

Table 2.

Correlation analyses of low protein expression of BRCA1/BRCA2/XRCC5 and RB/p53 protein alteration in relation to tumor type and smoking habit

CharacteristicsTotal, n*BRCA1, low protein, n (%)BRCA2, low protein, n (%)XRCC5, low protein, n (%)
Protein expression of RB in AD patients     
    Low 22 14 (64) 0.020 9 (41) 5 (23) 
    Normal 13 3 (23) 6 (46) 1 (8) 
Protein expression of p53 in smoker patients     
    Overexpression 27 9 (33) 9 (33) 9 (33) 0.084 
    Normal 29 12 (41) 10 (34) 4 (14) 
CharacteristicsTotal, n*BRCA1, low protein, n (%)BRCA2, low protein, n (%)XRCC5, low protein, n (%)
Protein expression of RB in AD patients     
    Low 22 14 (64) 0.020 9 (41) 5 (23) 
    Normal 13 3 (23) 6 (46) 1 (8) 
Protein expression of p53 in smoker patients     
    Overexpression 27 9 (33) 9 (33) 9 (33) 0.084 
    Normal 29 12 (41) 10 (34) 4 (14) 

NOTE: Ps with significance are shown as superscripts.

*

Although a total of 81 patients were analyzed, only the data for 35 adenocarcinoma patients and 56 smoker patients were shown.

Low mRNA expression and alternative splicing. RNA with high quality for multiplex RT-PCR analysis was extracted from frozen tissues of 87 NSCLC patients (Fig. 2A). Decreased or absence of BRCA1, BRCA2, and XRCC5 transcripts were shown by multiplex semiquantitative RT-PCR to occur in 30%, 37%, and 26% tumors, respectively.

Fig. 2.

Representative figures for BRCA1, BRCA2, and XRCC5 mRNA expression analysis by the semiquantitative multiplex RT-PCR assay. The results of their mRNA expression status are indicated above. + (positive) and − (negative) in the rows labeled mRNA indicate their expression status. N, normal lung tissue; T, tumor tissue of the lung. β-actin was used as the internal control for the analysis (an amplification fragment of 201 bp was used for BRCA1 and BRCA2 analyses; an amplification of 618 bp was used for XRCC5 analysis). The tumor tissues expressing low amounts of target transcript are indicated with arrows.

Fig. 2.

Representative figures for BRCA1, BRCA2, and XRCC5 mRNA expression analysis by the semiquantitative multiplex RT-PCR assay. The results of their mRNA expression status are indicated above. + (positive) and − (negative) in the rows labeled mRNA indicate their expression status. N, normal lung tissue; T, tumor tissue of the lung. β-actin was used as the internal control for the analysis (an amplification fragment of 201 bp was used for BRCA1 and BRCA2 analyses; an amplification of 618 bp was used for XRCC5 analysis). The tumor tissues expressing low amounts of target transcript are indicated with arrows.

Close modal

The presence of exon 5 or exon 12 alternative splicing of the BRCA1 and BRCA2 genes had been reported in brain tissue (24, 25). We therefore analyzed cDNA samples using primer sets specifically designed to detect the exon 5 for BRCA1 and the exon 12 for BRCA2 in these NSCLC samples. The data indicated that none of the cDNA samples showed a splicing variant of exon 5 in the BRCA1, whereas all samples exhibited an alternative splicing of exon 12 in the BRCA2. However, the exon 12 splicing variant of the BRCA2 gene was equally expressed in the normal lung tissue.

Promoter hypermethylation. 5′CpG hypermethylation has been shown to be the mechanism underlying the frequent aberrant expression of many tumor suppressor genes. Therefore, we studied promoter hypermethylation of the BRCA1, BRCA2, and XRCC5 genes using an MSP assay in this cohort of NSCLC (Fig. 3A). There were 30%, 42%, and 20% of NSCLC tumors showing promoter hypermethylation in the BRCA1, BRCA2, and XRCC5 genes, respectively, whereas no or low methylation was found in their matched normal lung tissue. Moderate-to-strong methylation of the regions with reported promoter activities from −1429 to −1314 of BRCA1, −133 to +98 of BRCA2, and −293 to −145 of XRCC5 was found by direct bisulfite sequencing in 11 tumor samples that were available for repeating experiments (Fig. 3B).

Fig. 3.

