Nasopharyngeal carcinoma (NPC) is an endemic cancer with particular geographic and ethnic distribution within southern China. According to the International Agency for Research on Cancer (IARC, http://www.iarc.fr), NPC among the Chinese accounts for 43.2% of the total worldwide incidence and 40.6% of mortalities. Currently, it poses one of the most serious health problems within southern China. Etiologically, nasopharyngeal cancer is a multifactorial disease, and current evidence supports the involvement of Epstein-Barr virus infections along with other genetic and environmental factors.1 Genetic alterations in oncogenes and tumor suppressor genes (TSGs) are believed to be critical in the multi-step process that lead to the development of tumors. In NPC, however, relatively little is known about the molecular alterations of tumor suppressor genes. Previous studies have reported that the well-known TSGs, such as p53, Rb and p16, though commonly inactivated in many other malignancies, were rarely mutated in NPC.2, 3, 4 Recently, a number of genes differentially expressed in NPC have been identified in studies based on EST microarray techniques; however, none of the genes has been confirmed as a specific TSG associated with NPC.5, 6
Chromosomal abnormalities are common in tumor cells and the recurrent allelic losses from specific chromosomal regions may target resident TSGs. Nonrandom losses at chromosome 3p21.3 region have frequently been demonstrated in many human malignancies including lung, breast, kidney, cervical and nasopharyngeal cancers, indicating that 1 or more TSGs within this region may be involved in these tumors.7, 8, 9, 10, 11, 12, 13 Recently, several candidate TSGs have been identified in a 630 kb homozygous deletion region at 3p21.3 in human lung cancer.14 Notably, BLU gene, isolated in a nesting 120 kb region of minimal homozygous deletion, exhibited reduced or absent expression in 70% of lung cancer cell lines. Moreover, a few missense mutations of the gene were discovered in lung cancer cell lines, establishing it as an attractive TSG candidate at 3p21.3 region.14 The BLU gene spans ∼4.5 kb in the genomic sequence and consists of 11 or 12 exons (corresponding to GenBank accession numbers U70880 and U70824, respectively), coding for approximately 2 kb alternatively spliced transcripts. The encoded BLU protein (U70824) contains 440 amino acids and the sequence of 394–430 amino acid residues at the carboxyl terminus encodes a conserved DNA-binding domain, known as MYND zinc finger domain. Because the MYND domain was found in some transcription repressors and suppressors of cell cycle entry, for example, AML/ETO, BS69, and suppressin protein,15, 16, 17, 18 it is postulated that BLU may similarly function negatively in cell growth like its homologous counterparts. Therefore, BLU should be considered one of the promising TSG targets at 3p21.3 involved in human cancers.
Loss of heterozygosity (LOH) on chromosome 3p has been detected in as many as 95–100% of NPCs and 75% of precancerous lesions, suggesting that inactivation of the TSGs on chromosome 3p might be an early, critical event in the development of NPC.13, 19 Evidence from functional studies has targeted the NPC-associated TSGs to a 11.2 cM fragment at 3p21.3 covering the entire 630 kb lung cancer homozygous deletion region.14, 20 Although several genes located at this region have been characterized as candidate TSGs in lung cancer,14, 21, 22, 23, 24, 25 our preliminary studies only revealed SEMA3B and BLU gene significantly down-regulated on mRNA levels in primary NPC tumors, whereas most other genes exhibited no mutations nor altered expression in NPCs(data not shown). So far, the tumor suppressor role of SEMA3B in cancer cell lines has been well established,22, 23 but the characteristic of BLU in most human cancers including NPC has remained unknown. Therefore, to explore whether BLU at 3p21.3 might play an important role in the pathogenesis of NPC, we examined both genetic and epigenetic changes of BLU in primary NPC biopsied tissues and NPC cell lines. We screened for mutations in the coding region of this gene by direct sequencing and analyzed its expression by RT-PCR. Furthermore, to elucidate the mechanism possibly responsible for transcriptional silencing of the gene, we analyzed the methylation status of CpG islands in BLU promoter region by methylation specific PCR (MSP) and bisulphite sequencing. Our findings indicated high incidence of alterations of expression and promoter hypermethylation of BLU in NPC, suggesting a role for BLU in development of this cancer.
