ISSN 1672-9145 Acta Biochimica et Biophysica Sinica 2006, 38(5): 349–355 CN 31-1940/Q

©Institute of Biochemistry and Cell Biology, SIBS, CAS

Genetic and Epigenetic Alterations of DLC-1, a Candidate TumorSuppressor Gene, in Nasopharyngeal Carcinoma

Dan PENG, Cai-Ping REN*, Hong-Mei YI, Liang ZHOU, Xu-Yu YANG, Hui LI, and Kai-Tai YAO*Cancer Research Institute, Xiangya School of Medicine, Central South University, Changsha 410078, China

Abstract The DLC-1 gene, located at the human chromosome region 8p22, behaves like a tumorsuppressor gene and is frequently deleted in diverse tumors. The deletion of 8p22 is not an uncommon eventin nasopharyngeal carcinoma (NPC), therefore we explored the expression levels of the DLC-1 gene in NPCsand NPC cell lines by reverse transcription-polymerase chain reaction. The results showed the mRNA levelof DLC-1 was downregulated. To identify the mechanism of DLC-1 downregulation in NPC, we investigatedthe methylation status of the DLC-1 gene using methylation-specific PCR, and found that 79% (31 of 39) ofthe NPC tissues and two DLC-1 nonexpressing NPC cell lines, 6-10B and 5-8F, were methylated in theDLC-1 CpG island. Microsatellite PCR was also carried out, and loss of heterozygosity was found at fourmicrosatellite sites (D8S552, D8S1754, D8S1790 and D8S549) covering the whole DLC-1 gene with ratiosof 33% (4 of 12 informative cases), 18% (2 of 11), 5% (1 of 18), and 25% (3 of 12), respectively. Takentogether, our results suggest that DLC-1 might be an NPC-related tumor suppressor gene affected by aberrantpromoter methylation and gene deletion.

Key words DLC-1; nasopharyngeal carcinoma; hypermethylation; loss of heterozygosity

Received: February 18, 2006 Accepted: March 6, 2006*Corresponding authors:Cai-Ping REN: Tel, 86-731-2355066; Fax, 86-731-4360094; E-

mail, rencaiping@xysm.netKai-Tai YAO: Tel, 86-731-4805451; Fax, 86-731-4360094; E-mail,

ktyao@fimmu.com DOI: 10.1111/j.1745-7270.2006.00164.x

The tumorigenesis of nasopharyngeal carcinoma (NPC)is a multi-step process involving various factors, includingEpstein-Barr virus infection and accumulation of epigeneticand genetic alterations. Because of its exceptionally highincidence in southern China, discovering the molecularbasis of NPC is a priority in our research. Genome-widemicrosatellite polymerase chain reaction (PCR) revealedhigh frequency of loss of heterozygosity (LOH) onchromosome 8p22, and published reports indicate 8p22might be a promising region containing candidate tumorsuppressor genes of NPC [1].

DLC-1 (deleted in liver cancer-1), a candidate tumorsuppressor gene, was isolated from human hepatocellularcarcinoma (HCC) by the PCR-based subtractive hybridiza-tion approach. DLC-1 shares high sequence similarity withrat p122RhoGAP [2], which is a GTPase-activating protein(GAP) for Rho family proteins [3]. Rho family proteins

play essential roles in regulating diverse biological functions,including cytoskeletal organization, cell adhesion, and cellcycle progression [4−6], and are known to be involved inRas-mediated oncogenic transformation [7]. A GAP servesas a negative regulator of Rho proteins by stimulating itsintrinsic GTPase activities [8], thus it may function as atumor suppressor by inactivating Rho proteins. Trans-fection of the DLC-1 cDNA into HCC cell lines withhomozygous deletions of the gene caused a strong inhibi-tion of cell growth [9]. In addition, transfection of theDLC-1 cDNA into non-small cell lung carcinoma cell linescaused a significant inhibition of cell proliferation and adecrease in colony formation in vitro, and abolishedtumorigenicity of non-small cell lung carcinoma cell linesin nude mice in vivo, which clearly showed that DLC-1might exert tumor suppressor activity [10].