A, representative figure of 5′CpG hypermethylation analysis of BRCA1, BRCA2, and XRCC5 using the MSP assay. + (positive) and − (negative) in the rows labeled methyl indicate their promoter methylation status. Primer sets used for amplification are designated as U for unmethylated or M for methylated genes. B, bisulfite sequencing analysis of BRCA1, BRCA2, and XRCC5 genes in 10 NSCLC samples. Methylated CpG sites at the level of 100% to 76% (•), 75% to 51% (), 50% to 26% (), and 25% to 0% (○).

Fig. 3.

A, representative figure of 5′CpG hypermethylation analysis of BRCA1, BRCA2, and XRCC5 using the MSP assay. + (positive) and − (negative) in the rows labeled methyl indicate their promoter methylation status. Primer sets used for amplification are designated as U for unmethylated or M for methylated genes. B, bisulfite sequencing analysis of BRCA1, BRCA2, and XRCC5 genes in 10 NSCLC samples. Methylated CpG sites at the level of 100% to 76% (•), 75% to 51% (), 50% to 26% (), and 25% to 0% (○).

Close modal

LOH at microsatellites in the three gene loci. We further investigated whether LOH of the BRCA1, BRCA2, and XRCC5 genes occurred in this series of NSCLC tumors (Fig. 4). Microdissected genomic DNA obtained from 87 matched pairs of primary tumors and nearby normal lung tissue was examined for the incidence of LOH at the intragenic and/or nearby microsatellite markers. The data indicated that 25%, 44%, and 48% of tumors were found to harbor LOH at the BRCA1, BRCA2, and XRCC5 loci, respectively.

Fig. 4.

Effects of 5′-Aza-dC treatment on mRNA expression (A) and demethylation (B) of BRCA1 and BRCA2 in CL1-5-F4 NSCLC cells. The absence of the M product after 5′-Aza-dC indicates the demethylation of the promoter in these cell lines. C, anchorage-independent tumorigenic assay of CL1-5F4 cells treated with 5′-Aza-dC. Left, photographic views for colony formation; right, counted ratio of colony number between DMSO control– and 5′-Aza-dC–treated cells.

Fig. 4.

Effects of 5′-Aza-dC treatment on mRNA expression (A) and demethylation (B) of BRCA1 and BRCA2 in CL1-5-F4 NSCLC cells. The absence of the M product after 5′-Aza-dC indicates the demethylation of the promoter in these cell lines. C, anchorage-independent tumorigenic assay of CL1-5F4 cells treated with 5′-Aza-dC. Left, photographic views for colony formation; right, counted ratio of colony number between DMSO control– and 5′-Aza-dC–treated cells.

Close modal

Concordance analysis of low mRNA/protein expression of BRCA1, BRCA2, and XRCC5 with their promoter hypermethylation and allelic imbalance

To investigate whether the low BRCA1, BRCA2, and XRCC5 mRNA and proteins expressions were coordinated with promoter hypermethylation and allelic imbalance at the gene locus, cross-tabulation analysis was done on the protein data to examine the correlations among mRNA expression, DNA methylation, and LOH status, using the Pearson's χ2 test (Fig. 5). Aberrant protein expression was significantly associated with low mRNA expression (BRCA1, P < 0.001; BRCA2, P < 0.001; XRCC5, P = 0.032) and with promoter hypermethylation (BRCA1, P = 0.008; BRCA2, P < 0.001; XRCC5, P = 0.047) but not with LOH of the gene. Furthermore, low mRNA expression was significantly associated with hypermethylation of the gene (BRCA1, P < 0.001; BRCA2, P < 0.001; XRCC5, P < 0.001).

Fig. 5.

Concordance analysis among protein expression, mRNA expression, and promoter methylation in (A) BRCA1, (B) BRCA2, and (C) XRCC5. Y-axis, percentage of cases; X-axis, type of comparison. +, positive protein expression, positive mRNA expression, and hypermethylation of promoter. Gray columns, the concordant group, such as loss of expression of both protein and mRNA. White columns, the discordant group, such as expression of protein without loss of mRNA expression.

Fig. 5.

Concordance analysis among protein expression, mRNA expression, and promoter methylation in (A) BRCA1, (B) BRCA2, and (C) XRCC5. Y-axis, percentage of cases; X-axis, type of comparison. +, positive protein expression, positive mRNA expression, and hypermethylation of promoter. Gray columns, the concordant group, such as loss of expression of both protein and mRNA. White columns, the discordant group, such as expression of protein without loss of mRNA expression.