MATERIAL AND METHODS
Cell lines, primary tumor biopsies and blood samples
NPC cell lines including CNE1, CNE2, SUNE1, HNE1 and 5-8F (a derivative of SUNE1 cell line established in our labs), along with 45 primary NPC biopsies, were used in the investigation of genetic and epigenetic changes of BLU gene. For noncancerous controls, 15 biopsied nasopharyngeal tissues diagnosed with chronic inflammation were used. All cell lines were maintained in PRMI 1640 supplemented with 10% fetal bovine serum. Tumor biopsies and nasopharyngeal epithelial control tissues were obtained from patients in the Cancer Center, Sun Yet-sen University (Guangzhou, China) and were confirmed histologically prior to analysis. Peripheral blood lymphocytes obtained from 80 healthy volunteers in southern China were included for mutation confirmations.
PCR and DNA sequencing
PCR sequencing analysis was performed for screening mutations in BLU gene. DNA samples from biopsies and cell lines were prepared using DNAzol extraction kits (Invitrogen, San Diego, CA). Leukocyte DNA was extracted according to standard procedures. Exon sequences were amplified by mixing genomic DNA (20 ng) with 0.2 μM of each primer in 25 μl reaction buffer containing 200 μM each dNTPs and Taq polymerase at 94°C for 35 sec, 55–60°C for 35 sec and 72°C for 1 min over 35 cycles. (Primers are available upon request.) PCR products were purified with PCR Product Pre-Sequencing Kit (Amersham Pharmacia Biotech, Uppsala, Sweden) and sequenced on an Applied Biosystems 377 DNA sequencer using Dye Deoxy Terminator Cycle Sequencing Kit (Applied Biosystems, Foster City, CA).
Analysis of BLU expression by RT-PCR
The expression of BLU in the NPC primary tumors and cell lines was examined by RT-PCR analysis. Total RNA was extracted with Trizol reagent (Invitrogen, San Diego, CA) and 1 μg of total RNA was reverse transcribed using Reverse Transcription System (Promega, Madison, WI). PCR was performed with upper primer from exon 2 (5′-AACCAGCAGCATGAGAACCT-3′) and lower primer from exon 5 (5′-AGTTTGCGGTGGCAATAGTC-3′) of BLU gene at 94°C for 35 sec, 60°C for 30 sec and 72°C for 30 sec over 35 cycles. The resulting PCR fragment was 338 bp. GAPDH gene was amplified as an endogenous control with primer pairs 5′-AATCCCATCACCATCTTCCA-3′ and 5′-CCTGCTTCACCACCTTCTTG-3′, generating a 580 bp fragment.
DNA samples from the 5 NPC cell lines were subjected to bisulphite sequencing to determine the methylation status of the promoter region of BLU. Genomic DNA was denatured in 0.3 M NaOH at 42°C for 30 min. Cytosines were sulphonated in 3.1 M sodium bisulphite(Sigma Chemical Co., St. Louis, MO) and 0.5 mM hydroquinone (Sigma Chemical Co.) at 55°C for 24 hr. The DNA samples were desalted through a column (Wizard DNA Clean-up System, Promega), desulphonated in 0.3 M NaOH and precipitated. We amplified DNA sequences by mixing bisulphite-treated DNA (20 ng) with 0.2 μM primers MF1 and MR1 in 25 μl reaction buffer containing 200 μM each dNTPs and Taq polymerase at 94°C for 45 sec, 50°C for 1 min and 72°C for 1 min over 40 cycles. We performed a nested PCR using 1/25 of the amplified products and internal primers MF2 and MR2 at 94°C for 30 sec, 60°C for 30 sec and 72°C for 30 sec over 35 cycles. The primer pairs and product sizes are listed in Table I. The PCR products were purified and sequenced as described above.