The DLC-1 gene was found to be located at 8p22, aregion of LOH in a number of human cancers such asliver, lung, breast, colon, prostate, and head and neckcancers [11−15]. Of note, DNA methylation of DLC-1was found in lung, liver, gastric and primitive neuroecto-

350 Acta Biochim Biophys Sin Vol. 38, No. 5

©Institute of Biochemistry and Cell Biology, SIBS, CAS

dermal tumors [10,16−18], which strongly suggests thathypermethylation in the DLC-1 promoter might performan important role in the transcriptional silencing of the gene.In this study, our data indicate that genetic and epigeneticalterations are involved in the inactivation of DLC-1 in NPC.

Materials and Methods

Samples and cell lines

Forty-one poorly-differentiated NPC biopsies (T1−T41)of primary tumors and an additional 20 NPC biopsies withmatched constitutional DNA from peripheral bloodlymphocytes were obtained from NPC patients withconsent before treatment at the Hunan Cancer Hospital(Changsha, China). The male to female ratio of the NPCpatients was 2.73:1 (30:11), and the age range was 31−62years (mean age, 48.15 years). In addition, 16 normalnasopharynx (NP) tissues were obtained from patientswithout NPC at Hunan Cancer Hospital. The tissues werestudied simultaneously as controls under similar experi-mental conditions. The male to female ratio of the patientswithout NPC was 3:1 (12:4), and the age range was 26−69 years (mean age, 46.42 years). All the specimens werereviewed by an otorhinolaryngologic pathologist. FreshNPC tissues or normal tissues were snap-frozen in liquidnitrogen and stored until required.

Four NPC cell lines were obtained: HNE1 and CNE2were from the Cancer Research Institute, Xiangya Schoolof Medicine, Central South University (Changsha, China);5-8F and 6-10B were from the Cancer Center, Sun Yet-Sen University (Guangzhou, China). All were maintainedin RPMI 1640 containing 10% fetal bovine serum at 37 ºC ina humidified 5% CO2 atmosphere. Cells were harvestedfor total RNA and genomic DNA extraction at 70%−80%confluence.

RNA preparation and reverse transcription-poly-merase chain reaction (RT-PCR)

Total RNA was prepared from NPC cell lines andcryopreserved NPC tissue samples or normal NP tissuesamples and was extracted with TRIzol Reagent(Invitrogen, Carlsbad, USA). RNA was quantified at 260nm by a spectrophotometer and RNA quality was assessedby visualization of clear 18S and 28S RNA bands afterelectrophoresis through agarose gels. cDNA was syn-thesized from total RNA using oligo(dT) as the primer witha commercially available reverse transcription system(Promega, Madison, USA). Reverse transcription was

performed in a total volume of 20 µl containing 5 mMMgCl2, 1×buffer, 1 mM dNTPs, 20 U RNasin, 14.4 UAMV reverse enzyme, 1 pM oligo(dT), 2 µg RNA. Thereaction was performed according to the instructions. Apair of primers was used to amplify the 299 bp region ofDLC-1, and the primer sequences were as follows: 5'-AGCCAATTCTGGAACCAAAC-3' (forward) and 5'-GGAAGACCCCAAGAAACACA-3' (reverse). At the sametime, a 550 bp fragment of glyceraldehyde phosphatedehydrogenase (GAPDH) was amplified as a control.Negative controls for PCR were run in reagent mixtureswithout RNA or reverse transcriptase. The PCR wasterminated at the exponential phases: 32 cycles for DLC-1and 24 cycles for GAPDH, including 1 cycle of hot startat 95 ºC for 5 min, followed by amplification at 94 ºC for30 s, 58 ºC for 30 s, and 72 ºC for 30 s, and a final extensionat 72 ºC for 5 min.

DNA preparation

Genomic DNA was prepared from NPC cell line pelletsand NPC tissues using an improved method of extractinghigh molecular weight DNA with phenol/chloroform, asdescribed previously [19]. Briefly, lysis overnight at 37 ºCin 500 µl salt/EDTA buffer, 25 µl 20% (W/V) sodiumdodecylsulfate and 25 µl proteinase K solution (2 mg/ml),followed by phenol/chloroform extraction and stored at−20 ºC.