Close modal

Reactivation of BRCA1 and BRCA2 by 5-Aza-dC treatment

5-Aza-dC increased mRNA expression. To determine whether BRCA1 and BRCA2 promoter methylation could be further linked to the loss of gene expression, CL1-5F4 NSCLC cell, which showed negative expression and promoter hypermethylation of the BRCA1 and BRCA2 genes, were treated with the demethylating agent 5-Aza-dC. As shown in the Fig. 3C, 5-Aza-dC successfully restored mRNA and protein expression and demethylated the promoter region in the cells that lacked BRCA1 and BRCA2 expression and harbored a methylated respective promoter.

5-Aza-dC decreased anchorage independence. To evaluate whether the reexpression of BRCA1 and BRCA2 resulted in low tumorigenicity, the effect of 5-Aza-dC on CL1-5F4 cancer cell growth was tested by the anchorage-independent assay. A significant decrease of anchorage-independent growth was observed in CL1-5F4 cancer cells reexpressing the BRCA1 and BRCA2 (Fig. 3D).

Previous studies have shown that lung cancer cells frequently exhibit marked chromosome instability (1, 31). It is possible that defects in the DSB repair system are involved in the genesis of chromosome instability in lung cancer. In an effort to better understand the molecular pathogenesis of the DSB repair pathway in NSCLC, we collected 98 primary NSCLC tumors and carried out a comprehensive molecular analysis, including mRNA/protein expression and DNA methylation/allelic imbalance of three genes (BRCA1, BRCA2, and XRCC5) encoding proteins known to function in the DSB repair pathway. We found that 64% (63 of 98) of NSCLC tumors showed alteration in one of the abovementioned genes, suggesting that DSB repair is potentially involved in the NSCLC tumorigenesis. Low expression of BRCA1 and XRCC5 was frequent in adenocarcinoma/RB–deficient and SCC/p53–deficient lung cancer, respectively. Note that low mRNA and protein expression of the BRCA1, BRCA2, and XRCC5 genes was most often attributable to their promoter hypermethylation. To our knowledge, our study is the first to identify deregulation of BRCA1, BRCA2, and XRCC5 genes in the same series of NSCLC samples. Our data provide new evidence for an alternative mechanism of NSCLC tumorigenesis by DSB repair deregulation.

Low levels of BRCA1/BRCA2 and XRCC5 protein expression occurred specifically in primary adenocarcinoma and SCC lung tumors, respectively. It is possible that the homologous recombination and nonhomologous end joining repair pathways differ in their requirements for DSB repair in different cell types (32). Homologous recombination may be responsible for the homologous template in non–smoking-related adenocarcinoma lung cancer development, whereas nonhomologous end joining may be required in the double-stranded end joining for smoking-related SCC lung cancer. In view of the apparent link between RB and p53 tumor suppressor genes in adenocarcinoma and smoker NSCLC tumorigenesis, respectively (29, 33), in this series of NSCLC patients with previously examined protein expression data of the RB and p53, we investigated the possible relationship of either BRCA1/BRCA2 or XRCC5 alterations. The data indicated that low BRCA1 expression was significantly associated with low RB expression, especially in adenocarcinoma-type lung cancer (Table 2). In addition, concurrent alterations in XRCC5 and p53 were the most frequent profile in smoking NSCLC patients (Table 2). These data were in agreement with the role of p53 and RB in the cell cycle checkpoint controls (34, 35). It has been shown that p53 deficiency can rescue the embryonic lethality and developmental defects of XRCC5−/− mice (11). RB binds to the BRCT domain of BRCA1 to mediate the ability of BRCA1 to suppress growth (36, 37). We therefore hypothesized a model in which alteration of the DSB repair pathway, perhaps by interacting with p53 and RB deregulation, is important in the pathogenesis of a subset of NSCLC. The function of these “gatekeepers” (p53 and Rb) and “caretakers” (XRCC5 and BRCA1/BRCA2) in regulating the cellular response to DSB in NSCLC warrants further investigation.