Table I. Primer Sequences Used for Methylation ASSAT and PCR Product Size
PCR product size (bp)
Methylation specific PCR (MSP)
Twenty-three primary NPC samples from individuals who were 30–57 years of age, which exhibited aberrant BLU expression, and 9 available noncancerous nasopharyngeal samples from individuals who were 30–72 years of age, which served as controls, were further examined for the promoter methylation status by MSP assay. Genomic DNA was first modified by bisulphite treatment as earlier described. Primer pairs MF1 and MR1, which are not biased for either methylated or unmethylated DNA but which preferentially amplify bisulphite-modified DNA, were used in primary PCR. The nested PCR specifically amplifying methylated DNA was performed using primer pairs MSPF and MSPR (listed in Table I) at 94°C for 30 sec, 65°C for 30 sec and 72°C for 30 sec over 35 cycles. The nested PCR specifically amplifying unmethylated DNA was performed using primer pairs UMF and UMR (listed in Table I) at 94°C for 30 sec, 58°C for 30 sec and 72°C for 30 sec over 35 cycles. The wild-type DNA not treated by bisulphite was used for negative controls.
Re-expression of BLU by 5′-aza-2′-deoxycytidine treatment
To determine whether BLU expression could be restored by application of a demethylation agent, the NPC cell line CNE2, highly methylated and showing no expression of BLU, was exposed to concentrations of 0, 1.0, 2.5, 5.0 and 10 μM of 5′-aza-2′-deoxycytidine (Sigma Chemical Co.), respectively, for 4 days. The media and drugs were replaced every 24 hr. RT-PCR was performed as described above.
DNA variations of BLU in NPCs
We screened all exons and partial sequences of introns of BLU in 45 primary NPC tumors and 5 NPC cell lines for mutation by direct sequencing. A nucleotide 599G> A substitution leading to a Gly160Arg amino acid alteration in exon 5 of BLU (U70824) was observed in 4 of 45 primary tumors but not in any of the cell lines. To clarify whether it was a pathogenic mutation or a polymorphism, this alteration was further examined in the lymphocytes of 80 healthy controls. However, the identical base substitution was also found in one healthy subject, suggesting it was more likely a polymorphism. Additionally, 2 synonymous mutations including Arg231Arg in exon 7 and Gln59Gln in exon 2, as well as a few nucleotide substitutions in introns, were also found in this gene. Otherwise, no presumably pathogenic mutations of BLU were detected in NPCs. The missense mutations Asp198Gln and Arg407Gln previously reported in lung cancer were not found in our NPC samples.
Loss of expression of BLU in NPC cell lines and primary tumors
Because the inactivation of TSG may be achieved by inhibition of transcription, we examined the expression of BLU in some of the primary tumors, NPC cell lines and the nasopharyngeal tissue samples with chronic inflammation serving as noncancerous controls. The splice variant corresponding to GenBank accession number U70824 was expressed in the nasopharyngeal epithelia as revealed by sequenced RT-PCR products. RT-PCR analysis revealed that all 15 control samples expressed BLU mRNA, whereas only 8 of 36 primary tumors exhibited transcription levels comparable to those in the noncancerous controls. BLU mRNA levels were either absent or reduced in 78% (28 of 36) of the tumors investigated. In the 5 NPC cell lines, CNE1, CNE2, SUNE1, HNE1 and 5-8F, BLU transcripts were also missing (Fig. 1).
Aberrant promoter hypermethylation of BLU in NPC cell lines and primary tumors
Aberrant promoter methylation in tumors has been determined to lead to the loss of gene expression of several TSGs in human cancers.21, 22 To assess whether the loss of BLU expression in NPC had resulted from promoter hypermethylation, we determined the methylation status of 20 CpG islands in a 269 bp 5′ region of BLU (from −292 to −23 bp of the ATG translation start site) by sequencing sodium bisulphite-modified DNA from 5 NPC cell lines (Fig. 2a). The BLU nonexpressing tumor cell lines, CNE1, CNE2 and HNE1, exhibited methylation of all CpG dinucleotide sites in this region, whereas SUNE1 and its derivative 5-8F, though not expressing BLU, showed only partial methylation in most or all CpG sites examined (Fig. 2b,c).
To confirm that promoter hypermethylation contributed to the silencing of BLU in the NPC cell lines, we assessed the effect of 5′-aza-2′-deoxycytidine, a drug that inhibits DNA methylase, on BLU expression. We exposed the BLU nonexpressing cell line CNE2 to 5′-aza-2′-deoxycytidine and found re-expression of BLU by this cell line (Fig. 2d), suggesting that repression of BLU transcription might at least be partially mediated by promoter methylation.