Bisulfite modification of DNA and methylation analysis

The sodium bisulfite reaction converts unmethylatedcytosine in DNA to uracil while leaving the methylcytosineunchanged, so that methylated and unmethylated allelescan be distinguished by methylation-specific PCR (MSPCR)or sequencing. Genomic DNA from tumor samples waspurified using the standard proteinase K digestion andphenol/chloroform extraction method, as described above.Then sodium bisulfite treatment was carried out using aprotocol modified from Clark et al. [20]. Ten microgramsof genomic DNA was denatured with 0.3 M NaOH andmixed with 300 µl of 10 mM hydroquinone (Sigma, St.Louis, USA), 5.2 ml of 3.6 M NaHSO3 (pH 5.0; Sigma),covered with paraffin oil then deaminated in the dark for4 h at 55 ºC. Bisulfite-treated DNA was purified withpurification columns (TaKaRa, Shiga, Japan). Subsequentlypurified DNA samples were desulfonated with 0.3 M NaOHat room temperature for 10 min, neutralized with ammo-nium acetate, ethanol precipitated, and resuspended in20 µl of Tris-EDTA buffer.

Bisulfite-treated DNA was amplified by PCR withmethylation-specific primer pairs described in previously

May, 2006 Dan PENG et al.: Genetic and Epigenetic Alterations of DLC-1 in NPC 351

http://www.abbs.info; www.blackwellpublishing.com/abbs

published reports [17]. The primer pairs were able todiscriminate between methylated and unmethylated allelesof the DLC-1 gene. MSPCR was carried out under thefollowing conditions: hot start at 95 ºC for 5 min, followedby 35 cycles of 94 ºC for 30 s, 55 ºC for 30 s, and 72 ºCfor 30 s, and a final extension at 72 ºC for 10 min. ThePCR reaction conditions for the unmethylated allele of theDLC-1 gene were the same as those for MSPCR.

Allelic loss analysis

To examine the allelic loss in the DLC-1 locus, weselected four microsatellite markers on chromosome 8p22,which covered a relatively wide chromosomal region ofapproximately 1.6 Mb encompassing the DLC-1 gene.Primers for amplification of microsatellite markers D8S549(maps to 1.0 Mb upstream of DLC-1), D8S1754 (locatesat DLC-1), D8S1790 (locates at DLC-1), and D8S552(maps to 0.2 Mb downstream of DLC-1) are availablethrough the genome database on the National Center forBiotechnology Information website (http://www.ncbi.nlm.nih.gov). The microsatellite markers were amplified byPCR from 50−100 ng DNA extracted from human NPCtissues and their matched blood samples which were usedas controls. Reaction was initiated at 95 ºC for 5 min,28 cycles of 94 ºC for 15 s, 58 ºC for 15 s, and 72 ºC for15 s, followed by a final elongation at 72 ºC for 5 min.

After amplification, 6−8 µl of the reaction mixture wasmixed with 8 µl of loading dye (95% formamide, 20 mMEDTA, 0.05% bromophenol blue, and 0.05% xylenecyanol), heat denatured, chilled on ice, then electrophoresedon a 6% polyacrylamide gel containing 8 M urea. The DNAbands were visualized by silver staining. LOH was scoredif one of the heterozygous alleles showed at least a 50%reduction in intensity in tumor DNA as compared with thematched blood DNA.

Grayscale scanning and statistical analysis

RT-PCR products were separated through 1.5% agarosegel containing ethidium bromide. The sizes of the RT-PCRproducts were 299 bp for DLC-1 and 550 bp for GAPDH.The intensity of each band was measured by Image MasterVDS (Pharmacia Biotech, Piscataway, USA), and analyzedby VDS software version 2.0 for band quantification. Theexpressions of DLC-1 in tumors and normal tissues wereinvestigated after they were normalized by transformingthem into two groups of ratios of the band intensity ofDLC-1 over that of GAPDH of the same sample. EachRT-PCR reaction was carried out in triplicate.