A high concordance was observed between alterations in protein and mRNA expression and promoter hypermethylation of the BRCA1, BRCA2, and XRCC5 genes. The data suggest that promoter hypermethylation is the predominant mechanism to inactivate the BRCA1, BRCA2, and XRCC5 genes. We deliberately designed the primers for MSP at regions with reported promoter activities (3841). Methylation status of the promoter region examined was correlated to the transcriptional activity of the genes and perhaps to tumorigenesis. This conclusion is further strengthened by the reexpression of mRNA expression together with demethylation at the promoter region in both the BRCA1 and BRCA2 genes and inhibition of anchorage-independent growth, using 5-Aza-dC treatment for NSCLC cell lines (Fig. 3C and D). In the future, it will be important to examine the effects of 5-Aza-dC on neoplastic phenotypes within a much larger cohort of samples so as to define more precisely the critical site in BRCA1, BRCA2, and XRCC5 for gene silencing.

Note that some patients with positive protein expression showed promoter methylation. This could be explained by the presence of several distinct tumor subpopulations, one of which has methylation and does not express the protein, whereas the others have no methylation and no protein expression. In addition, methylation occurred at only one allele so that the remaining allele still expresses the protein that we observed. We found that the allelic imbalance in chromosome regions that harbor the BRCA1, BRCA2, and XRCC5 genes was not associated with the altered expression of the respective proteins. This may be because inactivation of tumor suppressor genes occurs via a two-hit mechanism. Loss of only one allele may not result in loss of tumor suppressor protein expression (42). There were several patients with no demonstrable protein staining and no evidence of promoter methylation and LOH. We postulated that histone deacetylation further decreasing the access of transcription factors to their binding sites on gene promoters and resulting in gene silencing may be involved in these patients. Alternatively, unidentified genetic variants may be occurred in these samples.

In conclusion, our findings provided compelling evidence of low mRNA and protein expression in the BRCA1/BRCA2 and XRCC5 genes in lung adenocarcinoma and SCC, respectively, and suggested that promoter hypermethylation is the predominant mechanism in the deregulation of these genes. Our data support the premise that BRCA1, BRCA2, and XRCC5 are tumor suppressors in NSCLC, and that alteration in DSB repair pathway is involved in NSCLC tumorigenesis. Our clinical correlation data corroborated the cell model data by Sun et al. showing the deficiency of DSB repair in tumorigenic human bronchial epithelial cells treated with α-particle exposure (43). Whether the deficiency of DSB repair could promote tumorigenesis in NSCLC cells treated with other DNA damage reagents, perhaps by interacting with p53 and RB deregulation, needs further examination. In addition, BRCA1 has been shown to associate with the RAD50/MRE11/NBS1 complex that functions in both nonhomologous end joining and homologous recombination (44). Therefore, we tested the mRNA alteration of NBS1 in 75 NSCLC patients. The data indicated that 17% of patients had low mRNA expression of NBS1. However, low expression of NBS1 did not show any statistically significant association with clinicopathologic variables. Alterations in other components of DSB signaling, including RAD50, MRE11, XRCC4, and XRCC6, in NSCLC samples are under investigation to more clearly determine the etiology of this disease.

Grant support: National Science Council (The Executive Yuan, Republic of China) grants NSC94-2320-B-003-002 and DOH95-TD-G-111-004.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Originality: This is the first study to examine the alteration of DNA, RNA, and protein in double-strand break repair genes BRCA1, BRCA2, and XRCC5 in the same series of human cancer and to examine their relationship with the clinical significance in patients and growth variables in cell models. The study identifies that alteration of BRCA1 and BRCA2 was frequent in lung adenocarcinomas, whereas low XRCC5 protein was specific to squamous cell carcinoma. Promoter hypermethylation is the predominant mechanism in the deregulation of these genes.