We further analyzed the methylation status of BLU promoter by MSP in 23 primary NPC tumors that exhibited absent or reduced BLU expression, which were derived from individuals who were 30–57 years of age, and in 9 non-neoplastic nasopharyngeal epithelial tissues with chronic inflammation from individuals who were 30–72 years of age. Genomic DNA was treated with sodium bisulphite and tested for the presence of methylated and unmethylated CpG dinucleotides in the promoter region covered by MSP primers. All of the primary NPC samples amplified unmethylated bands which were expected because of the heterogeneity of the NPC tumors, or contamination of normal tissues or infiltrating lymphocytes. Among the 23 carcinomas examined, 17 exhibited methylation in the BLU promoter region. By contrast, methylated alleles were detected in only 2 samples derived from individuals who were 63 and 72 years of age, respectively. Representative examples are displayed in Figure 2e. The methylated status of BLU promoter in cell line CNE2 and partial methylation status in SUNE1 were also confirmed by MSP.
The high frequency of allelic loss on chromosome 3p in NPC has been described previously by our group and others.13, 19 Allelic loss of chromosome 3p, including 3p21.3 region, was observed even in 75% of precancerous lesions and in 73.9% of normal nasopharyngeal epithelia from high-risk groups,13 strongly suggesting that the inactivation of TSGs in this region may be causally involved in the earliest steps of pathogenesis of NPC. Functional studies have targeted tumor suppressor activity to an 11.2 cM fragment at 3p21.3 region in a NPC cell line, making it a promising locus for the identification of NPC-associated TSG (s).20
The BLU gene, residing within a 120 kb subregion of minimal homozygous deletion at 3p21.3, has recently been identified as a candidate TSG due to the occurrence of missense mutations and loss of its expression in lung cancer.14 At present, little is known about the function of this gene in normal cell physiology. However, BLU protein contains a conserved MYND domain, suggesting its possible role in transcription regulation. The MYND domain contains repeated patterns of cysteine and histidine residues that are reminiscent of zinc finger involved in specific protein-protein interactions. The domain is also present in human proteins including AML/ETO fusion protein in acute myelogenous leukemia, cell cycle inhibitor suppressin and BS69 protein, a transcriptional co-repressor and adenovirus E1A-binding protein.15, 16, 17, 18 Recent studies have indicated that the MYND domain can bind N-CoR, thereby recruiting a histone deacetylase complex that causes transcriptional repression.18, 26, 27 Since the MYND domain is required for repression of basal transcription by the AML/ETO protein and for binding of adenoviral oncoprotein E1A by the BS69 protein,15, 17 it may also be essential for the BLU to be possibly involved in negative transcriptional regulation in cells. Interestingly, Ansieau et al.17 has recently found that the MYND domain in the BS69 protein bound to the EBNA2 oncoprotein of Epstein-Barr (EB) virus. Since the EB virus plays an important role in NPC tumorigenesis, it is speculated that BLU may function as a tumor suppressor via binding and interference with certain EB virus oncoproteins through MYND domain mediated interaction.
We assessed the TSG candidacy of BLU in NPC by examining both genetic and epigenetic changes of this gene in primary NPC tumors and cell lines. Intron-specific primers used for mutation analysis, encompassing each exon of BLU in turn, amplified the expected PCR products in each sample, suggesting that no homozygous deletion had occurred in this gene. Although a missense alteration of Gly160Arg in exon 5 was found in 4 NPC patients, it also occurred in 1 unaffected individual, suggesting the probability of a polymorphism. It is notable that, although not in the MYND domain, this substitution is a nonconservative one that may introduce positive charge into the amino acid sequence. Therefore, additional studies are needed to determine whether this polymorphism has functional consequences or whether it may be associated with the risk to NPC in our population. Our results have provided no evidence to verify the presence of presumably pathogenic mutations in the coding sequence of BLU. This indicated that, in the corresponding tumors, BLU was probably not a primary target for the mutagenic process that leads to carcinogenesis. In similar results obtained by Lerman et al.,14 only a low mutation rate of BLU in lung cancer was found; therefore, it was proposed that mechanisms other than mutation, such as transcriptional inhibition, may have led to the inactivation of this gene. Indeed, we have observed frequent alteration of mRNA expression of BLU in NPC. Absent or reduced mRNA levels of BLU were found in 78% of primary tumors and in all tumor cell lines examined, suggesting that the BLU function had been abrogated by transcriptional repression in most NPCs.