Otherwise, the association of DLC-1 expression withhistological type of tissue, gender and node metastasis was

examined by Mann-Whitney U-test. The association ofDLC-1 methylation with gender was analyzed by Fisher’sexact test, and the association of DLC-1 methylation withage was analyzed by Mann-Whitney U-test, as appropriate.All statistical analysis was performed using spss version10.0 statistical software for Windows (SPSS, Chicago,USA). P<0.05 was regarded as statistically significant.

Results

Downregulation of DLC-1 expression in NPC tissuesand NPC cell lines

To evaluate the relationship between mRNA expressionof DLC-1 and NPC, RT-PCR was carried out in 41 samplesof NPC tissues and 16 normal NP tissues as well as fourNPC cell lines. Our analysis revealed that all the normalNP tissues demonstrated stable DLC-1 gene expression,which is consistent with the earlier finding that DLC-1was widely expressed in normal tissues [2]. The DLC-1mRNA level was normalized against the housekeeping geneGAPDH (Fig. 1). DLC-1 mRNA was undetectable in

Fig. 1 DLC-1 mRNA levels assessed by reverse transcription-polymerase chain reaction(A) Obvious downregulation of DLC-1 in four nasopharyngeal carcinoma (NPC)cell lines (HNE1, CNE2, 6-10B and 5-8F) compared with normal nasopharyngeal(NP) tissue (N3). (B) Obvious downregulation of DLC-1 in primary NPC biopsies(cases T1–T3) compared with normal NP biopsies (cases N1–N3). H2O was usedas the negative control, and glyceraldehyde phosphate dehydrogenase (GAPDH)was used as the endogenous control. M, marker.

352 Acta Biochim Biophys Sin Vol. 38, No. 5

©Institute of Biochemistry and Cell Biology, SIBS, CAS

test, P=0.003) (Table 2). No relationship was foundbetween methylation and gender in NPCs (Fisher’s exacttest, P=0.682) (Table 3).

Allelic deletion of DLC-1 in NPCs

Previous reports manifested that the deletion of theDLC-1 gene was associated with downregulation of

Fig. 2 Hypermethylation of the DLC-1 gene in nasopharyn-geal carcinoma (NPC)(A) Methylation-specific polymerase chain reaction (MSPCR) results of the DLC-1 gene in NPC tissues and NPC cell lines (6-10B and 5-8F). Normal nasopharyn-geal tissues (N1 and N2) and NPC tissues (T3) showed unmethylated (U) poly-merase chain reaction products only, whereas NPC tissues (T1, T2 and T4) andNPC cell lines (6-10B and 5-8F) demonstrated both unmethylated (U) and methy-lated (M) polymerase chain reaction products. (B) A map of the CpG island of theDLC-1 gene spanning 668 bp upstream of the ATG translation start codon. Thetranscription start site is marked. The locations of MSPCR primers that recognizeunmethylated (uF and uR) or methylated (mF and mR) DNA are shown.

Table 3 Correlation of DLC-1 methylation with the genderof patients with nasopharyngeal carcinoma (NPC)

DLC-1 methylation Cases (n)

Male Female

+ 22 5− 9 3No difference in DLC-1 methylation between NPC male and female patients

(P=0.682).

13 cases, and the overall DLC-1 expression in NPCs wasreduced significantly (P=0.001) (Table 1). DLC-1transcripts were nearly undetectable in HNE1, CNE2,5-8F and 6-10B NPC cell lines. Statistical analysis revealedno significant difference in expression of DLC-1 betweenmales and females or node metastasis and non-nodemetastasis (P=0.063 and P=0.057, respectively). We alsoranked the DLC-1 expression and age, and no statisticalcorrelation between expression and age was found usingSpearman’s rank coefficient analysis (P=0.370).

Table 2 Correlation of DLC-1 methylation with the age ofnasopharyngeal carcinoma patients

DLC-1 methylation Cases (n) Values

Median (years)† P

+ 31 47.5 0.003− 8 32.9

† Median age of patients. P<0.05 was regarded as statistically significant.