1
Tseng RC, Chang JW, Hsien FJ, et al. Genomewide loss of heterozygosity and its clinical associations in non small cell lung cancer.
Int J Cancer
2005
;
117
:
241
–7.
2
van Gent DC, Hoeijmakers JH, Kanaar R. Chromosomal stability and the DNA double-stranded break connection.
Nat Rev Genet
2001
;
2
:
196
–206.
3
Khanna KK, Jackson SP. DNA double-strand breaks: signaling, repair and the cancer connection.
Nat Genet
2001
;
27
:
247
–54.
4
Chen J, Silver DP, Walpita D, et al. Stable interaction between the products of the BRCA1 and BRCA2 tumor suppressor genes in mitotic and meiotic cells.
Mol Cell
1998
;
2
:
317
–28.
5
Davies AA, Masson JY, McIlwraith MJ, et al. Role of BRCA2 in control of the RAD51 recombination and DNA repair protein.
Mol Cell
2001
;
7
:
273
–82.
6
Moynahan ME, Chiu JW, Koller BH, Jasin M. Brca1 controls homology-directed DNA repair.
Mol Cell
1999
;
4
:
511
–8.
7
Patel KJ, Yu VP, Lee H, et al. Involvement of Brca2 in DNA repair.
Mol Cell
1998
;
1
:
347
–57.
8
Xu X, Weaver Z, Linke SP, et al. Centrosome amplification and a defective G2-M cell cycle checkpoint induce genetic instability in BRCA1 exon 11 isoform-deficient cells.
Mol Cell
1999
;
3
:
389
–95.
9
Lee H, Trainer AH, Friedman LS, et al. Mitotic checkpoint inactivation fosters transformation in cells lacking the breast cancer susceptibility gene, Brca2.
Mol Cell
1999
;
4
:
1
–10.
10
Koike M. Dimerization, translocation and localization of Ku70 and Ku80 proteins.
J Radiat Res (Tokyo)
2002
;
43
:
223
–36.
11
Difilippantonio MJ, Zhu J, Chen HT, et al. DNA repair protein Ku80 suppresses chromosomal aberrations and malignant transformation.
Nature
2000
;
404
:
510
–4.
12
Ford D, Easton DF, Stratton M, et al. The Breast Cancer Linkage Consortium. Genetic heterogeneity and penetrance analysis of the BRCA1 and BRCA2 genes in breast cancer families.
Am J Hum Genet
1998
;
62
:
676
–89.
13
Thompson D, Easton DF. Cancer Incidence in BRCA1 mutation carriers.
J Natl Cancer Inst
2002
;
94
:
1358
–65.
14
The Breast Cancer Linkage Consortium. Cancer risks in BRCA2 mutation carriers.
J Natl Cancer Inst
1999
;
91
:
1310
–6.
15
Korabiowska M, Tscherny M, Stachura J, Berger H, Cordon-Cardo C, Brinck U. Differential expression of DNA nonhomologous end-joining proteins Ku70 and Ku80 in melanoma progression.
Mod Pathol
2002
;
15
:
426
–33.
16
Boettger MB, Sergi C, Meyer P. BRCA1/2 mutation screening and LOH analysis of lung adenocarcinoma tissue in a multiple-cancer patient with a strong family history of breast cancer.
J Carcinog
2003
;
2
:
5
.
17
Esteller M, Silva JM, Dominguez G, et al. Promoter hypermethylation and BRCA1 inactivation in sporadic breast and ovarian tumors.
J Natl Cancer Inst
2000
;
92
:
564
–9.
18
Gras E, Cortes J, Diez O, et al. Loss of heterozygosity on chromosome 13q12–14, BRCA-2 mutations and lack of BRCA-2 promoter hypermethylation in sporadic epithelial ovarian tumors.
Cancer
2001
;
92
:
787
–95.
19
Taron M, Rosell R, Felip E, et al. BRCA1 mRNA expression levels as an indicator of chemoresistance in lung cancer.
Hum Mol Genet
2004
;
13
:
2443
–9.
20
Marsit CJ, Liu M, Nelson HH, Posner M, Suzuki M, Kelsey KT. Inactivation of the Fanconi anemia/BRCA pathway in lung and oral cancers: implications for treatment and survival.
Oncogene
2004
;
23
:
1000
–4.
21
Niwa Y, Oyama T, Nakajima T. BRCA1 expression status in relation to DNA methylation of the BRCA1 promoter region in sporadic breast cancers.
Jpn J Cancer Res
2000
;
91
:
519
–26.
22
Miyamoto K, Fukutomi T, Asada K, et al. Promoter hypermethylation and post-transcriptional mechanisms for reduced BRCA1 immunoreactivity in sporadic human breast cancers.
Jpn J Clin Oncol
2002
;
32
:
79
–84.