Recently, promoter hypermethylation has been recognized as a common mechanism leading to inactivation of TSGs.28, 29 Several TSGs or genes associated with tumorigenesis, such as RB, VHL, BRCA1, DAPK and RASSF1A, have been demonstrated to have their expression silenced by tumor-acquired promoter hypermethylation, some of which occur in NPC.30, 31, 32, 33, 34 In BLU gene, we also detected a high frequency of promoter hypermethylation and found it correlated with the loss of BLU expression in NPC. We observed extensive promoter methylation in 3 BLU nonexpressing cell lines, CNE1, CNE2 and HNE1, and restoration of BLU expression by exposing 1 cell line, CNE2, to the methylase inhibitor 5′-aza-2′-deoxycytidine. Furthermore, 74% (17 of 23) of primary tumors in which BLU was down-regulated exhibited aberrant promoter methylation, in contrast to the situation that no methylation was shown in most age-matched noncancerous controls. Infrequent methylation of TSGs in normal epithelia related with increasing age has been reported in colon mucosa and in bladder urothelium.35, 36, 37, 38 Here we also found methylation of BLU in 2 normal nasopharyngeal epithelia derived from individuals who were 63 and 72 years old but not in younger cases (30–55 years old) that matched the NPC groups (30–57 years old) in age. It is suggested that BLU may be methylated both in a cancer-specific and an age-related manner, the latter of which may conform to the hypothesis that age-related methylation in normal epithelia mark and may lead to field defect in association with acquired predisposition to cancer. Our results suggested a role of promoter methylation in silencing of BLU gene in the NPC. However, there may also be the possibility of other mechanisms involved in repression of BLU transcription in this cancer. In the tumor cell lines SUNE1 and its derivative 5-8F, only partially methylated CpG sites were observed despite the fact that BLU transcripts were lost in the cells. Consistently, 26% (6 of 23) of primary tumors exhibited neither methylation in their promoter regions nor normal BLU mRNA levels. Therefore, alternative mechanisms other than promoter hypermethylation, such as promoter mutation, loss of transcriptional activators, binding of suppressor proteins to the promoter etc., might also affect BLU transcription. Given its location within a region frequently undergoing LOH in NPC,13, 19 together with our results, we surmise that allele loss combined with promoter hypermethylation, as well as other unknown mechanisms, may contribute to the inactivation of BLU in NPC.
In summary, we have identified a high frequency of alterations of the BLU gene in nasopharyngeal carcinoma. Absence or reduction of expression was observed in 78%(28/36) of primary tumors, 74%(17/23) of which demonstrated hypermethylation in the BLU promoter region, in contrast to the situation in most noncancerous nasopharyngeal epithelia. This strongly suggests that inactivation of BLU mediated by promoter hypermethylation or other unknown mechanisms may be closely associated with the development of NPC. Based upon observations that BLU is down-regulated in the majority of lung and nasopharyngeal cancers, in addition to its localization in a frequently deleted region and its putative function in transcriptional regulation, we suggest that BLU may be one of the critical target TSGs on chromosome 3p21.3 involved in tumorigenesis. Nevertheless, it cannot be excluded that not BLU alone, but BLU together with other genes in this region, may exert a tumor suppressor function. Recent studies have identified multiple genes neighboring BLU, such as RASSF1A, SEMA3B, SEMA3F, FUS1, 101F6, NPRL2 etc. but not BLU itself, as candidate TSGS because of their tumor suppressor activities upon introduction into lung or ovarian cancer cell lines.21, 22, 23, 24, 25 Therefore, our results may specifically apply to NPC, and there may be a possible combined effect of several of these genes in the region contributing to the NPC carcinogenesis. Further studies are required to address this hypothesis and to determine the specific role of BLU in the development of NPC.