Methylation status of the DLC-1 CpG island in NPCtissues and NPC cell lines

To explore the potential role of CpG island methylationin the transcriptional silencing of the DLC-1 gene, wechecked the methylation status in the NPC tissues byMSPCR, which can specifically amplify the methylatedand unmethylated alleles, after chemically modifying theisolated DNA with sodium bisulfite [17]. Our MSPCRanalysis showed that the DLC-1 CpG island was methylatedin 79% (31 of 39) of the NPC tissues and two NPC celllines (5-8F and 6-10B) with loss of DLC-1 expression.Methylated DLC-1 alleles were also observed in 30.7%(4 of 13) of the normal samples. A representative illustrationof MSPCR is shown in Fig. 2(A). In tumorous samples,T1, T2 and T4 showed both the methylation-specific andunmethylation-specific bands, but T3 showed only theunmethylation-specific band. In non-tumorous samples,N1 and N2 showed only the unmethylation-specific band.A significant difference in age was found between themethylated and unmethylated groups (Mann-Whitney U-

Table 1 Correlations of DLC-1 expression with clinicopatho-logical features

Feature Cases (n) Value

Median† P

Type Tumor 41 0.4507 0.001Normal 16 0.9331

Gender Male 42 0.6364 0.063Female 15 0.5853

Node + 27 0.5679 0.057− 14 0.4900

† median of relative amount (intensity of DLC-1 over that of glyceraldehydephosphate dehydrogenase of the same sample). P<0.05 was regarded as statisti-cally significant. +, node metastasis; −, non-node metastasis.

May, 2006 Dan PENG et al.: Genetic and Epigenetic Alterations of DLC-1 in NPC 353

http://www.abbs.info; www.blackwellpublishing.com/abbs

DLC-1 in multiple tumors [2]. To determine whether thedownregulation of DLC-1 in NPCs was due to genomicdeletion of DLC-1, the allelic status of DLC-1 was inves-tigated by microsatellite analysis. As shown in Fig. 3,D8S552 and D8S549 flank a 1.6 Mb region on chromo-some 8p22 containing the DLC-1 locus, and D8S1754 andD8S1790 are intragenic markers mapped in the DLC-1gene. The LOH frequencies for D8S552, D8S1754,D8S1790, and D8S549 were 33% (4 of 12 informativecases), 18% (2 of 11), 5% (1 of 18), and 25% (3 of 12),respectively. In total, 35% (7 of 20) of informative NPCsdemonstrated LOH of at least one marker: case 6 displayedLOH at site D8S549; case 8 displayed LOH at site D8S552;case 10 displayed LOH at site D8S549; case 13 displayedLOH at site D8S1790; and case 20 displayed LOH at siteD8S552. Cases 9 and 16 were likely to have lost one alleleof DLC-1, for these two cases displayed LOH at both ofthe intragenic markers and the flanking markers. Thesefindings indicated that allelic loss at the DLC-1 locus wasnot uncommon in NPCs.

several kinds of tumors acting like a tumor suppressorgene [9,10,21−23]. In our study, RT-PCR of DLC-1 inNPC tissues and NPC cell lines showed manifestdownregulated expression of DLC-1 mRNA compared withnormal NP tissues, therefore DLC-1 might be a candidateNPC tumor suppressor gene. The loss of DLC-1 mRNAexpression was not only observed in tumors with genomicdeletions but also in tumors without homozygous deletion,such as HCCs and gastric cancer cells [16,17], suggestingthat multiple mechanisms are responsible for inactivatingDLC-1 in these tumors.

To find the reasons leading to the reduced level ofDLC-1 mRNA, we carried out LOH studies to investigatethe allelic status of DLC-1. While, our results showed that35% (7 of 20) informative cases of NPC biopsies demon-strated LOH in at least one site, strongly suggesting thatdownregulation of DLC-1 expression might not be onlydue to genomic deletions compared with the moreconspicuous downregulation of DLC-1 detected in NPCsand epigenetic mechanism remained investigated.