23
Moll U, Lau R, Sypes MA, Gupta MM, Anderson CW. DNA-PK, the DNA-activated protein kinase, is differentially expressed in normal and malignant human tissues.
Oncogene
1999
;
18
:
3114
–26.
24
Claes K, Vandesompele J, Poppe B, et al. Pathological splice mutations outside the invariant AG/GT splice sites of BRCA1 exon 5 increase alternative transcript levels in the 5′ end of the BRCA1 gene.
Oncogene
2002
;
21
:
4171
–5.
25
Bieche I, Lidereau R. Increased level of exon 12 alternatively spliced BRCA2 transcripts in tumor breast tissue compared with normal tissue.
Cancer Res
1999
;
59
:
2546
–50.
26
Chang KW, Kao SY, Tzeng RJ, et al. Multiple molecular alterations of FHIT in betel-associated oral carcinoma.
J Pathol
2002
;
196
:
300
–6.
27
Hilton JL, Geisler JP, Rathe JA, Hattermann-Zogg MA, DeYoung B, Buller RE. Inactivation of BRCA1 and BRCA2 in ovarian cancer.
J Natl Cancer Inst
2002
;
94
:
1396
–406.
28
Derks S, Lentjes MH, Hellebrekers DM, de Bruine AP, Herman JG, van Engeland M. Methylation-specific PCR unraveled.
Cell Oncol
2004
;
26
:
291
–9.
29
Chen JT, Chen YC, Chen CY, Wang YC. Loss of p16 and/or pRb protein expression in NSCLC. An immunohistochemical and prognostic study.
Lung Cancer
2001
;
31
:
163
–70.
30
Chen JT, Chen YC, Wang YC, Tseng RC, Chen CY. Alterations of the p16(ink4a) gene in resected nonsmall cell lung tumors and exfoliated cells within sputum.
Int J Cancer
2002
;
98
:
724
–31.
31
Masuda A, Takahashi T. Chromosome instability in human lung cancers: possible underlying mechanisms and potential consequences in the pathogenesis.
Oncogene
2002
;
21
:
6884
–97.
32
Essers J, van Steeg H, de Wit J, et al. Homologous and non-homologous recombination differentially affect DNA damage repair in mice.
EMBO J
2000
;
19
:
1703
–10.
33
Wang YC, Chen CY, Chen SK, Cherng SH, Ho WL, Lee H. High frequency of deletion mutations in p53 gene from squamous cell lung cancer patients in Taiwan.
Cancer Res
1998
;
58
:
328
–33.
34
Schwartz D, Rotter V. p53-dependent cell cycle control: response to genotoxic stress.
Semin Cancer Biol
1998
;
8
:
325
–36.
35
Classon M, Harlow E. The retinoblastoma tumour suppressor in development and cancer.
Nat Rev Cancer
2002
;
2
:
910
–7.
36
Aprelikova ON, Fang BS, Meissner EG, et al. BRCA1-associated growth arrest is RB-dependent.
Proc Natl Acad Sci U S A
1999
;
96
:
11866
–71.
37
Yarden RI, Brody LC. BRCA1 interacts with components of the histone deacetylase complex.
Proc Natl Acad Sci U S A
1999
;
96
:
4983
–8.
38
Rice JC, Massey-Brown KS, Futscher BW. Aberrant methylation of the BRCA1 CpG island promoter is associated with decreased BRCA1 mRNA in sporadic breast cancer cells.
Oncogene
1998
;
17
:
1807
–12.
39
DiNardo DN, Butcher DT, Robinson DP, Archer TK, Rodenhiser DI. Functional analysis of CpG methylation in the BRCA1 promoter region.
Oncogene
2001
;
20
:
5331
–40.
40
Davis PL, Miron A, Andersen LM, Iglehart JD, Marks JR. Isolation and initial characterization of the BRCA2 promoter.
Oncogene
1999
;
18
:
6000
–12.
41
Ludwig DL, Chen F, Peterson SR, Nussenzweig A, Li GC, Chen DJ. Ku80 gene expression is Sp1-dependent and sensitive to CpG methylation within a novel cis element.
Gene
1997
;
199
:
181
–94.
42
Kohno T, Yokota J. How many tumor suppressor genes are involved in human lung carcinogenesis?
Carcinogenesis
1999
;
20
:
1403
–10.
43
Sun JF, Sui JL, Zhou PK, et al. Decreased efficiency of γ-ray-induced DNA double-strand break rejoining in malignant transformants of human bronchial epithelial cells generated by α-particle exposure.
Int J Radiat Biol
2002
;
78
:
773
–80.
44
Zheng L, Li S, Boyer TG, Lee WH. Lessons learned from BRCA1 and BRCA2.
Oncogene
2000
;
19
:
6159
–75.