There is a growing realization that LOH at a given tumorsuppressor gene locus is not a prerequisite for neoplasia,following the intensive comprehension of epigeneticalterations in tumorigenesis. Methylation of the CpG islandis an alternative way of making genes inactive. Structurally,the 5'-upstream region from the start codon of the DLC-1gene (GenBank accession No. AF404867) has a high G+Ccontent (73%), which meets the criteria for a CpG island[Fig. 2(B)]. Recently, hypermethylation of the DLC-1promoter region was reported in HCC, lung cancer andgastric cancer unceasingly [10,16−18,24]. MSPCRanalysis in our study showed that the DLC-1 CpG islandwas methylated in 79% (31 of 39) of the NPC tissues.Confusingly, the theme at the center of this controversy iswhether significance can be attached to the methylatedpromoter in NPCs, considering those CpG islands methy-lated in four samples of 13 normal tissues. General reasonsfor the methylation present in histologically normal sampleshave been provided by others [24], such as infiltratingtumor cells, epigenetic field defect, imprinting, or tissue-specific methylation. Notwithstanding, we want to discussindividual possible mechanisms according to our results.

Promoter hypermethylation can occur in conjunctionwith allelic loss and/or mutation and is regarded as analternative form of “knockout” in bi-allelic inactivation.According to accepted knowledge, within the wholesequence of the CpG island of a gene, methylation takingplace at transcription factor binding sites takes over-whelming responsibility in silencing a certain gene. Actually,Bais et al. [25] reported that complex high to low methy-

Fig. 3 Allelic losses of the DLC-1 locus in primary nasopha-ryngeal carcinoma (NPC)(A) Three representative cases (numbered 9, 10 and 16) of primary NPC showingallelic losses at three microsatellite sites (D8S552, D8S549 and D8S1754) cover-ing the DLC-1 locus. Arrows point to the alleles deleted in the tumor DNA. N,DNA from matched peripheral blood lymphocytes; T, DNA from tumor biopsies.(B) Deletion map of all cases observed. ■, loss of heterozygosity; □, retention ofalleles; ○, none informative; ND, not done.

Discussion

Since DLC-1 was isolated from human HCC by PCR-based subtractive hybridization approach, subsequentreports have shown that DLC-1 was associated with

354 Acta Biochim Biophys Sin Vol. 38, No. 5

©Institute of Biochemistry and Cell Biology, SIBS, CAS

lation levels are found in primary breast tumors and theirnormal counterparts at multiple regions within the promotersequence of a potential tumor suppressor gene namedCBFA2T3B. They revealed that only a few cell lines dis-played clear hypermethylation in association with reducedexpression of CBFA2T3B in breast cancer. A second-roundreal-time MSPCR (quantitating methylation levels)combined with real-time RT-PCR (examining mRNA levels),still revealed the strong correlation between reducedexpression and “hypermethylation” at several certainregions out of a “basal” methylation existing in the promoterCpG island. For this reason, we selected those primerswith amplification products residing within a consensusand predicted specificity protein (Sp1) binding site inMSPCR. No data has directly proved the existence of theSp1 binding site in DLC-1 utilized in MSPCR by experiment,even though the potential transcription factor binding sitesare delicately predicted, precisely designed and adoptedby many researchers, including us. Methylation of DLC-1found in normal NP tissues might be due to the undefinedSp1 binding site. However, further studies using DLC-1promoter constructs are required to identify the functionalroles of the Sp1 binding sites in DLC-1 silencing.

Furthermore, a more interesting finding in our researchis that a significant difference was found in age between themethylated group and the unmethylated group (P=0.003).Higher age tends to correlate with higher frequency ofmethylation. The available explanation for this result comesfrom the role of the environment, particularly carcinogens,in causing epigenetic changes. Previously published reportsshowed that aberrant methylation of RASSF1A wasassociated with exposure to smoke and correlated to along-term smoking habit [26,27]. It is not a rare phenome-non that hypermethylation of RASSF1A has been detectedfrom cells in the sputum and bronchioloalveolar lavagesof smokers [28], which might also provide us with analternative way of studying the methylation appearing insome normal tissues in our study. Perhaps there is some-thing associated with the methylation of DLC-1 in our livingcircumstances, even though there has been no report, untilnow, demanding intensive investigation. Andrew P.Feinberg hypothesized that genetic and epigenetic alterationsmight interact, in that epigenetic alterations might influencethe effect of subsequent genetic insults during the courseof cancer initiation and the probability of cancer develop-ment depends on the presence of epigenetic alterationsafter genetic alterations have occurred [29]. Thus, con-sidering the environmental effects on epigenetic alterations,as we age, the number of epigenetic errors increasefollowed by the increasing probability of carcinogenic

progression after genetic changes.Lack of DLC-1 expression was detected in one medullo-

blastoma cell line [18], in which no genomic deletion,somatic mutation, or promoter hypermethylation wasfound. A similar observation was reported in two DLC-1-nonexpressing gastric cancer cell lines without detectablemethylated alleles of DLC-1 [17]. In this study, the authorswere able to reactivate DLC-1 expression by treating thesecells with a histone deacetylase inhibitor [17]. Thus, thereis another epigenetic mechanism mediating transcriptionalsilencing of DLC-1 at least in gastric cancer, which ishistone deacetylation. These findings have led to thespeculation that perturbation in the chromatin environmentwith histone deacetylation might contribute to transcrip-tional silencing of the DLC-1 gene in gastric cancer cells.

In summary, our results indicate that mRNA levels ofDLC-1 were downregulated in NPC, and promoterhypermethylation and LOH of DLC-1 were commonlyfound in NPC tissues. The results suggest that bothpromoter hypermethylation and LOH of DLC-1 might haveoccurred in NPC, and they might take major responsibilityfor the downregulation of DLC-1 in NPC.

References

1 Shao JY, Wang HY, Huang XM, Feng QS, Huang P, Feng BJ, Huang LX etal. Genome-wide allelotype analysis of sporadic primary nasopharyngealcarcinoma from southern China. Int J Oncol 2000, 17: 1267−1275

2 Yuan BZ, Miller MJ, Keck CL, Zimonjic DB, Thorgeirsson SS, PopescuNC. Cloning, characterization, and chromosomal localization of a gene fre-quently deleted in human liver cancer (DLC-1) homologous to rat RhoGAP.Cancer Res 1998, 58: 2196−2199

3 Homma Y, Emori Y. A dual functional signal mediator showing RhoGAPand phospholipase C-delta stimulating activities. EMBO J 1995, 14: 286−291

4 Hall A. Rho GTPases and the actin cytoskeleton. Science 1998, 279: 509−514

5 Moon SY, Zheng Y. Rho GTPase-activating proteins in cell regulation.Trends Cell Biol 2003, 13: 13−22

6 Etienne-Manneville S, Hall A. Rho GTPases in cell biology. Nature 2002,420: 629−635

7 Qiu RG, Chen J, McCormick F, Symons M. A role for Rho in Rastransformation. Proc Natl Acad Sci USA 1995, 92: 11781−11785

8 Morii N, Kumagai N, Nur-E-Kamal MS, Narumiya S, Maruta H. Rho GAPof 28 kDa (GAP2), but not of 190 kDa (p190), requires Asp65 and Asp67 ofrho GTPase for its activation. J Biol Chem 1993, 268: 27160−27163

9 Ng IO, Liang ZD, Cao L, Lee TK. DLC-1 is deleted in primary hepatocellu-lar carcinoma and exerts inhibitory effects on the proliferation of hepatoma celllines with deleted DLC-1. Cancer Res 2000, 60: 6581−6584

10 Yuan BZ, Jefferson AM, Baldwin KT, Thorgeirsson SS, Popescu NC, ReynoldsSH. DLC-1 operates as a tumor suppressor gene in human non-small celllung carcinomas. Oncogene 2004, 23: 1405−1411

11 Fujiwara Y, Ohata H, Emi M, Okui K, Koyama K, Tsuchiya E, Nakajima T,

May, 2006 Dan PENG et al.: Genetic and Epigenetic Alterations of DLC-1 in NPC 355

http://www.abbs.info; www.blackwellpublishing.com/abbs

et al. A 3-Mb physical map of the chromosome region 8p21.3-p22, includinga 600-kb region commonly deleted in human hepatocellular carcinoma,colorectal cancer, and non-small cell lung cancer. Genes Chromosomes Can-cer 1994, 10: 7−14

12 Patel U, Grundfest-Broniatowski S, Gupta M, Banerjee S. Microsatelliteinstabilities at five chromosomes in primary breast tumors. Oncogene 1994,12: 3695−3700

13 Cunningham C, Dunlop MG, Wyllie AH, Bird CC. Deletion mapping incolorectal cancer of a putative tumour suppressor gene in 8p22-p21.3. Oncogene1993, 8: 1391−1396

14 Bova GS, Carter BS, Bussemakers MJ, Emi M, Fujiwara Y, Kyprianou N,Jacobs SC, et al. Homozygous deletion and frequent allelic loss of chromo-some 8p22 loci in human prostate cancer. Cancer Res. 1993, 53: 3869−3873

15 Wu CL, Roz L, Sloan P, Read AP, Holland S, Porter S, Scully C, et al.Deletion mapping defines three discrete areas of allelic imbalance on chromo-some arm 8p in oral and oropharyngeal squamous cell carcinomas. GenesChromosomes Cancer 1997, 20: 347−353

16 Wong CM, Lee JM, Ching YP, Jin DY, Ng IO. Genetic and epigeneticalterations of DLC-1 gene in hepatocellular carcinoma. Cancer Res 2003, 63:7646−7651

17 Kim TY, Jong HS, Song SH, Dimtchev A, Jeong SJ, Lee JW, Kim TY etal. Transcriptional silencing of the DLC-1 tumor suppressor gene by epige-netic mechanism in gastric cancer cells. Oncogene 2003, 22: 3943−3951

18 Pang JC, Chang Q, Chung YF, Teo JG, Poon WS, Zhou LF, Kong X et al.Epigenetic inactivation of DLC-1 in supratentorial primitive neuroectodermaltumor. Hum Pathol 2005, 36: 36−43

19 Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning. 2nd ed. New York:Cold Spring Harbor Laboratory Press 1989

20 Clark SJ, Harrison J, Paul CL, Frommer M. High sensitivity mapping of

methylated cytosines. Nucleic Acids Res 1994, 22: 2990−299721 Yuan BZ, Zhou X, Durkin ME, Zimonjic DB, Gumundsdottir K, Eyfjord JE,

Thorgeirsson SS et al. DLC-1 gene inhibits human breast cancer cell growthand in vivo tumorigenicity. Oncogene 2003, 22: 445−450

22 Zhou XL, Thorgeirsson SS, Popescu NC. Restoration of DLC-1 gene expres-sion induces apoptosis and inhibits both cell growth and tumorigenicity inhuman hepatocellular carcinoma cells. Oncogene 2004, 23: 1308−1313

23 Goodison S, Yuan J, Sloan D, Kim R, Li C, Popescu NC, Urquidi V. TheRhoGAP DLC-1 functions as a metastasis suppressor in breast cancer cells.Cancer Res 2005, 65: 6042−6053

24 Dammann R, Strunnikova M, Schagdarsurengin U, Rastetter M, Papritz M,Hattenhorst UE, Hofmann HS et al. CpG island methylation and expression oftumour-associated genes in lung carcinoma. Eur J Cancer 2005, 41: 1223−1236

25 Bais AJ, Gardner AE, McKenzie OL, Callen DF, Sutherland GR, KremmidiotisG. Aberrant CBFA2T3B gene promoter methylation in breast tumors. MolCancer 2004, 3: 22−37

26 Kim DH, Kim JS, Ji YI, Shim YM, Kim H, Han J, Park J. Hypermethylationof RASSF1A promoter is associated with the age at starting smoking and apoor prognosis in primary non-small cell lung cancer. Cancer Res 2003, 63:3743−3746

27 Golub TR, Slonim DK, Tamayo P, Huard C, Gaasenbeek M, Mesirov JP,Coller H et al. Molecular classification of cancer: Class discovery and classprediction by gene expression monitoring. Science 1999, 286: 531−537

28 Honorio S, Agathanggelou A, Schuermann M, Pankow W, Viacava P, MaherER, Latif F. Detection of RASSF1A aberrant promoter hypermethylation insputum from chronic smokers and ductal carcinoma in situ from breast cancerpatients. Oncogene 2003, 22: 147−150

29 Feinberg AP. The epigenetics of cancer etiology. Semin Cancer Biol 2004,14: 427−432

Edited byMing-Hua XU